Section Iv, 1. the Experiments of Table Ix, C, Repeat Those of a With

duplication of groups; and D repeats those of C with longer exposure.

The sound for A was that of a small organ-pipe (Ut 4) blown by mouth. As in A of the preceding table the observers did not know in a given experiment with which group the sound would be given, but, as before, it was given the same number of times with the first as with the last. For B the multiplied sound was produced by an electric bell with a wooden gong. This was adopted in preference to metal because of the prompt ceasing of the sound after the stroke,--a very necessary condition when this sound accompanied the first group, that it might be clearly connected with its own group. A metal gong was used for the single sound, that the two might not be too unequal in loudness. Its vibrations were deadened by a rubber band, and each bell was controlled by a floor-button. For C and D a higher sound, from the same pipe unstopped, was used in preference to the former, for the reason that in certain experiments performed just previously the lower sound had been used and was presumably very familiar. So in order that the sound might be brought, if possible, afresh to the attention, the change was made.

TABLE IX

A B

132 experiments 44 experiments each with 3 each with 2 subjects. 180 subjects. 88 with 1 subject with 1 subject

Exposure = 1/25 sec. Exposure = 1/25 sec.

Sound\No No Many One No Sound tendency- Sounds Sound tendency_ Subjects 4 1 2

Av.% of difference in favor of 5.4 18.2 2.2

C D

44 experiments each 88 experiments each

Exposure = 1/25 sec. Exposure = 1/4 sec.

Sound No Sound No Sound No Sound No tendency tendency Subjects 1 1 2

20.4 4.6 2.2

The results of these experiments may be summarized as follows: (1) The figures give evidence of but two cases out of eleven where sound was influential. (2) Duplication of groups is not effective in developing evidence of the influence of sound. (3) Increased length of exposure works, as in former cases, to lessen the influence of the modifying factor. (4) The introspections are to the effect that the sound seems to be entirely without influence upon the judgment, beyond the distraction it brings in the earlier stages of work. Sometimes it dropped wholly out of consciousness. Sometimes the distraction seemed to last longer. One observer reported, when D was taken, that he felt as if the sound sometimes increased and sometimes decreased the apparent numerousness. In some other experiments not directly upon this point, but later to be reported, a sound was used; and one observer reported that it seemed to become functionally connected with certain gaps in the groups, as though the puff had blown a hole in the group. Here its effect was of course to emphasize negative factors. It appears thus that the sound might function in opposite directions at different times, somewhat in accord with the particular character of the visual presentation. We should expect, then, to have percentages that look insignificant. (5) We shall not have failed to notice the difference between touch and auditory stimuli in the feeling of influence upon the number-judgment. If we seek a cause for the superior influence of touch, we may perhaps find it in the fact that practical experience has trained us to disregard in any case of judgment such simultaneous presentations as were employed for auditory stimuli; while a definite tap upon the brow is a rather unusual experience likely to attract notice to itself in spite of attempts at abstraction. As one observer said, who took part in both kinds of experiments, the touch seemed more "intimate."

3. The Influence of Kinæsthetic Impression.

The method consisted in the employment of active effort upon a fist dynamometer or a wooden handle during the appearance of one of the groups. The handle was preferable because noiseless. The effort was made with the left hand because the right was used in recording. The amount of it was left to the observer's regulation, with the one instruction that its presence be made decidedly evident but without too great fatigue. The cards of Section IV 1 and 2 were used in the One-Group Apparatus. Similarly again the experiments were repeated with the duplicate-group cards. I present no table here because the figures show practically no influence of the effort. On one subject 176 experiments were made; on a second 132; on a third 88.

It is interesting to note here certain results obtained from one observer when he was in what he described as an active attitude toward the groups, in which he seemed to rouse himself to an unusual pitch of concentration upon the visual situation. This was evidently a condition of increased effort to abstract. Without abstraction he gave 26 to 6 in favor of the effort while with abstraction this tendency had fallen off to 30 to 17. The strength of the tendency is thus strongly indicated. Another observer felt a kind of motor difference between the groups; he expected the effort-group to look larger and felt additionally excited, a scattered activity, while he was passive toward the other group. Perhaps this account puts a little meaning into his small per cent. That his power of abstraction was effective here is hinted by his remark that he felt a difference in the groups even when he judged them equal. The third observer found no subjective evidence that effort modified his judgment.

V. THE "ERRORS" OF EXPERIMENTATION

Throughout the foregoing experiments has been involved the possibility of some one of the three "errors" of experimentation, those of time, space, and distribution, and sometimes all three. Their effect on the results, if it existed, was, to be sure, eliminated in the well-known way, but their existence, if actual, would raise an interesting problem. It was possible, in the case of every group of experiments, to rearrange the tables in such a way as to bring out the evidence for any tendency to overestimate, for instance, the first group as against the second, the right as against the left, or one kind of irregular distribution as against another.

The distribution-error must have a word of explanation. It refers to a tendency to give more wrong judgments in favor of one kind of irregular distribution than of the other kind with which, in a given card, it is mated. In the construction of a set of cards several forms of irregular internal arrangement were used, in order that the judgment might not be one merely of form, and of course on any given card the forms were not the same. Elimination of the effect of these form-differences from the results involved the appearance of any given one as many times in connection with one of the two contrasting factors studied in a given experiment as with the other. Thus two sets of forms were carried through an experimental series--a source of error indeed, but avoidable only by such means as were used to escape the effects of the space-error. Analysis would show which, if either, of the two sets received more judgments in its favor, resulting in further evidence as to the extent to which the judgment of relative number is a function of distribution, and as to the fineness of discrimination for such differences.

Now the tables, when thus rearranged, show that these errors exist to a surprisingly large extent. In many cases their causes, whatever they are, seem to be the controlling factors in the judgment of relative number.

Barring the experiments of Section III, in which the space-error has largely been accounted for, I now propose to gather in one survey all the results of those analyses that have given us the information of the existence of these errors, and all the material of later tables that bears on this point, and to test them by further experimentation. I will begin with the space-error.

TABLE X

Av. % of Av. % of Av. % of difference in difference in difference in favor of favor of favor of

Cases Right Cases Left Cases No tendency

Angier 1 25 6 17.9 9 5.7 Davison 5 26.4 1 10.8 6 6.2 Dunlap 7 18.3 4 4.4 Holt 2 13.6 7 17.6 4 3.2 Hylan 8 23.6 7 6.7 Meakin 2 19.2 1 29.6 8 5.5 Meriam 2 16.5 2 12.9 7 6.8 Moore 2 11 3 16.6 7 3.1 Peterson 1 13.6 1 12.2 9 4.3 Rogers 4 21.9 2 15.9 6 4.5 Rouse 3 15.5 8 5.1 Shaw 3 20.9 5 18 7 6.1 Windate 1 22.8 4 19.5 7 4.5 Yerkes 6 26.6 8 5.2 Henry 1 10 2 16.6 3 8.1 Woods 3 19.3 3 4.8

1. The Space-Error.

Table X presents to us a summary of the values of the space-error tendency. (1) Taken as a whole they fall into all the three classes that are possible; (a) favoring right; (b) favoring left; (c) no marked tendency. (2) There is no observer that does not at some time show a fairly marked tendency. (3) All the observers fall into (c) and all but five into both (a) and (b). (4) More favor the left than the right,--50 to 35. (5) This survey makes it clear that the observers agree neither with themselves nor with each other in the direction of influence exerted by the causes underlying the space-error.

a. Special Experiments to establish the Facts. It might be suspected that irregularities would be more apparent where other factors such as we have been studying enter to complicate the situation from the point of view of pure relative position of the two groups. Table XI presents the answer to this query. The cards used contained groups of gray circles (Gray Darker, Prang) arranged in equal areas of the usual size and shape. The distribution-error was eliminated, though not by duplication, and the small-difference cards were retained. The Two-Group Apparatus was used, with an exposure of 6/5 sec.

TABLE XI

88 experiments with each

Right Left No tendency

Subjects 4 7 3

Av. % of difference in favor of 25.6 23.1 3.4

The results give us again our inevitable three classes, and in many cases a difference-value surprisingly large when we reflect on the simplicity of the conditions. That the omission of complicating features was of importance is shown by the fact that more of the observers (11 out of 14) show a marked error than in any other case. Clearly enough the various factors introduced tend to eliminate the space-error, but when in any case it does enter, it is even then capable of rising to as high a degree on the whole as in the uncomplicated series, as is shown by the fact that in but four cases does the new value surpass the best of the old, and in three of these by a trifling amount.

It is interesting to note that the three cases of minimum space-error show a well-defined tendency to be determined by distribution.

b. Possible Bases of this Error. The outcome of these special experiments is that the factors found in the groups are at least not directly responsible for the situation that we are considering. The divergence among the observers shows this. In hunting after the cause for this apparent influence of side, we look first for changes in the peripheral, and then in the central, processes that precede the judgment. The material used for the experiments of Table XI seems approximately to have equalized all the objective factors in the two groups. How could there be anything further in the peripheral process whereby group could be differentiated from group? The most evident thing is that the visual stimulus is received in a different way from the two groups. There is a definite peripheral mechanism whose factors seem essentially to be two, however variously they may be combined: (a) The relative amount of time given to each group; (b) the order in which the groups are viewed.

The observers were instructed and continually reminded to equalize the amount of attention devoted to the group; but as this is not wholly a voluntary matter, the possibility of failure to conform has to be reckoned with. Experiment must therefore be employed to test the influence of these factors before one can fall back upon a central process as the cause for this tendency to favor a side.

c. Its Relation to Differences between the Groups in Length of Look. The material was the same used for the experiments of Table XI. The method was the same with the following necessary exceptions: The longer exposure was double the shorter (4/5 sec. to 2/5 sec.), and 2/5 sec. elapsed between the two. Further, the experiments were so arranged as to equalize the influence of the order of exposure with respect to both side and relative length. The means for effecting successive exposure took the earlier form described in the introduction to Section II.

TABLE XII

88 experiments with each

Longer Shorter No tendency

Subjects 6 7

Av. % of difference in favor of 18 6.5

These facts are yielded by Table XII: (1) There are but two classes of observers, as no tendency exists to favor the group of longer exposure. (2) The time-error shows a considerably more marked tendency than the length of look, which is indeed somewhat surpassed by the space-error. (3) The persistence of the space-error, even among those that reveal a tendency in length of exposure, shows that the factor of relative difference in length of look cannot account for it. The persistence of it, too, when the order of exposure is controlled, even though the conditions are not wholly adapted to the study of this latter factor, suggest at least that the space-error is independent of even that order; but into this we shall make special enquiry. (4) The judgment of number is independent of the amount of eye-movement devoted to the fixation of the objects in a group. This conclusion, so far as the actual movement is concerned, is established by the fact that so many observers favor the shorter look; and by all the experiments with the One-Group Apparatus where an exposure of 1/25 sec. was used, since that time was too short to admit of movement. That ideated movement is likewise insignificant appears from the fact of marked error arising in the material where the groups were duplicates. Here no motives to different movements could lie in the material.

d. Its Relation to the Order in which the Groups are viewed. Table XIII gives us the results of the enquiry. The experimental conditions were not changed except as to the length of exposure. Each group was given 3/5 sec.; and half the experiments were performed in the order right-left and half in the reverse order.

TABLE XIII

88 experiments with each

First Last No tendency

Subjects 3 9 2

Av. % of difference in favor of 17.5 28.3 1.7

The results may be thus summarized: (1) The order of exposure is notably influential upon the judgment of relative number, giving the usual three classes, with the tendency to overestimate the last group well in the lead. (2) The persistence of the space-error under these relatively simple conditions shows conclusively that it is not a function of the order of exposure. The two are independent variables.

2. The Time-Error.

In pursuit of our enquiry we must survey the facts as they are given in the various experiments already reported and later to be reported. These facts are gathered into Table XIV, which furnishes the following items of significance: (1) All the observers, with the exception of Rouse, show at some time a definite tendency. One case only is given for him in this table, but other experiments not included in the tables from which the present is drawn confirm this fact by the ratio 29 to 30. (2) There is a rather striking consistency in the several observers. (3) The predominance of the last group is marked.

TABLE XIV

Av. % of Av. % of Av. % of difference difference difference in favor of in favor of in favor of

Cases First Cases Last Cases No tendency

Angier 7 19.1 4 4.9 Baldwin 4 20.5 1 3.4 Bell 1 18.2 4 6.3 Davison 2 31.2 Dunlap 1 11.4 1 2.2 Holt 10 19 2 6 Hylan 2 16 5 25.4 5 4.9 Johnston 2 25.2 4 35.5 4 4.2 Meakin 2 39.7 Meriam 1 29.6 1 2.2 Miller 3 17.1 4 5.5 Moore 1 11.4 1 2.2 Olmsted 1 15.2 Peterson 1 17 1 9 Rogers 1 10.2 1 5.6 Rouse 1 1.2 Shaw 6 16.7 5 7 Windate 1 11.4 1 1.2 Yerkes 2 42.7

a. Relation of the Error to the Absolute Length of the Total Exposure. Table XV is set to answer this question. It is unsatisfactory in that but two observers took part in both XIII and XV. The material used for judgment consisted of the same cards used in the earlier experiment, but presented now in the One-Group Apparatus. The time of exposure was changed from 3/5 sec. to 1/25 sec. for each group. As the space-error was eliminated, the tendency to a time-error, if present at all, would presumably have freer play.

But the difference-values of the new table are for the most part very small. We have thus the further fact about the time-error that, under the conditions studied, it appears to be independent of the absolute length of exposure, when the groups are equal in this respect. To this we may add another fact drawn from Table XIV, that with the One-Group Apparatus the time-error is greater on the whole where the groups are differentiated by other factors. Thirdly, the values for Table XIII show that with all complicating factors withdrawn, except the differences in position, the error is at a maximum. This may be significant of the effect of space-differences upon that error, or, more probably, be due to the general difference between work by daylight and work in a dark room by artificial light. We shall be better able to consider this later.

TABLE XV

88 experiments with each of two subjects. 176 experiments with one subject. 154 experiments with one subject. 66 experiments with one subject.

Exposure = 1/25 sec.

First Last No tendency

Subjects 1 4 Av. % of difference in favor of 15.2 6.8

3. The Distribution-Error.

The last of our three "errors" of experimentation is now before us. We may recall once more the meaning the term has had for us in these studies. It points to a tendency discovered by the use of those cards where all objective factors were in the course of a series equalized,--a tendency to mass one's judgments in favor of a particular arrangement of the circles; though each group had been constructed with a view to filling the given area as homogeneously as an irregular arrangement would allow.

As in the two "errors" preceding, so here we must get possession of the facts that gave rise to the present enquiry. Table XVI presents them to us, gathered out of all the tables wherein such a tendency has been technically reckoned with. But first a few words of explanation are needed to make the new table intelligible. Two sets of results are found in its two parts. In each set the particular group-arrangements employed and the frequency of their appearance are exactly the same. The two sets differ, as their headings suggest, in that the material for the second set was formed out of the first by replacing the small-difference cards by those having equal groups. Such a change as this might affect the proportion of judgments given in favor of the two sets of arrangements in a particular series, and these new results are, therefore, no longer fully comparable with the earlier ones. In presenting the directions of tendency in the results, it is impossible here, as in all the similar cases throughout the tables, to name a factor as a standard in whose favor all the judgments in the plus column should be understood as given,--impossible for this reason that, because the very method by which the circles were distributed in the groups, the experimenter was unable to satisfy himself as to the significant differences in the arrangements. All the results, however, when analyzed on this basis, were recorded consistently, so that consistencies and agreements among the observers might be readily apparent. We can now understand in part what Table XVI has to say to us.

TABLE XVI

A No Cases Class 1 Cases Class 2 Cases tendency

Angier 4 19.1 2 6.3 Davison 3 23.1 Dunlap 2 13.7 1 2.2 Holt 3 15.5 1 25 2 4.5 Hylan 1 11.4 2 14 3 6.7 Johnston 5 30.4 Meakin 3 34.9 Meriam 1 16 2 6.8 Miller 5 32.1 Moore 1 39.8 2 5.7 Olmsted 1 27.2 Peterson 3 42.1 Rogers 1 11.4 2 5.7 Rouse 2 14.2 Shaw 6 29.1 Windate 3 30 Yerkes 3 17.8

B No Cases Class 1 Cases Class 2 Cases tendency

Angier 3 2.9 Davison Dunlap Holt 1 22 1 21.6 2 3.1 Hylan 2 50.8 1 11.4 1 8.4 Johnston Meakin Meriam Miller Moore Olmsted Peterson Rogers Rouse Shaw 2 26 1 0 Windate Yerkes

Here as elsewhere the per cents recorded indicate the average per cent of difference in favor of a given class.

Here are the facts, first of A: (1) The only lapses from consistency are confined to two observers; and in both these cases there is but a single break in a uniform trend. (2) With three exceptions all agree in the trend of their difference-values. Of these three--Holt, Hylan, and Moore--the last furnishes but one significant value, and so must be left out of the reckoning on this point. (3) Of the 64 cases, 50 rise above 10%, some far beyond, showing the importance for the judgment of relative number of this factor of distribution. (4) Of the 50 cases, 45 agree in tendency. (5) That with this surprising agreement we have still a few exceptions, adds another item to the growing array of evidence on behalf of the importance of some subjective factor for the number-judgment. As to the nature of this factor we are yet in the dark. (6) To these facts B of this same table adds the further information that the observers inconsistent in the old are not consistent in the new, while the consistent still maintain their record.

a. Analysis of the Experimental Conditions of Distribution. At once we are interested to enquire for the factors underlying these results. To put ourselves upon the right track we must first consider what factors are involved in any such arrangement of objects as we have used in the material for these studies, and then, more precisely, we may ask in what way such arrangements could differ significantly. Finally, by an experimental trial-and-error process, we may solve our problem.

The groups of objects in our material were arranged in an area marked out in each corner by a circle. Within this area the circles were set irregularly, with the result that the group, as a mass of objects distinguished from a homogeneous background, had a more or less irregular outline whose irregularity varied with different internal arrangements. Within its outlines this area presented a mixed pattern of bright and dark. While the total enclosure marked off by the corner circles was always the same and theoretically the relative amounts of brightness and darkness in equal groups was likewise the same, yet practically differences, more or less slight, might enter through the changing character of the rude outlines whose ideal completeness could scarcely be brought out of a black background by the uninitiated. The amount of this difference is sometimes surprising to one whose chief thought of the group has been as vignetted in process of construction. As the objects are pushed toward the edges the central spaces open out; as they are withdrawn toward the interior gaps appear in the margin.

It is not a very easy task to fill an area with objects in irregular arrangement in such a way that no sections of vacancy or filling stand out by contrast against the remainder of the same element. To succeed in this is to fill the area homogeneously. But the chances are good that some vacant patch will get slightly the better of its neighbors or some section of circles will gather a little more closely than the surrounding circles; or perhaps a gap in the outline will be unexpectedly intrusive. Now in a given area the circles of one part cannot become more thickly massed without a corresponding enlargement of the vacancies of the other parts, and of course the converse is as true; but this theoretical situation may be quite out of ken at the moment when the group is seen. Either member of this pair of complements may stand out vividly in the field and its fellow quite escape perception. The very nicety with which in practical affairs we have to make a reliable comparison of this sort shows what suspicion of accuracy the off-hand judgment has bred. And further, the widening of a gap or thickening of the filling in one small part of a group may give a complementary loss to the rest of the group small enough to be unperceived when distributed throughout the larger section.

Two factors must therefore be considered as possibly significant in moving the judgment,--vacancies and filling; and with the former must be reckoned indrawing of the outline. Psychologically, increase in the prominence of either of these factors would be all one with their objective increase. With respect to the direction of their influence upon the judgment of number the increase of vacancies must signify the waning, and the increase of filling the waxing, of the objective number in the group.

It is in advance altogether probable that the results gathered into Table XVI were brought about by these two factors, at least in large part. And we have also in these factors the possibility of two types; for as we saw above, increased vacancies in one part involves increase of filling in another, and conversely. So the interesting question turns upon the altogether disproportional representation of types. Which is the type of the majority?

b. Experimental Test of Hypotheses. The question was put to the test of experiment. This was done by using groups in which now vacancies and now filling were objectively emphasized in contrast with the usual homogeneous group. First the vacancies. A set of cards was prepared after the method previously used to eliminate the distribution-error without duplication of groups on any one card. (See Section II.) In the present case, however, the two sets of arrangements were definitely differentiated as already indicated. One set had a homogeneously filled area, the other a prominent vacancy within or gap in the edge. The size of these variations was kept pretty close to the limit of noticeableness, that the increase in compactness of the other portions might be as slight as possible. It was experimentally necessary to free the material as far as might be from ambiguity, and practically important to avoid rousing the suspicions of the observers and the resulting reflections. It seemed very likely that the strength of the tendency shown by the distribution-error was due to its appearance in situations where the observers knew that other factors were being tested.

The general method already described was used in preparing the groups that gave objective prominence to compacted parts of the filling. To fulfil the conditions outlined above was here even more difficult than in the first set; and the cause will appear in the sequel. The small-difference cards were omitted and the One-Group Apparatus used.

A further attempt was made to head off reflection by a subterfuge. It had been found that, among the factors whose influence on the judgment had been studied, hearing had been as little effective as any. So the small stopped pipe used for those experiments was again brought into service and the error resulting eliminated in the usual way. Incidentally our new tables will thus give us further information about the effect of this factor, though of course under conditions that are theoretically highly unfavorable, since we are forcing upon the attention of the observers other factors that experience has shown them only too ready to seize upon. So if a tendency traceable to the factor of hearing should appear, we ought perhaps to give it somewhat more than its face value.

TABLE XVII

Exposure = 1/25 sec.

Homogeneous Vacant No tendency

Angier 50 Baldwin 53.4 Bell 52.2 Holt 44.4 Hylan 51.2 Johnston 56.8 Miller 4.2 Shaw 29.6

Exposure = 1/4 sec.

Homogeneous Vacant No tendency

Angier 51.2 Baldwin 55.6 Bell Holt 13.6 Hylan 52.2 Johnston 62.6 Miller 25 Shaw 14.8

88 experiments each Exposure = 1/25 sec.

Homogeneous Compact No tendency

Angier 39.6 Baldwin 35.2 Bell 3.4 Holt 27.2 Hylan 39.6 Johnston 44.4 Miller 16 Shaw 2.2

44 Experiments. 88 Experiments.

The per cents recorded indicate the average per cent of difference in favor of a given factor.

Now we are ready to inspect the results. Table XVII, A is the outcome of the attempt to emphasize vacancies. Its experiments with 1/25 sec. exposure were repeated with one of 1/4 sec. as Table XVII, B, shows. In Table XVII, C, the emphasis of compactness is concerned.

For convenience we may again resort to a summary outline in extracting the meaning from these tables. First Table XVII, A. (1) All the observers but one agree in favoring the homogeneous, most of them with very high difference-values. (2) Miller alone gives no tendency, and his notes show a conflict between the increased vacancy and the increased compactness. In other words, his discrimination was too keen for the material. Under the circumstances he constitutes no exception to the conclusion that the vacancy objectively emphasized was the cause for an underestimation of its group.

From Table XVII, B, we learn the following: (1) All the observers save one favor the homogeneous group, in most cases by large values. (2) The difference in the length of exposure seems to have no significance for this tendency, since, while Holt and Shaw decline, Miller rises in the scale.

Table XVII, C, gives us these facts: (1) The difference-values have noticeably fallen off. (2) We have again the customary three classes, but with homogeneous leading as in the earlier tables. (3) By his present favoring of the compact, Miller has now appeared in all three classes, while Holt has developed the preference for the compact that was budding in XVII, B. (4) The presence of four well-marked preferences for the homogeneous shows that the vacancies in the compact group were more significant for the number-judgment than was the increased compactness of the filling, and that in spite of the experimental effort to the contrary. (5) The decrease of this tendency and the growth of the opposing, indicates that the judgment is determined in either case by the more vivid factor.

The conclusions to be drawn from these facts lie close at hand. (a) The results in Table XVI, with their disproportionate division into classes, were evidently due to the tendency of three observers to note the filling and of the rest to be concerned with the vacancies. (b) The judgment of relative number under these conditions is primarily a judgment of vacancies. (c) The subjective factor of vividness determines the direction of error, and may attach to either vacancies or filling, though it usually attaches to the former.

It may not be out of place here to speculate a bit as to the probable cause for so close a dependence of the number-judgment upon what has no number, so to say; upon an object that has no standing in the official conclusion. The situation seems to be fundamentally based upon the conditions that determine contrast. In a homogeneous field no part stands out. Introduce a small object quite different in brightness or complementary in color and the attention is drawn instantly to it, but internal differences in its content are quite lost in the common quality by which it differs from the ground. A case somewhat analogous is furnished by our material, particularly in the One- and Two-Group Apparatus. The small group is so unified by its contrast with the field that internal differences must be made out with relative effort. Now internal differences are necessary to the numerical character demanded of it, and they can be brought out in no way save by attending to the vacancies and so isolating parts in the threatening unity, each in a kind of space-matrix. The most careful observer could not do better on his way to truth; and that is why the error was so much larger when the factor of space-differences was studied.

That group is normally the more numerous in which the vacancies are less completely developed under observation. We say "normally" here by virtue of the speculation just completed as to the best method of attaining a judgment objectively true. For a man thus proceeding, our proposition is a sound statement of fact, to which the following results of our experiments bear witness. (a) The experiments recorded in Table XII on Relative Difference in Length of Look shows no exception of a value equal to 10% to the general statement that all tendencies, when any existed, were in the direction of favoring the shorter group. The shorter the time of exposure the less completely would the vacancies develop. (b) Table IV, E, shows that without exception the darker group tends to be judged the more numerous. (c) Table XXI shows for each subject that in a shorter exposure the absolute number seems considerably greater than in a longer exposure.

No comment seems necessary to concentrate the force of such evidence. If we carry our proposition to the detailed results of our separate studies in factors of distribution, we shall find that it helps us to understand those few exceptions to the general trend of observers as they appear in Tables II and XVII. The exceptions there favored the groups in which compactness of parts went along with certain large vacancies. Possibly enough they refused to fall in with the objective analysis, and, disregarding the prominent vacancies, devoted themselves to a development of the vacancies within the compacted parts.

c. The Factor of Hearing. The time-error analyses of the experiments of Table XVII have already contributed their facts to the special section dealing with that error. But one or two interesting facts have remained unnoticed in the sound-analysis. In the experiments of Table XVII, A, there is a single case of marked tendency to favor the sound group. With the lengthened exposure of B, this tendency, as usual, disappears; but returns in C to some extent and two other observers share it. A fourth markedly favors the group without sound. So the experiments of this last table present as marked external evidence as we have for the influence of hearing upon the judgment. These facts are presented in Table XVIII.

TABLE XVIII

A 44 experiments with each of 4 subjects, 88 with each of 3.

Exposure = 1/25 sec.

No- No Sound Sound tendency Subjects 1 6

Av. % of difference in favor of 27.2 4.1

B 88 experiments with each of 3 subjects, 44 with each of 3.

Exposure = 1/4 sec.

No- No Sound Sound tendency Subjects 6

Av. % of difference in favor of 5.9

C 88 experiments each

Exposure = 1/25 sec.

No- No Sound Sound tendency Subjects 3 1 3

Av. % of difference in favor of 12.9 20.4 4.5

It is a further curious fact, well sustained by these same experiments, that where there is some confusion, each of the factors present has a better chance to determine the judgment. The values for both time-error and sound rise higher for the majority in C than in A or B.

VI. THE INFLUENCE OF FACTORS IN THE SAME SENSE-FIELD UPON THE JUDGMENT OF ABSOLUTE NUMBER

The nature of the enquiry that we have been pursuing through so many pages is such that it may be raised exactly as well in the case of absolute as in that of relative number. There appears to be no reason why in this new field the results should not be exactly comparable with those in the old, to be taken indeed as a kind of test for the interpretation to be put upon the old. Without a single exception, unless it were imposed by a technical difficulty, all the earlier factors could be studied with the new purpose. Our practical interest to go to such lengths would depend pretty largely upon the results of first attempts. If wholly confirmatory, these would probably suffice.

The experimental conditions were of the simplest. The 3-8 in. steel balls of Section III were again pressed into service as objects for the number-judgment. They were thrown loosely into a fixed black frame, 20 cm. square. To avoid suggestive noises, its undersurface was made of a thick piece of felt covered with black cloth; and the whole rested of course on a black-topped table. The exposures were 2 sec. long, timed by watch-ticks. Between experiments the observer held a cardboard screen between him and the objects. When conditions were ready for a new judgment, closing his eyes he lowered the screen, opening his eyes again at the word of command and shutting them at the close of the experiment.

Of course the observers felt that their judgments were for the most part extremely vague. With small numbers they had a greater feeling of confidence. Yet altogether it was surprising with what readiness an absolute number-judgment would spring up in the presence of any given collection whatever within the limits set by the experimental series. Sometimes the observers thought that they made rough calculations on the basis of the filling in a unit of area. So far as this held it would tend to cut off the more astonishing departures from correctness, and it would probably advantage the smaller groups more than the large. Still it was entirely too rough a method to prevent the influence of the factors introduced, as the results will show. There was no time for systematic counting, which, in any case, the observers knew to be forbidden.

The figures in which the observers reported their judgments of absolute number have a value that is chiefly qualitative. The marked inconsistencies and disagreements are our guarantee for this statement. With all the observers there was but the loosest association between group-appearance and number-name. The innumerable variations in internal space-relations were of course responsible. For one observer a particular name probably had a quantitative significance far in excess of its value for another observer in this respect. To one man 100 might have meant about the same as 60, for example, to his neighbor. On the whole they were parsimonious; but Baldwin decidedly not.

A more or less constant influence was exerted on any given judgment by the comparison of the presented group with the traces of the preceding still in mind. The observers felt, however, that the judgment was largely independent of such comparison, and its fluctuations give some credence to this feeling.

The numbers chosen ranged by fives, from 25 to 100. In four cases a number was immediately repeated that rough suggestions as to the definiteness of the judgment and its dependence upon the actual number might be gained. These were indeed but rough suggestions, since, with certain exceptions to be noticed later, the arrangement was disturbed between times; but they made possible a closer watch upon the flickering of the judgment than could be kept by a mere repetition of the series. In the latter case it might be unstable and yet relatively firm in the other. A standard series is here recorded. Its order was determined by drawing the numbers out of a heap, but the repetitions were inserted arbitrarily.

1. 95 2. 25 3. 35 4. 65 5. No change 6. 30 7. 90 8. 85 9. 45 10. 100 11. 50 12. No change 13. 60 14. 40 15. No change 16. 70 17. 80 18. 55 19. 75 20. No change

1. Absolute Number under Standard Conditions.

An indispensable preliminary for the present study is the establishment of a standard. Unless we know something in advance about the characteristics of the judgment of absolute number in relatively simple conditions, we shall be unable to tell what influence, if any, to attribute to the modifying factor in later experiments. Having then decided as to the general conditions under which we will study the problem we must make these the standard conditions of our work; and having discovered the nature of the judgments given under them, measure up to these results in all that is to follow. These standard conditions have already been set forth in the introduction to this section. The results are recorded in Tables XIX and XX.

TABLE XIX

KEY: St = Standard Sc = Scattered Co = Compact

Subject = Baldwin Subject = Miller

Trials with each number 6 3 3 4 2 3

Original Numbers St Sc Co St Sc Co

25 1 0 -6 -10 -6 -4 30 3 5 -5 -12 -7 -6 35 10 7 -3 -14 -4 -7 40 10 13 -8 -15 -8 -11 40 10 8 -8 -14 -10 -12 45 10 18 -3 -21 -13 -10 50 17 23 2 -19 -15 -17 50 15 22 2 -17 -15 -14 55 27 40 0 -18 -3 -12 60 19 33 5 -20 -23 -17 65 27 53 2 -18 -3 -7 65 24 57 5 -13 -13 2 70 38 77 0 -19 8 -8 75 31 73 0 -24 10 -20 75 28 78 -5 -24 8 -8 80 36 83 -5 -14 5 -18 85 54 85 5 -19 -8 0 90 54 90 5 -13 8 -3 95 48 73 5 -30 20 -15 100 61 87 7 -13 10 -7

The figures recorded are the average of the algebraic sums of other figures that represent the difference between the actual and the estimated number. Fractions are replaced by an added unit, if the value is 1/2 or over.

Baldwin never underestimated the scattered group, and only once the standard; but 31 times the compact. Miller only once overestimated the standard group, and but 6 times the compact; but 13 times the scattered.

Turning to these tables we notice at once, as characteristic of all the observers, the following facts: (1) Wide variation from objective correctness. (2) A far wider discrepancy with the larger numbers than with the smaller. Miller does not wholly agree here. His judgments by series show inconstancy, tending at first to follow the rule, but in the last two series to a maximum error near the middle. Certain remarks of this observer suggest that possibly in the latter case reflection as to the convenience of certain actual numbers for manipulation may have had influence. The three earlier series of Hutchison conform to the rule. The remainder, on the contrary, show no definite progression in tendency. It should be noted here that both Miller and Hutchison were more inclined than the other two observers to rough calculation. The effect of its adoption or of increased practice in it is shown by the disappearance of the characteristics of the earlier series. We have thus in these two cases a doubleness of standard that we must not fail to consider in our later comparisons. (3) There is a pronounced instability of judgment, as shown by the fluctuations for the same number in different series, and especially in successive judgments, of the same in any given series. (4) There is a general tendency to judge in multiples of five. That there should be any splitting of fives, particularly in the large numbers, might be regarded as mere caprice. Not so did it seem to the observers. They were conscious of an apparent absurdity in it where judgments were necessarily so vague; but they insisted that this stood for a kind of qualitative shading in the perception which threw out the choice of the round numbers just above and below. (5) The number is on the whole underestimated, three observers agreeing in this respect; but the fourth shows a very large and consistent tendency in the opposite direction.

In spite of the manifold special inconstancies and disagreements, these general tendencies are decidedly well-featured in the results. We may say that we have found a kind of standard illusion that will serve us for a guide through our later studies.

2. The Influence of Distribution.

TABLE XX

Subject = Hutchison Subject = Olmsted Trials with each number 3 2 4 First Second Original Standard Standard Mixed Small numbers Series Series Sizes Sizes 25 -5 -2 -3 -1 30 -4 0 -5 0 35 -6 -5 10 1 40 -11 -5 -5 -13 40 -12 -9 -13 -11 45 -7 -4 -15 -10 50 -14 -10 -20 -12 50 -13 -8 -15 -15 55 -10 -7 -20 -11 60 -19 -18 -25 -20 65 -15 -7 -20 -5 65 -15 -10 -10 -18 70 -14 -3 -15 -20 75 -21 -17 -28 -23 75 -25 -20 -20 -24 80 -24 -17 -30 -19 85 -17 -13 -35 -13 90 -13 -7 -25 -20 95 -25 -13 -40 -23 100 -22 -13 -35 -24

6 2 3 Mixed Small Standard Sizes Sizes_ -8 -8 -12 -11 -14 -16 -8 -5 -14 -18 -17 -20 -13 -18 -22 -18 -20 -18 -19 -18 -23 -22 -15 -27 -24 -28 -23 -22 -23 -27 -28 -28 -33 -22 -28 -23 -27 -30 -28 -32 -23 -27 -33 -25 -32 -27 -23 -33 -27 -38 -38 -33 -40 -37 -39 -55 -40 -36 -15 -37

For the meaning of these figures see under Table XIX. Hutchison overestimated the standard group only 5 times, never the mixed-size group, and 9 times the small-size group. Olmsted never overestimates at any time.

The first of the modifying factors to be considered has to do with the arrangement of the objects. Hitherto they had been thrown loosely into the frame. Now in successive studies they were, first, well scattered over the surface and, second, brought together into several compact nuclei. The last arrangement was adopted in preference to that of a single mass as being less open to comparison with preceding judgments and to judgment on the basis of form and size of group.

The results are shown in Table XIX: (1) The effect of scattering the objects is very markedly to raise the apparent number. Baldwin's preceding overestimations soar still higher; while Miller's former tendency to underestimation is checked to such an extent that 13 overestimations appear. (2) The effect of compacting the objects is just as markedly in the opposite direction. Baldwin gives 31 underestimations, and Miller reverts in a measure to his former type. (3) When similar arrangements were up for study in relative number we found two classes of observers, one favoring the compact, the other the scattered. The present results of Baldwin and Miller put them into the latter class.

3. The Influence of Complexity of Group-Content.

This new factor of complexity in the content of the group was realized experimentally by making up the collection out of steel balls of two sizes, 1/8 in. and 3/8 in. The former looked almost infinitesimal beside the latter. The same objective numbers were still maintained and divided between the two sizes except where in so doing a five must be broken. In such a case the extra five went to the larger balls.

The results are found in Table XX. Olmsted shows no definite influence of the new factor. Hutchison, however, shows a very evident decrease in his estimations, when comparison is made with his second standard series. With the earlier series the new results rather closely correspond. That the latter are not simply a vacillating reversion seems fairly clear from this observer's account of his method. The small balls, he says, did not distinctly come in visually. To his judgment of the large he added an amount based on a very insecure estimate of the small. The number of the latter seemed from time to time pretty constant.

This situation corresponds very fully to that in the investigation of the same factor by use of a group of mixed colors, where relative number was in question. (Section II.) The tendency there discovered was to neglect the other colors in favor of one which thus surpassed the others in vividness. There as here the mixed group seemed smaller.

4. The Influence of Size of Objects.

A study of this factor was made possible by substituting for the usual objects steel balls of a smaller size, 1-4 in. The results are contained in Table XX. They are not so striking as those obtained in our study of distribution. Still the influence of this new factor is evident, in the reduction of the apparent number. Olmsted shows this more generally for the smaller numbers. We find it in Hutchison when we compare the new results with the second standard series. This tendency to underestimation increases in the two final series of the present set. At the beginning of these two he remarked that he thought he had been overestimating the group. This tendency of smaller size to reduce apparent number was found true for the majority of observers in our earlier study of relative number.

5. The Influence of the Length of Exposure.

I found that in relative number the shorter the look the more marked was the influence of certain factors. Reports of the observers making this seem highly probable happened in this way: When working with the One-Group Apparatus in relative number the shutter of the camera would occasionally stick, leaving a group exposed beyond its usual time. The effect of this upon some but not all the observers was to cause a noticeable shrinking in numerousness. Of those questioned, the only one failing to notice this effect is included among the observers in this new study.

To test this possibility resort was had to the One-Group Apparatus as affording a more satisfactory means for getting different lengths of exposure of small absolute magnitude. Cards were prepared containing a single group of larger area (67 × 82 mm.) than had been used for relative number. The objects were the usual white circles. Each corner was marked as usual; and, by reason of the number involved, the outline of the area was more regular than had been true in the earlier work. The number of circles on each card varied by steps of two from 16 to 30, giving eight cards in all. The series was arranged irregularly as before, and two of the cards repeated immediately upon their first presentation, making ten experiments in one set. The order of the series follows:

1. 24 2. 22 3. 26 4. No change 5. 18 6. 28 7. 16 8. 20 9. 30 10. No change

Two time-magnitudes were used for comparison,--1-25 sec. and 1 sec. The latter was managed with bulb exposure. All the experiments with the shorter time were made before those with the longer had been begun. The results are given in Table XXI. So far as the material is comparable, we may include in our comparison the standard experiments of Tables XIX and XX with 2 sec. exposure.

TABLE XXI

KEY: A = 1/25 sec. B = 1 sec.

Baldwin Miller Hutchison Olmsted Actual Numbers A B A B A B A B Number of trials with each number 5 6 5 5 5 4 4 4 16 6 9 -2 -3 11 -1 2 -5 18 8 10 -4 -1 12 -1 -3 -5 20 34 19 1 3 19 1 1 1 22 37 17 6 6 14 6 4 -2 24 49 26 12 5 14 10 5 0 26 70 33 15 8 12 9 5 4 26 79 37 21 9 14 4 9 4 28 93 42 29 11 18 7 8 7 30 106 52 38 12 18 5 19 10 30 103 50 45 11 20 4 19 4

For the meaning of these figures see note under Table XIX.

The outcome may be thus summarized: (1) The apparent number is inversely proportional to the length of exposure. The tables show a perfectly clear progression from 2 sec. to 1-25 sec. All those that formerly underestimated are brought into the opposite class. (2) The results of the earlier experiments are confirmed on the whole with respect to the occurrence of greater errors with the larger numbers. (3) Baldwin's overestimation reaches astonishing heights. (4) These new facts for absolute number are quite in accord with Table XII, where, under the conditions of interpretation laid down, the tendencies were wholly in favor of the shorter look.

The issue of these tentative experiments in absolute number confirms the teaching of our studies in the related field. Absolute number, like relative, has been found largely subject to a modifying influence of certain factors. In the new field, too, distribution has asserted its supremacy among these, and similar effects of shortening exposure have been observed. There has been variation among the observers and some shifting of tendency, both of which point as before to the coöperation of some subjective factor in our results. Indeed the whole situation, as opened by these preliminary studies, indicates a theoretical interpretation that for both fields is at bottom one. So to an attempt to reach such an interpretation the next section will be devoted.

VII. THEORETICAL DISCUSSION

1. The Fact of Modification.

That such an influence upon the judgment of number should have been exercised by the factors considered seems in many cases to receive an adequate account on the principle of association. Our practical experience in the simultaneous variability of number and certain other characteristics of a group of objects has been such as to lead us into illusions when the two no longer vary together. In such a case, when we have no time to count, we are actually led to see a group as smaller or larger in accordance with the variations perceived in the associated factor. This interpretation is supported by the fact that on the whole the space-factors were more markedly influential in creating illusions than were any others. For those cases, however, in which the modification was effected by a factor unconnected with number, as color, or the simultaneous stimulation of other senses by irrelevant objects, it appears that the mere occurrence of greater total stimulation during the appearance of one group is sufficient to create illusion, either through failure of the observer to discriminate between the relevant and the irrelevant, or because he is led through fear of disturbance to overemphasize the other group.

2. The Direction of Modification.

The foregoing account of the general fact throws no light upon the direction of the influence. Why should a given factor make a group seem more numerous and not less? Why should it affect one man in one way and his neighbor in another? Why should it vary with the same man at different times? Appearances no less contradictory than these are what we must face in carrying a theoretical account to completion. The following propositions with appended commentary are offered in satisfaction of these requirements.

a. Differences in vividness among the factors determine differences in number.

Our study of the factor of distribution in Table XVII, where it was possible in a measure to control the vividness, furnishes evidence for this proposition. Introspective reports in other cases confirm this view by showing that the direction of the attention, the popular way of stating our proposition, was the determining feature. This will receive further support in our discussion of the following proposition.

b. If the vivid factor or complex be positive, i. e., associated in experience with the numerous, or if it be neutral, its group will seem the more numerous. If negative, i. e., associated in experience with the few, its group will seem the less numerous.

The experiments upon the effect of distribution support this proposition, especially as set out in Table XVII. When the vacancies in a given group were made vivid, the other group seemed more numerous; when its filling surpassed in vividness, the judgment was given for it. We have other confirmation in the fact that lengthening the time of exposure reduced the absolute number. Take also this note of one observer on the material in Table II, C:

"I noticed that I had set the open spaces in the outlined group over against the lack of them in the homogeneous, without paying much attention to the nearness together of the spots in the lines of the outlined. Then for a time my attitude was quite vacillating. I found my attention drawn to the nearness together of the spots in part of the outlined group so strongly that if I did not turn it voluntarily to the fact that the other was filled without any large open spaces, I was led to call almost any outlined group the larger. Toward the end of the experiment I got back into my original attitude, in which the outlined group seemed to have its spots hardly more thickly arranged in any part than the homogeneous, and to have also the bare spots and so to be the fewer."

That the vividness of a neutral factor or complex increases the apparent number was suggested by comments of the observers. One observer reported of the material in Table IV: "The greater brightness of red gave it more importance. The natural thing seemed to be to give the red the judgment. The gray fought more for recognition." And again: "The red seems a vitalized space and the dots more omnipresent, also the red lasts longer in memory and is there more vivid, so that often in cases of doubt, where the decisive comparison was made in memory, the red may have been given the vote. Often there was an immediate unanalyzed feeling that if the groups had been both of the same color, the judgment would have been for the gray." In both cases his results showed this tendency. Another observer, whose results agree with the former, found that his eye was directed involuntarily toward the red.

This fact was put to a special test. In the material of Table II, B, a card, in which the pattern group had an appearance strikingly different from the normal, was introduced, the two groups being objectively equal. With three observers the effect was overestimation of this group, and with a fourth, the suppression to equality of a previous overestimation of the opposite group. This fact, together with the observers' comments, seems to justify the conclusion that the vividness of a neutral factor or complex was the determining condition of the judgment. That the observers did not all show positive results in this experiment may be set down to the difficulty in controlling the subjective conditions of vividness. Of course the space-relations within the new pattern were different from those in the old. The only justification for taking no account of these is the character of the introspections themselves.

It should be said of the red group that beside its vividness it had characters mentioned by other observers that might independently have made its number seem greater. It was called "dazzling," "blurred," and its area seemed increased. In this respect the effect of the color should be discussed as a special case of distribution or object-size.

The vividness tested in this special way seems due to contrast, in the one case with surroundings, in the other case with the expected. Such a judgment is very far removed from the normal bases, rather more so, it would seem, than even those where a group had sound or touch accompaniment; for in the latter case there could be no question about the "moreness"; the only doubt could be about its legitimacy. Of the precise extent to which this cause of vividness has operated throughout our studies, even where spatial differences have been concerned, we cannot be sure. The patterns of the materials in B and C of Table II seem to offer that possibility. That it should enter anywhere opens, indeed, the entire field.

c. The observers fall into the following classes on the basis of the character of the association:

(1) Relatively fixed association,

(a) involving correct adjustment to objects (vividness of relevant factors);

(b) involving incorrect adjustment to objects (vividness of irrelevant factors).

(2) Relatively unstable association.

It will be remembered that in the case of nearly every factor studied under Relative Number, we found three classes of observers,--those favoring one group, those favoring the other group, and a so-called "no-tendency" class. The bases of classification were, first, the relative constancy in the character of the error, and, secondly, its direction. In this third or "no-tendency" class were really lumped off two kinds of observers, not separated at the time because our special interest did not demand it. Along with those that gave large errors in both directions was a much smaller class that gave a relatively large proportion of correct judgments; but could never claim any one observer all the time. In the new classification of Proposition c this mixed composition is recognized by dividing it between (1) (a) and (2). The prime condition of correct judgment is asserted to be one in principle with that of the illusion,--namely, vividness, but in this case vividness of relevant factors. Our original "tendency" classes both fall under (1) (b).

Proposition c is merely an attempt to apply the principles of association and vividness to an organization of our results. The types in question have no hard-and-fast connection with any particular observer; they rather represent a kind of ideal fixation of opposite tendencies playing through all.

3. The Time-Error.

So far the time-error has been left without interpretation. The chief facts to be considered were: (a) Divergence of error and general trend in favor of last. (b) Individual inconsistencies. (c) Occasional absence.

We are in a position now to invoke at once the principle of vividness to account for the existence of the error and the vividness of recency to account for the predominant tendency to favor the last of the two groups exposed. In this respect this error may be classed with the effect of red. That is to say, a factor or complex, directly through its vividness and not indirectly through its association with the numerous or the few, draws the judgment after it. Here the content of the group is the effective thing, not the character of the vacancies.

But the observers do not all agree in the direction of the time-error nor are they always consistent. Here we shall get help from a proposition offered in the discussion of the distribution-error in which it is asserted that the group seems the more numerous in which the vacancies are less developed under observation. We have already noted a decided tendency to depend in judging upon the vacancies. Let us suppose that the two groups presented in succession differ with respect to the success of the observer in developing these vacancies. If this be true, that difference may well depend upon the occurrence of maximal attention during the exposure of but one of the groups. The tendency of the majority to overestimate the second group suggests that the attention is likely to be at a maximum when the experiment begins. If, on the other hand, it ripen later, or if the observer seek to rescue the second group from relative unclearness, then we should have the first group overestimated. The time-error would disappear for those that could attend alike to both. Clearly enough this account is decidedly hypothetical.

4. The Space-Error.

Our attempt to reduce this error to one of time in some form was proven a failure. The facts brought out indicate that at the bottom is some subjective factor thus far not isolated. This factor is not a preference going directly with right- or left-handedness because on the surface at least it runs in the observers independent of such asymmetry. A single bit of available introspection would seem, however, to point to some relation of that sort; for one observer, who favored the left, felt that a group on that side gained an importance that was somehow due to the greater absolute value of a weight in the left than in the right hand. Even if this be decisive for him it will still be inapplicable to errors in the opposite direction unless we assume that with variations in bodily energy the emphasis is cast now in one, now in the other, direction, after the analogy of those two types of man to be found in our social experience, for whom respectively mountains are molehills and molehills mountains. Such successive differences in type in a single individual would then find an intelligible account in the shifting tides of that bodily energy. It is to be noted that the observer just quoted once, but only once, made a decisive reversal of his error from left to right.

It may occur to some one that the use of two observers sitting side by side may have given them a preference for one position of the groups. In the first place care was taken that both groups should be as readily seen from one point of view as from the other. Secondly and chiefly, there is no regularity among the observers in this respect.

It is not unlikely that a chance aspect of a particular group develops an emphasis that gives the mechanism of subjective adjustment a particular bent that for a time is relatively independent of the objective situation. Still the fact that there are some cases of persistence in type is rather damaging to this assumption and speaks rather for the earlier one. That one, if true, seems indeed adequate to account for the situation. As an hypothesis it accords with analogous physiological facts; but its weakness lies in imposing the burden of a strong tendency upon asymmetrical differences that may be in comparison relatively slight. Finally, these studies furnish no proof that the bodily condition of an observer of a particular type corresponds to the demands of the hypothesis.

Summary: 1. The estimation of relative number in the visual field is modified by group-area, internal distribution, order, and complexity in group-composition; by the size, form, color, brightness, and complexity of the individual members; and by the character of the environment. It is further modified by factors contributed by the objects through other senses, as in active pressure, special differences in pressure character, active weight, and that complex of muscular and spatial factors arising when a group is observed under the condition of eye-muscle strain. The judgment is also influenced by factors outside the group in the field of touch, but not in that of kinæsthetic impressions.

2. On the whole the most influential factors were those lying in the space-characters of the groups; while those of least moment were contributed by other objects in other fields of sensation. Hearing was very nearly ineffectual.

3. In very many cases the observers fell into three groups, one of no-tendency, and a second and third showing opposite tendencies with respect to the factor investigated.

4. With a majority of observers there is a tendency to underestimation in the judgment of absolute number, though with a single observer the tendency is directly the reverse. Scattering the objects increases, and compacting diminishes, the apparent number. The smaller the size of the objects the fewer, under conditions, do they appear; while heterogeneity in group-composition lessens the number for one observer and has no apparent effect upon the other.

5. The apparent absolute number of objects is inversely proportional to the length of exposure of a group; and in relative number the influence of a factor was on the whole greater for shorter exposures.

6. The marked tendency to a space-error was found to be independent of differences between the groups in the length of look, and of the order in which they were viewed.

7. The distribution-error is grounded in a fundamental tendency to base the judgment of relative number upon the character of the vacancies in a group; though a secondary tendency to depend upon the filling was shown to exist. The subjective factor of vividness, attaching now to one and now to the other of the foregoing factors, determines which shall be operative, though it usually is joined to the first. The ground for the primary tendency may very well be the necessities imposed upon discrimination by the material. The contrast effect between the large black background and the brighter objects tends to unify the latter, which, to be discriminated as a number, must be split up by an emphasis of the vacancies.

8. The time-error is possibly due to differences in power to dismember the groups exposed in succession in one experiment, while its variations in direction seem adequately accounted for by differences in the time at which attention becomes maximal during the progress of a single test.

9. The ground for these facts of modification is found in the strength of the association between these several factors and the elements that signify number.

10. The basis for the different tendencies found among the observers is the differing vividness among several factors. If the vivid factor is associated with the idea of numerousness, or is in this respect neutral, its group will seem more numerous. If it has been associated with the idea of fewness, its group will seem less numerous. The difference between the two classes of "tendency" and "no-tendency" lies in the fact that for the latter either only correct clues are vivid, or else there is so frequent an alternation in vividness of opposing incorrect clues that through any given series no tendency appears, while for the "tendency" class misleading clues are without shifting in the ascendant.

At the time when these experiments were completed, no work precisely upon this problem had been published. Since then, however, Dr. J. F. Messenger has issued a monograph entitled The Perception of Number (Psych. Rev., Mono. Supp., vol. 5, no. 5), of which certain parts fall within the scope of the present studies. He was concerned with the estimation of absolute number and was primarily interested to discover the nature of the number-judgment. The reader of both articles will find agreement between the results and interpretations here recorded and such part of Messenger's work as has been a common object of study,--viz., the factors of distribution and size.

FOOTNOTES:

TIME-ESTIMATION IN ITS RELATIONS TO SEX, AGE, AND PHYSIOLOGICAL RHYTHMS

BY ROBERT M. YERKES AND F. M. URBAN

The desirability of a statistical study of time-estimation was suggested to us by a note concerning "sex-differences in the sense of time" which was published in Science recently by Prof. Robert MacDougall. By comparing the time-estimates of groups of men and women consisting of fifteen individuals each, MacDougall discovered that for intervals of from one quarter of a minute to a minute and a half, the women exhibited a far stronger tendency to overestimate than did the men, and were at the same time markedly less accurate. The nature and extent of the overestimation discovered by MacDougall are indicated by the results presented in the accompanying table from Science. The numerals, 1, 2, 3, and 4 refer to different fillings of the intervals (listening to reading, marking letters, etc.), the signs + and - to over- and under-estimation respectively.

Period, One Minute.

Sex 1 2 3 4 Men +29" + 1.3" +22" - 3.5" Women +66 +22.0 +80 +24.0

These apparent sex-differences in time-estimation demand further attention, first, because the number of individuals studied by MacDougall, as he recognized, is too small to establish the fact of the existence of such differences, and, second, because if the differences really do exist they should be studied in their relations to age and the fundamental physiological rhythms.

It seemed probable that further investigation of this subject might reveal some important facts concerning the development of the ability to estimate time in the individual, the significance of various conditions for time-estimation, the psychology of sex, and the relations of rhythms to personal affinities, antipathies, and motor capacities.

In this report the results of a statistical study of the sex-differences in time-estimation are discussed, and in later papers we shall present the results of investigations of the relations of time-estimation to age and to individual and sex rhythms, and attempt to work out a convenient and serviceable rhythm-formula. The need of such a formula for expressing individual rhythms is obvious, as is also the need of comparative studies of individual and sex rhythms.

TIME ESTIMATION

Name Place Age Date

ORDER OF TESTS TIME IN SECONDS

Time of Male Female Intervals. (17 years) (17 years) No. 1 No. 9

1. Idleness 108̋ 70̋ 120̋ 2. Reading 36 30 118 3. Writing 72 36 60 4. Estimating 18 15 30 5. Reading 108 90 68 6. Idleness 36 35 60 7. Writing 18 10 10 8. Estimating 108 100 125 9. Reading 72 100 66 10. Idleness 72 75 58 11. Writing 36 25 22 12. Estimating 72 60 60 13. Reading 18 14 15 14. Writing 108 130 59 15. Estimating 36 30 41 16. Idleness 18 10 18

How did you estimate the interval when you were asked to estimate it as accurately as you could?

n o f e y m i q r s a d r g d e s t k n w e r a x u p x z y o n d f n o d c a e h p m a l g s r w y t b c k p s o n q a r v q c o m p v r i c p k t o s n q z r l x m i h u v o q g P p f u t o i c n g s c a r n o t c d a a o b i a r s a d e r w o a i e r g l c r t h f s o r a e n s i o c r b x g r z b h o w l t s

Number of letters counted 85 88 Pulse-rate 72 81

The experimental data now to be considered were obtained as follows. Record-sheets of the form reproduced above were printed, with blanks for age and name of subject, place, date, for sixteen judgments of time-intervals (numerals 1 to 16), for a statement of the subject's method of estimating time, for the number of letters counted in thirty seconds, and for the pulse-rate. Four intervals were used, 18, 36, 72, and 108 seconds, and for each of these intervals judgments were taken under the four conditions designated on the record-sheet as idleness, reading, writing, and estimating. In the experiments the intervals were not given in order of regular increase or decrease of the length of interval, nor were all the judgments for any one interval taken together, but instead, for the purpose of avoiding the influence of expectation of a particular interval or filling, they were arranged irregularly in the order of column two of the record-sheet. This column, as also columns three and four, which are specimen series of judgments for a male and a female respectively, of course were not printed on the record-sheets which were supplied to the subjects.

The experimental procedure was as follows:

(1) Each subject was given a record-sheet.

(2) The experimenter was provided with a record-sheet on which the time of the intervals numbered from 1 to 16 was given. Care was taken that the subjects should not know the length of the intervals before the experiments.

(3) The beginning of each interval was indicated to the subjects by the word "start" uttered distinctly by the experimenter; the end, by the word "stop."

(4) Before beginning the sixteen tests the experimenter gave a thirty-second interval as a standard of judgment. The experiment then proceeded with only sufficient pause between judgments to allow of the recording of estimates by the subjects.

(5) During the filling called "idleness" the subject did not pay special attention to the estimation of the time, but instead permitted his attention to wander.

(6) During "reading" the experimenter read aloud to the subjects.

(7) During "writing" the subjects wrote from the dictation of the experimenter.

(8) During "estimating" the subjects judged the interval as accurately as they could, by whatever method they chose except the use of a time-piece.

(9) Each subject recorded his judgment of the length of an interval in seconds at the appropriate place on the record-sheet as soon as the interval was ended.

(10) The question following judgment number 16 on the sheet was answered as soon as the sixteen judgments had been recorded.

(11) The number of letters counted in thirty seconds was determined by the use of the lines of letters at the bottom of the sheet. The subjects began at the left of each line and counted singly as many letters as they could between the "start" and "stop" signals of the experimenter. They then marked the last letter counted and immediately recorded, in the place provided on the record-sheet, the number of letters counted.

(12) The pulse was counted by the experimenter immediately after the experiment when possible and the rate recorded on the sheet.

(13) The experimenter avoided delays, interruptions, or other irregularities in the course of the series of experiments.

The materials for our discussion of the sex-differences in time-estimation consist of the judgments of 251 males and 274 females. The majority of the males were students in Harvard College, the majority of the females, in Radcliffe and Smith Colleges. The remainder of the records were obtained in Ohio State University, Pomona College, and West Chester State Normal School. The authors gratefully acknowledge their indebtedness for assistance in the obtaining of records to Professors A. H. Pierce, T. H. Haines, W. H. Scott, D. R. Major, A. M. Smith, H. A. Miller, and B. T. Baldwin. The males ranged in age from 17 to 23 years, the females from 17 to 20. The total number of judgments, the distribution of which among the various ages is shown in Table 1, is 4014 for the males, 4375 for the females.

Despite the fact that our experiments are open to the criticisms of all work done under variable conditions and by different experimenters, it cannot be doubted that the results indicate certain sex-differences in time-estimation which suggest additional problems. For the present we refrain from interpretations for the most part and state merely the statistical results of the investigation.

Previous studies of the "time-sense" and the conditions which influence time-estimation suggested to us the desirability of examining our data with reference to (1) sex-differences in estimates of intervals, (2) age-differences, (3) the influence of different fillings, and (4) differences dependent upon the length of the interval. The results have been studied, therefore, with reference to the significance of sex, age, filling, and length of interval, but as no marked age-differences appeared, the detailed tables which were constructed to exhibit the results for the subjects of each year of age have not been printed.

In all the tables the results for males and females are presented separately. The judgments for the sixteen intervals are arranged with reference to the length of the interval, not in the order in which they were taken; all the 18̋ intervals, for example, are grouped (Table 2). The letters I, E, R, W, refer to the fillings of the intervals.

TABLE 1

NUMBER OF SUBJECTS, AGE, SEX, AND NUMBER OF JUDGMENTS

Males

Age No. of subjects No. of judgments

17 yrs. 16 256 18 27 432 19 40 639 20 67 1071 21 50 800 22 35 560 23 16 256

Totals 251 4014

Interval No. of judgments

18" 1004 36" 1003 72" 1003 108" 1004

Total 4014

Females

Age No. of subjects No. of judgments

17 yrs. 73 1160 18 57 911 19 64 1024 20 80 1280

Totals 274 4375

Interval No. of judgments

18" 1092 36" 1094 72" 1096 108" 1093

Total 4375

A general survey of the individual records, all of which for any one year and sex were tabulated, for convenience of examination, on a single large sheet of coördinate paper, showed that the judgments vary within a wide range and are very inexact. Table 2 exhibits the number of correct judgments for each sex, interval, and filling. Of the 4014 male judgments only 96 (2.39%) were correct; of the 4375 female judgments only 46 (1.05%) were correct. The number of correct judgments decreases as the length of the interval increases. For the 18-second intervals there were 7.37% for the males, 2.48% for the females, while for the 108-second intervals there were only 0.10% and 0.37% respectively.

TABLE 2

FREQUENCY OF OCCURRENCE OF CORRECT JUDGMENTS

Males

18̋ 36̋ 72̋ 108̋ I E R W I E R W I E R W I E R W Σ % 29 26 11 8 5 6 3 0 3 3 0 1 0 1 0 0 96 2.39 Totals 74 = 7.37% 14 = 1.40% 7 = 0.70% 1 =0.10% 96 2.39

Females

7 15 1 4 2 4 1 0 2 5 0 1 2 1 0 1 46 1.05 Totals 27 = 2.48% 7 = 0.62% 8 = 0.73% 4 = 0.37% 46 1.05

List of abbreviations which occur in the tables.

Σ always designates the sum of the results of the column which it heads.

I, E, R, W refer respectively to the intervals of idleness, estimating, reading, and writing.

The % sign refers to the value of the result in question in terms of the total number of judgments.

C refers to the results of the letter-counting test.

The male judgments for the 108-second intervals range from 11 to 300 seconds. If random guesses be made within these limits the probability of the occurrence of right guesses (108") would be 1 in 290; therefore among 1004 guesses (the number of male judgments for 108-second intervals) 3.5 would be right. In the experiment only one judgment of the 1004 was correct. For the other intervals, with the exception of 18 seconds, the number of correct judgments is only slightly greater than random guessing would have given. Both males and females, however, show considerably more correct judgments for 18-second intervals than the number of probable right guesses. Within the range of the male judgments and for their number 16.9 right guesses might be expected, for the females 10.9. In contrast with these numbers the experiments furnished 74 and 27 correct judgments respectively.

It is noteworthy that for those intervals which are most frequently correctly judged, not only is the number of correct judgments greater for the males than for the females (the ratio of the percentages is about 3 to 1), but the ratio of the number of correct judgments to the probable number of right guesses is also greater for the males.

The female judgments vary within a wider range and are less often correct than the male. For the latter the total number of correct judgments is more than twice that for the former.

Another interesting fact concerning the judgments of the time-intervals is that certain numerals occur in the last place of a judgment more frequently than we should expect if their occurrence depended on random guessing. Tables 3 and 4 exhibit the results of an analysis of the data made for the purpose of studying this fact. In Table 3 the frequency of occurrence of the digits 0, 1, 2, 3, etc., in the last place of the male judgments is given for each filling under the four intervals. For example, the digit 0 occurred in the last place of the male judgments for the interval reading 36 seconds 98 times, as we learn by referring to the first line and third row of the second column of Table 3.

Examination of Tables 3 and 4 shows at once the marked preference of the subjects for 0 and 5. The percentage of male judgments which end in 0 is 41.50; of female 58.51. Similarly the percentages of occurrence of the digit 5 for the males is 24.41, and for the females 23.11. Only two of the other digits (2 and 8) occur with a frequency of over 5%.

Among the 4014 male judgments 0 occurred as a final digit 1666 times, 5, 980 times. Among the 4375 female judgments 0 occurred 2560 times, 5, 1011 times. In the male judgments 0 occurred about four times as often as it would in random guessing; in the female, almost six times as often.

Comparison of Tables 3 and 4 indicates that the occurrence of 0 is 17.10% greater for the females, while that of 5 is 1.30% greater for the males. The sum of the percentages of occurrence of 0 and 5 for the males is 65.91, therefore the probability that a male judgment ends in one of these digits is almost twice that in favor of any other digit. For the females the sum of the same percentages is 81.62, and the probability of occurrence of 0 or 5 is therefore more than four times that of the other eight digits.

Tables 3 and 4 show that even numbers occur more frequently than uneven as final digits. Of the total number of judgments 2461 (3063) end with even digits and 1553 (1312) with uneven.

TABLE 3. FREQUENCY OF OCCURRENCE OF THE NUMERALS (0 TO 9) AS FINAL DIGIT OF THE JUDGMENTS COMBINED RESULTS FOR ALL MALES

18̋ 36̋ 72̋ I E R W I E R W I E R W

0 70 57 72 76 131 77 98 90 124 74 120 125 1 5 11 9 9 10 9 8 7 1 18 5 2 2 20 15 21 18 5 19 16 15 11 20 5 14 3 9 17 14 10 8 11 10 8 6 14 12 4 4 8 12 14 9 3 14 12 3 4 14 8 7 5 67 58 56 63 62 54 60 85 72 59 62 73 6 13 17 12 13 9 14 11 6 10 9 7 7 7 15 22 14 12 6 15 11 10 5 14 13 4 8 36 30 29 31 15 19 16 20 11 17 11 11 9 8 12 10 10 2 18 9 7 7 12 8 3

108̋ I E R W Σ % C

0 151 121 127 153 1666 41.504 21.92 1 4 11 10 2 121 3.014 7.57 2 9 12 9 8 217 5.406 13.15 3 3 12 14 3 155 3.861 7.17 4 6 18 9 3 144 3.587 7.57 5 54 45 52 58 980 24.414 10.36 6 8 7 6 4 153 3.812 8.76 7 4 7 10 8 170 4.235 7.17 8 8 13 5 9 281 7.004 8.76 9 4 5 9 3 127 3.164 7.57

TABLE 4. FREQUENCY OF OCCURRENCE OF THE NUMERALS (0 TO 9) AS FINAL DIGIT OF THE JUDGMENTS COMBINED RESULTS FOR ALL FEMALES

18̋ 36̋ 72̋ I E R W I E R W I E R W

0 120 117 106 126 169 124 182 151 176 145 175 194 1 4 8 3 2 6 7 1 3 2 3 2 1 2 17 18 12 13 9 22 7 11 7 11 3 4 3 4 9 9 11 5 9 3 7 6 15 10 4 4 2 11 13 10 1 6 4 2 9 14 3 1 5 89 63 91 85 62 61 60 79 56 55 65 63 6 9 12 11 6 5 13 7 5 4 6 3 3 7 5 5 7 7 4 9 5 3 2 7 3 3 8 16 22 17 12 10 13 5 10 10 9 6 1 9 5 9 5 1 3 9 0 2 2 9 4 0

108̋ I E R W Σ % C

0 226 197 196 186 2560 58.515 26.14 1 1 5 2 0 50 1.143 9.09 2 4 11 6 3 158 3.611 5.30 3 3 7 4 7 113 2.583 5.68 4 2 4 2 2 86 1.966 11.36 5 32 44 49 57 1011 23.109 14.39 6 0 9 3 1 97 2.217 6.44 7 0 5 2 5 72 1.646 6.44 8 5 11 8 7 162 3.703 10.61 9 1 10 2 4 66 1.508 4.55

In order that the probability of the occurrence of even and uneven numbers may be calculated, those judgments which end in 0 and 5 must be subtracted from the total number of judgments, for the occurrence of these two digits is apparently due to a constant influence. The problem may be formulated thus. First, what is the probability that a judgment is determined by the constant influence in favor of 0 and 5? Second, what is the probability of even and uneven numbers, when the influence in favor of 0 and 5 is eliminated? The calculated probability of 0 or 5 is 0.65919 (0.81622) and therefore the probability that a given judgment is not determined by this influence is 0.34081 (0.18378). The probable limits of these numbers are 0.00505 (0.00395).

After the subtraction of those judgments which end in 0 or 5, there remain 1368 (804), of which 795 (503) are even and 573 (301) uneven. The probability of an even number is 0.58115 (0.62562) and the inverse probability of an uneven number is 0.41885 (0.37438). The probable limits of these numbers are 0.00900 (0.01027). There are therefore even chances that the percentage of occurrence of even numbers is between the limits 57.215 and 59.015 (61.535 and 63.589), or outside these limits.

Statistical studies have already proved that in random guessing even numbers occur somewhat more frequently than uneven. It is therefore worthy of notice that in these results the frequencies of even numbers are not uniformly greater than that of uneven; for with the exception of the digit 6 in the female judgments, the digits next to 0 and 5, i. e., 9 and 1, 4 and 6, occur with least frequency.

In the case of the number of letters counted in a half minute, also, it appears (see last column (C) of Tables 3 and 4) that 0 and 5 occur more frequently in the last place than chance would lead us to expect. In contrast with the results for the time-judgments, in the same tables, the percentages of occurrence of the various digits in counting present less marked differences. For the males 3 and 7 occur least frequently, for the females 2 and 9.

To sum up the results of our examination of the materials with reference to the occurrence of digits in the final place of the judgments, the order of decreasing frequency of the various digits is 0, 5, 8, and 2. Of the others 3 and 7 occur more frequently than 4 and 6, with one exception. The statement that even numbers in general occur more frequently than odd must be modified by the statement that in these results the digits next to 0 and 5, namely, 9, 1, 4, and 6 occur with least frequency. These statements hold for both males and females, but for the latter the frequency of occurrence of 0 is far greater than for the males.

These results clearly indicate that the judgments are not random guesses. In seeking further for some explanation of the surprising frequency of occurrence of judgments which end in 0 or 5, we discovered that certain numbers occur very frequently, namely, the multiples of 15, 30, and 60. In order to exhibit this tendency quantitatively Tables 5 and 6 have been constructed.

In these tables will be found tabulated the number of times 15 and multiples of it which are not also multiples of 30 or 60 occur for any given interval and filling. Likewise are tabulated the frequencies of occurrence of 30 and multiples of it which are not multiples of 60, and finally, of 60 and its multiples. The numbers as they occurred in the three categories run as follows:

15 30 60 45 90 120 75 150 180 105 210 240 135 270 300 165 330 360 195 390 420

Fifteen and its multiples, as given above, are arranged in one division of the tables, thirty and sixty each in its own separate division. The line at the bottom of the tables marked Σ gives the frequency of occurrence of these three groups of numbers for all the subjects and for each interval and filling.

As is shown by the percentage of frequency columns of the tables, in no instance do the multiples of 15 constitute less than 19.52% of the male judgments and 23.81% of the female judgments. The lowest frequency for any of the four intervals is 20.32% of the total number of judgments. The maximum frequency for the males (43.03%) and for the females (56.57%) is for the interval idleness 108 seconds. That the male and female maxima should fall on the same interval is interesting.

TABLE 5

FREQUENCY OF OCCURRENCE OF 15, 30, 60, AND THEIR MULTIPLES

Males

15 30 60 Σ % Average

{I 48 5 0 53 21.12 {E 42 9 0 51 20.32 18̋ {R 43 8 0 51 20.32 20.32 {W 45 4 0 49 19.52

{I 29 60 7 96 38.25 {E 17 41 5 63 25.20 36̋ {R 22 41 6 69 27.41 29.57 {W 36 28 5 69 27.41

{I 29 18 46 93 37.05 {E 25 7 34 66 26.29 72̋ {R 28 26 33 87 34.66 34.70 {W 28 42 32 102 40.80

{I 25 31 52 108 43.03 {E 16 23 20 59 23.51 108̋ {R 22 30 39 91 36.25 36.16 {W 25 37 43 105 41.83

Σ 480 410 322 1212 % 11.96 10.21 8.02 30.12

TABLE 6

FREQUENCY OF OCCURRENCE OF 15, 30, 60, AND THEIR MULTIPLES

Females

15 30 60 Σ % Average

{I 58 30 0 88 32.47 {E 30 35 8 73 26.64 18̋ {R 56 26 3 85 31.02 28.48 {W 41 24 0 65 23.81

{I 31 55 39 125 45.62 {E 21 36 17 74 27.21 36̋ {R 31 63 33 127 46.35 38.86 {W 41 45 13 99 36.26

{I 29 28 60 117 42.70 {E 17 18 52 87 31.75 72̋ {R 26 37 52 115 41.97 41.60 {W 30 51 56 137 50.00

{I 12 45 98 155 56.57 {E 21 27 58 106 38.83 108̋ {R 23 36 84 143 52.19 47.83 {W 24 31 64 119 43.75

Σ 491 587 637 1715 % 11.22 13.42 14.56 39.22

If no influence worked in favor of the multiples of 15 in these experiments only one judgment in thirty would be 15 (3.33%) and only one in sixty, 30 or 60 (1.67%). For the probability of the occurrence of 15, 30, and 60 is 1/60 + 1/60 + 1/30 = 1/15, as is obvious from the fact that among sixty consecutive numbers (1 to 60) there are four which are multiples of 15. According to probability we should expect multiples of 15, 30, and 60 to occur 268 times among the 4014 male judgments and 292 times among the 4375 female judgments. As a matter of fact there are 1212 such judgments for the males, 1715 for the females. The probability that a male judgment is a multiple of 15, 30, or 60 is 0.3012 (probable error 0.0049); for a female judgment the probability is 0.39199 (probable error 0.0050).

These statistics indicate that the subjects are constantly and strongly influenced in favor of judgments which are simple fractions of a minute. Closer inspection of the tables gives some suggestion of the nature of this influence.

Comparison of the four intervals (Tables 5 and 6) with respect to the occurrence of simple fractions of a minute shows that the frequency of such numbers increases rapidly as the length of the interval increases. The various percentages of frequency for males and females and for the four intervals are again presented here for convenience of comparison.

18̋ 36̋ 72̋ 108̋

Males 20.32% 29.57% 34.70% 36.16% Females 28.48 38.86 41.60 47.83

As is obvious from these figures both frequency and rate of increase are far higher for the females than for the males.

Examination of the percentages (totals) at the bottom of Tables 5 and 6 reveals another remarkable sex-difference; for the frequency of occurrence of 15 and its multiples regularly decreases for males from the 15 to the 60 class, whereas for females it regularly increases.

Undoubtedly the time-judgments of these experiments were strongly influenced by thought of the conventional time-unit, the minute, for in all quantitative work there are errors in favor of the standard of measurement and simple fractions thereof. In the present instance this tendency to favor the unit was strengthened, perhaps, by the giving of a half-minute interval as a standard for comparison at the beginning of the tests.

Two explanations of the sex-differences above mentioned are suggested by our study of the data. One is the fact that the females are less exact than the males; the other that they generally overestimate the intervals, whereas the males often underestimate them. One's estimate of an interval is determined partly by confidence of accuracy. The longer an interval the less we feel able to estimate it accurately, and, as a consequence, the more frequently it is judged as the same as the time-unit or a simple fraction of that unit. The females are less exact in their estimates than the males, and less exact for long than for short intervals, and as an accompaniment of their inexactitude we find the frequent occurrence of multiples of 15, 30, and 60.

But confidence of ability to estimate accurately must be considered in connection with the fact which suggests our second explanation, namely, that the female estimates are higher than the male. Tables 7 and 8 show that the females almost invariably overestimate the intervals rather largely, while the males sometimes underestimate considerably. The range of the male judgments is from 1 to 300, of the female from 1 to 400. Obviously the chance of occurrence of 15, 30, 60, and their multiples varies with the range. The greater the range the greater the probable frequency of 30 and 60 in comparison with 15. In random guessing the probabilities of the occurrence of 15, 30, and 60 for long and short intervals is the same, but our results show that this is not true in the case of these time-estimation judgments. It seems possible, therefore, that the sex-differences referred to are due to the fact that the intervals seem longer to the females, and that, therefore, a feeling of greater inexactitude than would be felt for shorter intervals leads to the choice of simple fractions of a minute more frequently than in the male judgments and more frequently for the long than for the short intervals.

It is of interest in this connection to note that the length of a second is usually underestimated by females, overestimated by males. The average number of seconds counted in half a minute by twenty men and twenty women was as follows:

Men. M. 30.4, M.V. 8.7, R.V. 34.94. Women. M. 38.9, M.V. 10.6, R.V. 36.70

These figures would seem to indicate that the overestimation of the intervals of these experiments by the females is due to the use of a time-unit which is shorter than that of the males (although presumably of the same length).

We cannot with certainty say whether inaccuracy of judgment stands in the relation of condition or consequence of the occurrence of simple fractions of a minute, but it would appear that the female tendency to overestimate is responsible for the sex-differences already noted. For whatever be the facts concerning longer intervals the second as judged by the female is considerably shorter than that of the male.

Since a complicated periodicity of frequency in the distribution of the judgments is exhibited in the results of Tables 3-6 it is obvious that the distribution-curve will have a tertiary mode for each number ending in 0 or 5, a secondary mode for 15, 30, 60, and their multiples, and a primary mode which may or may not coincide with one of the secondary or tertiary modes. Extreme irregularity is characteristic of the distribution-curve. Different groups of judgments, as, for example, those for the two sexes, those for the different intervals, etc., give somewhat different forms of distribution, for the frequency of occurrence of 0 and 5, as well as of the multiples of 15, is variable.

These facts are important in connection with the selection of an interval for the construction of the distribution-curves, in that they indicate how large the interval or class of the distribution-curves and tables should be.

It is clear from the results of Tables 3 and 4 that the smallest interval which can be of value is 10 seconds, for a smaller interval would necessarily exhibit irregularities due to the greater frequency of 0 than of 5. The question is whether the interval can be so enlarged, without the loss of all details of the nature of the distribution, that every class will represent the influence of the same conditions. For this purpose only three intervals are possible: 10, 30, and 60 seconds. Of these 30 and 60 are undesirable because the interval 60 gives classes which are so large that all details of the distribution are lost, while 30 exhibits only a few details without doing away with the periodicity due to the preference for multiples of 60.

The further question remains, with which digit should the interval end, in order that uniformity of conditions for the various classes may be gained? Theoretically there are ten possibilities, but of these all except two, 0 and 5, are excluded by reason of the unequal frequency of the various digits already discussed. In favor of 0 is the fact that all the classes thus formed are of equal size, i. e. 1-10, 11-20, etc., whereas for 5 the first class would differ from the others in being only half as large, 1-5. This, however, is only a slight disadvantage, for there are very few judgments which fall in this class. On the other hand, since 0 is the final digit of most frequent occurrence, classes ending in 5 have the advantage of placing the value of greatest weight in the middle. On the whole it seemed desirable to arrange the judgments in 10 second classes, beginning with the class 1-10. But for purposes of comparison the male judgments have been distributed in classes of 10 seconds, which end in 5, 1-5, 6-15, 16-25, etc.

TABLE 7

DISTRIBUTION OF MALE JUDGMENTS IN 10" CLASSES

Classes 18̋ 36̋ 72̋ 108̋ C I E R W I E R W I E R W I E R W

1 - 10" 31 16 69 136 3 1 4 18 1 1 1 4 0 0 0 0 0 11 - 20 174 179 142 108 40 34 45 97 4 2 5 23 1 2 1 7 0 21 - 30 34 47 35 7 99 106 98 82 15 9 26 65 5 0 6 23 0 31 - 40 7 7 3 56 69 69 28 25 16 34 47 10 2 6 31 1 41 - 50 4 2 2 37 23 22 18 41 43 54 33 20 9 19 28 0 51 - 60 1 9 13 9 6 64 68 53 35 31 17 27 32 7 61 - 70 5 3 0 0 37 42 28 18 27 26 22 19 8 71 - 80 0 1 3 1 23 34 16 9 33 34 35 20 34 81 - 90 1 1 1 15 15 15 8 30 40 43 29 58 91 -100 0 9 8 7 1 23 35 25 17 65 101 -110 0 2 6 3 3 10 26 17 6 33 111 -120 0 8 5 5 2 29 16 17 16 29 121 -130 0 3 1 2 0 5 15 7 8 10 131 -140 0 1 1 1 1 6 9 5 0 4 141 -150 1 0 1 0 5 5 8 4 1 151 -160 1 0 2 3 2 2 1 161 -170 1 0 2 1 0 2 171 -180 1 1 5 4 5 2 181 -190 0 1 2 1 191 -200 3 2 2 2 201 -210 0 1 1 0 211 -220 0 0 0 0 221 -230 1 1 0 0 231 -240 2 0 1 0 241 -250 1 1 1 251 -260 0 0 261 -270 0 0 271 -280 0 0 281 -290 0 0 291 -300 1 1

Totals 251 251 251 251 251 250 251 251 251 251 251 250 251 251 251 251 251

As a result of these groupings of the male judgments it appeared that the former method gives a far more regular distribution than the latter. In view of this result and the above considerations, Tables 7 and 8 were constructed by the use of 10-second classes, beginning with 1-10. In these tables (column C) the distribution of the letter-counting results has been included for convenience of comparison of the two kinds of judgments as to form of distribution.

As instances of the general form of distribution of the judgments the curves have been plotted for letter-counting, Fig. 1 A. (Males ----, Females, ... ,) for idleness 36 seconds, Fig. 1 B, and for idleness 108 seconds, Fig. 2. The distribution of the letter-counting judgments in classes of 10 is very regular in comparison with that of the several time-estimation judgments. For idleness 36 seconds there are several fairly distinct modes, and for idleness 108 seconds the modes are still more numerous and more marked.

TABLE 8

DISTRIBUTION OF FEMALE JUDGMENTS IN 10̋ CLASSES

18̋ 36̋ 72̋ 108̋ Classes I E R W I E R W I E R W I E R W C

1 - 10 72 20 88 133 5 6 5 33 2 0 1 6 0 1 0 2 0 11 - 20 114 119 109 94 33 29 40 83 7 5 15 25 2 2 7 10 0 21 - 30 50 87 54 29 70 57 73 74 16 13 25 52 4 1 3 27 0 31 - 40 17 17 10 6 51 87 53 30 24 21 24 30 2 1 11 14 0 41 - 50 10 14 7 4 40 35 39 20 35 18 37 32 14 6 17 34 2 51 - 60 2 10 3 1 43 30 40 14 48 44 47 53 30 18 35 36 5 61 - 70 1 3 0 3 11 7 5 2 28 50 20 21 8 11 23 22 18 71 - 80 3 2 1 2 5 8 2 10 29 32 23 14 14 21 20 14 57 81 - 90 0 0 1 1 9 3 6 2 27 29 28 14 41 35 35 19 67 91 -100 0 2 0 1 4 2 1 16 11 15 5 25 35 17 11 50 101 -110 0 0 0 1 1 1 4 7 6 0 9 24 7 6 25 111 -120 0 0 2 3 6 3 17 20 10 8 52 38 33 16 19 121 -130 2 0 0 1 0 3 4 3 3 4 15 6 8 13 131 -140 0 1 1 0 1 2 1 4 6 9 6 6 5 141 -150 0 1 0 1 5 4 4 2 15 13 16 7 1 151 -160 1 1 0 0 0 4 2 0 3 7 2 2 2 161 -170 0 0 0 1 1 0 0 3 2 0 1 171 -180 1 0 0 7 4 7 4 22 14 16 22 181 -190 0 0 1 0 1 0 2 2 3 0 191 -200 0 0 1 1 3 0 3 5 6 6 201 -210 1 0 0 0 0 0 1 0 1 0 211 -220 0 0 1 0 0 1 2 0 0 221 -230 0 0 1 0 0 1 0 6 2 231 -240 0 2 1 0 1 3 3 2 4 241 -250 0 1 0 0 3 0 0 251 -260 0 0 2 0 0 1 261 -270 0 0 1 0 0 0 271 -280 0 0 0 0 0 0 281 -290 0 0 0 0 0 0 291 -300 1 2 4 2 1 1 301 -310 0 0 0 0 311 -320 1 0 0 0 321 -330 0 0 0 0 331 -340 1 0 0 1 341 -350 0 0 351 -360 1 1 361 -370 0 371 -380 0 381 -390 0 391 -400 2

Totals 271 274 274 273 274 273 274 273 274 274 274 274 274 273 274 272 264

Tables 7 and 8 show that the range of the judgments increases with the length of the interval judged, and that the modal class is always much nearer the lower than the upper limit. Asymmetry is characteristic of the distribution of organic data, and in certain instances, as for example writing 18 seconds, males, the choice of a 10-second class interval results in extreme asymmetry, and one is reminded of the tables which Fechner gave as examples of his logarithmic method in statistics.

It is not to be expected that a method of grouping should be found which will give regularity of distribution throughout, but it is important that there should be regularity about the mode. In the table of distribution for the males (Table 7) all the intervals from idleness 18 seconds to reading 72 seconds are regular. The remaining intervals, with the exception of estimating 108 seconds, are irregular.

Trial proves that for these intervals increase of the size of the class to 30 seconds is sufficient to give regular distributions, as is obvious from Table 9. Grouping by 30-second classes gives regularity for most of the female judgments, but for idleness 108 seconds and writing 108 seconds there are still slight irregularities, as Table 10 indicates.

Tables 7 and 8 show that the distribution is far less regular for the females than for the males. The fact that it becomes regular when the class interval is increased to 30 seconds suggests that the irregularities of distribution which appear in the tables are due to those influences which favor simple fractions of a minute and not to the small number of judgments.

Having now noted certain important characteristics of the time-estimation judgments and the nature of their distribution, we may examine the arithmetical means and other statistical quantities which have been determined for our data. Those quantities which have been determined for the several ages, intervals, and fillings as well as for the sexes are: (1) The Mean (M. in tables), (2) the average variability (M. V.), (3) the positive variability (+ V.), (4) the number of judgments with positive variation (No. + V.), (5) the negative variability (- V.), (6) the number of judgments with negative variation (No. - V.), (7) the relative variability (R. V.) = M + V/M × 100.

Since the sums of the positive and the negative variations are equal, it is possible to make certain of the accuracy of the means and average variabilities by comparison of the + V. and - V. As this was done in all cases we feel confident of the reliability of the statistical quantities presented in the tables.

In Table 11 have been arranged the various quantities as determined for the males and females for each interval. The values given in this table are averages of the determinations made for the several ages separately.

TABLE 9

DISTRIBUTION OF CERTAIN MALE JUDGMENTS IN 30̋ CLASSES

Classes. W. 72̋ I. 108̋ R. 108̋ W. 108̋

1- 30 92 6 7 30 31- 60 115 61 52 91 61- 90 35 90 100 68 91-120 6 62 59 39 121-150 1 16 20 12 151-180 1 9 7 6 181-210 3 5 3 211-240 3 1 0 241-270 1 1 271-300 1

250 251 251 251

TABLE 10

DISTRIBUTION OF FEMALE JUDGMENTS IN 30̋ CLASSES

18̋ 36̋ 72̋ 108̋ Classes I E R W I E R W I E R W I E R W 1- 30 236 226 251 256 108 92 118 190 25 18 41 83 6 4 10 39 31- 60 29 41 20 11 134 152 132 64 107 83 108 115 46 25 63 84 61- 90 4 5 2 6 25 18 13 14 84 111 71 49 63 67 78 55 91-120 0 2 0 3 8 9 5 37 38 31 13 86 97 57 33 121-150 2 0 2 2 1 9 10 8 9 25 37 28 21 151-180 1 2 0 0 8 9 9 4 28 23 18 25 181-210 1 0 2 1 4 0 6 7 10 6 211-240 0 2 3 0 1 5 5 8 6 241-270 0 1 0 3 3 0 1 271-300 1 2 4 2 1 1 301-330 1 0 0 0 331-360 1 1 1 1 361-390 0 391-420 2

Totals 271 274 274 273 274 273 274 273 274 274 274 274 274 273 274 272

TABLE 11

MEANS, ETC., FOR EACH SEX, INTERVAL AND FILLING

KEY: Ma. = Males Fe. = Females

M. M. V. +V. No. +V. Ma. Fe. Ma. Fe. Ma. Fe. Ma. Fe.

{I 17.7 20.74 5.4 10.35 5.8 15.48 15.9 22.75 18" {E 19.5 25.55 4.9 9.84 6.0 13.98 16.9 23.75 {R 15.5 18.46 4.9 9.14 6.3 13.03 14.6 24.25 {W 11.5 15.58 3.7 8.56 4.6 11.83 14.3 25.75

{I 33.3 42.81 9.1 16.55 12.6 20.63 14.2 28.00 36" {E 33.1 41.54 8.4 15.23 10.7 23.34 14.2 23.00 {R 32.1 41.71 8.4 16.42 9.7 20.78 16.3 27.75 {W 24.7 30.10 9.0 14.71 10.9 22.08 14.4 23.50

{I 63.3 73.00 17.2 27.20 20.9 33.72 15.7 27.75 72" {E 63.1 77.13 16.0 26.56 18.8 37.27 16.9 24.75 {R 57.9 70.78 17.3 30.30 20.9 39.28 15.2 26.75 {W 51.2 54.93 19.8 24.21 23.7 29.47 15.0 38.25

{I 92.7 113.37 29.8 40.13 35.3 44.08 15.2 32.00 108"{E 99.8 114.88 26.3 36.38 29.3 49.26 14.9 26.25 {R 90.1 100.47 28.3 40.18 33.9 47.90 15.3 34.25 {W 75.5 87.45 32.4 45.33 40.8 61.12 14.9 34.25

-V. No. -V. R. V. Ma. Fe. Ma. Fe. Ma. Fe.

{I 5.7 7.78 20.0 45.00 30 49.83 18" {E 4.4 7.69 19.0 44.75 25 39.10 {R 4.4 7.07 21.2 44.25 31 49.26 {W 3.2 7.14 21.6 42.50 33 54.55

{I 8.1 14.01 21.7 40.50 27 38.32 36" {E 7.2 11.57 21.7 45.25 25 36.66 {R 7.5 13.84 19.6 40.75 26 38.84 {W 8.7 11.24 21.5 44.75 36 48.64

{I 15.5 22.83 20.2 40.75 27 36.87 72" {E 14.7 20.98 19.0 43.75 27 34.43 {R 15.7 24.91 20.7 41.75 30 42.77 {W 16.0 21.26 20.9 36.75 37 43.84

{I 25.9 37.87 20.7 36.50 32 35.34 108"{E 24.4 29.51 19.9 42.00 26 31.57 {R 24.8 34.60 20.6 39.75 31 39.95 {W 27.5 35.86 19.9 42.75 42 51.67

The following facts are revealed by comparison of the results for the two sexes. Without exception the means for the females are larger than those for the males. All but one of the sixteen intervals (E 18̋) are underestimated by the males, whereas all but six are overestimated by the females. The amount of over- and under-estimation is given in Table 12. In every instance the females overestimate in comparison with the males. The mean variability is very much greater for the females, as is also the relative variability. If variability be taken as a measure of reliability of judgment the males are far superior to the females.

As is obvious from Table 12, both under- and over-estimation increase with increase in the length of the interval. For 18̋ intervals they are least, for 108" intervals greatest.

The influence of the fillings is marked. The writing intervals without exception are judged as shortest; reading gives the next shortest intervals, while sometimes idleness, sometimes estimating, comes third. In order of increasing length of average estimates of the intervals the fillings stand: writing, reading, idleness, estimating. As a rule the averages for the idleness and the estimating intervals are nearly the same, but it is worthy of note that the females always overestimate to a greater extent when estimating than when idle. This is another indication of the discrepancy between the female time-unit and the objective unit.

About ninety per cent of the subjects estimated by some counting method. The methods most frequently used were "counting seconds," counting "1 and 2 and 3 and 4, etc.," counting the swings of a pendulum, tapping, and counting imaginary watch-ticks.

The above statements might suggest that the overestimation characteristic of the female judgments is due to a more rapid counting rhythm. This, however, is not true, for the letter-counting tests indicate a slightly more rapid rhythm for the males, 93.42 as opposed to 91.89.

TABLE 12

AMOUNT OF OVER- AND UNDER-ESTIMATION OF INTERVALS

Males. Females. Males. {I - 0.3 + 2.74 Age 17 18 19 20 21 22 23 {E + 1.5 + 7.55 Over- or under- -8.55 18̋ {R - 2.5 + 0.46 estimation -3.60 -7.50 -13.02 -13.49 -8.71 -13.20 {W - 6.5 - 2.42 No. + 6 2 2 0 2 2 2

{I - 2.7 + 6.81 No. - 10 14 14 16 14 14 14 {E - 2.9 + 5.54 36̋ {R - 3.9 + 5.71 {W -11.3 - 5.90 Females.

{I - 8.7 + 1.00 {E - 8.9 + 5.13 Age 17 18 19 20 72̋ {R -18.1 - 1.22 Over- or under- {W -24.8 -17.07 estimation + 2.38 - 4.16 - 2.57 + 2.55

{I -15.3 + 5.37 No. + 11 8 8 12 {E - 8.2 + 6.88 No. - 5 8 8 4 108̋{R -17.9 - 7.53 + = overestimation {W -32.5 -20.55 - = underestimation

In Table 13 are the means, variabilities, errors, etc., for the letter-counting tests. In this table we have presented the values of the various statistical quantities for the several ages of both males and females, for there are certain interesting differences which should be noted. Similar age-tables have been prepared for all of the other results, but in no case have noteworthy differences appeared. From Table 13 it will be observed that the males on the average count more letters in thirty seconds than do the females, and at the same time make more errors. The mean and relative variabilities for the sex-groups are almost the same. Curiously enough the number of letters counted as well as the accuracy of counting decrease for the females with age (17 to 20 years being the age-limits of the group under consideration). The males within the same age-limits increase in rapidity of counting, but decrease in accuracy. In the examination of Table 13 it is to be remembered that the 17 and 23 year groups of males contain only 16 individuals each, and therefore cannot be compared to advantage with the other groups. Both mean and relative variabilities decrease from 17 to 20 years for the females, whereas for the males there is no constancy in the direction of the change.

TABLE 13

MEANS, ETC., FOR LETTER-COUNTING

Males.

Age 17 18 19 20 21 22 23 Average M. 92.81 92.48 94.00 95.87 92.66 97.06 89.06 93.42 M. V. 10.29 15.31 16.50 12.90 7.81 13.04 19.20 13.58 +V. 11.76 22.96 16.50 11.70 8.13 16.30 21.94 15.61 No. +V. 7 9 20 37 24 14 7 16.96 -V. 9.15 11.48 16.50 14.40 7.51 10.87 17.09 12.57 No. -V. 9 18 20 30 26 21 9 19.00 R. V. 11.1 16.5 17.6 13.5 8.4 13.4 21.6 14.59 Errors 0.88 1.04 2.10 3.03 1.70 0.89 2.75 1.77

Females.

Age 17 18 19 20 Average M. 97.74 95.51 89.21 85.09 91.89 M. V. 17.06 14.56 11.30 11.28 13.55 +V 18.67 18.49 12.06 14.56 15.94 No. +V. 32 21 30 31 28.50 -V. 15.71 12.14 11.66 9.21 12.18 No. -V. 38 32 31 49 37.50 R. V. 17.5 15.2 12.7 13.3 14.70 Errors 1.35 1.46 1.49 1.65 1.47

SUMMARY

(1) The length of a second is slightly overestimated by men, greatly underestimated by women.

(2) Intervals of from 18 to 108 seconds are usually slightly underestimated by men (ages 17 to 23 years), and greatly overestimated by women (ages 17 to 20 years).

(3) The time-estimates of women are far more variable than those of men, and on the whole markedly less accurate.

(4) Both men and women favor estimates which end in 0 or 5, as well as simple fractions of a minute, but the tendency is stronger in women than in men. Over one third of the estimates reported in this paper were 15 seconds or simple multiples thereof.

(5) In letter-counting the groups of subjects studied (251 men and 274 women) exhibited differences just the opposite of those in time-estimation, for the men counted more rapidly and less accurately than did the women.

(6) Of the four fillings for the intervals used in the experiments, "writing" gave the smallest estimates of the intervals, listening to "reading" next, while "idleness" and "estimating" were conditions in which the intervals seemed much longer to both men and women.

(7) This preliminary study of sex-differences in time-estimation, by which it has been learned that women overestimate and are notably inaccurate in comparison with men, is to be followed and supplemented by the results of an investigation now in progress concerning the relations of sex-differences in time-estimation to age and physiological rhythms.

FOOTNOTES:

ASSOCIATIONS UNDER THE INFLUENCE OF DIFFERENT IDEAS

BY BIRD T. BALDWIN

The purpose of the following investigation was to study the influence of two or more starting-points on the train of associated ideas. It was begun in October, 1902, and concluded in January, 1905. Nineteen graduate students acted as subjects, and the experiments were conducted with each individually for one hour per week, except in a few instances where two subjects were present together. Occasionally the experimenter acted as subject in order to get a clearer insight into introspective data. There are recorded here thirteen groups of one hundred and eight sections, including eight hundred and fifty-five separate experiments, with a sum total of eleven thousand, one hundred and thirty-five named associations.

So far the field has not been studied experimentally, although Cordes, while attempting to observe the effect of an unconscious intermediating factor in 'mediate association,' noticed that the accompanying factor (particularly with a tone) sometimes combined with the starting-point in determining the associated series. The fact is simply mentioned, and Scripture's conclusions from the same are inadequate when he states, "In general we may say that two simultaneous ideas have an effect that depends on their relative masses; if one of the ideas is over-poweringly weighty the next idea will be chiefly influenced by it, but if the two are nearly balanced the next idea will be the result of the two." Miss Calkins makes a near approach to the problem in a study of "Mental Combination," where auditory words were given as nearly simultaneously as possible and the subject was required to remain for four seconds in silence and then to write the train of imagery which passed in six following seconds; the words as auditory starting-points were finally excluded from the experiments "because the first word pronounced tends often to establish itself so firmly that its association-images are proof against the intrusion of the second word, which has therefore no chance to be grasped with the first." These results differ from the ones here recorded which were obtained independently and under different conditions. In her valuable monograph on Association there is in one instance a suggestion of the problem, but no results are given.

The two or more starting-points were nonsense syllables, concrete or abstract words, pronounced by the investigator or shown on cards, or finally, pictures. The associations continued for a period of fifty seconds unless otherwise indicated. The same letters denote the same subjects throughout all the experiments, K. and R. being women. In the examples given, Roman numbers have been used, in place of the letters, as it was found impossible to get representative series which were entirely devoid of personal references.

After the associations had been "jotted down," the subjects in each case copied them, noting also the suggesting ideas, and giving, in many cases, a number of introspective notes. It is important to mention that in all of the work these records of suggesting ideas were made by the subjects without any questions on the part of the investigator.

GROUP I. TWO NONSENSE SYLLABLES

In order that a standard might be obtained that would serve as a basis for subsequent experiments, two nonsense syllables were pronounced with equal emphasis, as starting-points, and the subject was asked merely to wait until both had entered consciousness and not to favor nor inhibit either. If the subject stated that associations arose between the pronunciation of the starting-points the results were discarded.

Turning to the notes we find that K. reported, "The first just hovered around," and the associations followed the second, or as V. states it, "Sof was carried subconsciously for a while without influence," while By. frequently noticed that "The first may be in the background on the point of breaking in but be inhibited involuntarily," or in another, "It was frequently present but exerting no influence." A. "forgot what the other word was, but felt it trying to get a hand in." Ro. goes further and states in terms similar to A., "The words which might have been suggested by 'taf' were wholly inhibited, though I am sure 'taf' was present to me throughout the series, but was not efficacious as against the other syllable." It is unnecessary to multiply instances to show that the wraith, so to speak, of one may linger for a short time, or occasionally for the entire series, but that the associations are determined by the other. Or the lingering wraith of a starting-point may hover ineffectively in consciousness and exert no determining influence on the series until toward its close; then, however, although both starting-points still persist, they exchange places so far as effectiveness is concerned, and the former wraith becomes the primary factor in effecting the series of associations. For example, "The two syllables were present all the way through, but the former exhausted itself as a word-suggester after a few words." Again Ro. gives examples where "One alone may be in consciousness and determine the associations when the other may enter without effect upon the series, or may determine the series, the first persisting but losing its effect"; and a case of both ideas persisting and each alternately exerting an influence was noted by M. These last two conditions are exceptional.

Studying carefully the notes and graphic representations as given in the tables, it is apparent that one starting-point may be followed independently and exclusively; we find "'fef' suggesting 'theft,' and the subsequent associations monopolizing the field of consciousness and 'tuz' not again appearing." (V.) The results show that in the one hundred and seven experiments one was followed exclusively only twenty times. The following would be an example:

III. Naf-Tam

--tambour tam --experiment tambour --psychology experiment --laboratory experiment --space psychology --time space --ego space and time --Ebbinghaus ego --discussion psychology --posted notice psychology --door posted notice --gray door --stairs door --fishes laboratory --water fishes --decay fishes --animal fishes --meat fishes --odor meat

In other cases one starting-point may be followed and then the other for a few or many associations, each "occupying the mind to the exclusion of the other for a number of words, when suddenly the other appears and suggests a new series," which V. calls "a curious zigzagging in consciousness." This may take place at any point along the series. As Ro. indicates, "There was rivalry here; 'yud' wholly inhibited all the 'zid' associations, even 'zid' itself for the time, then, 'yud' apparently being exhausted, 'zid' entered again into the focus."

In contrast to the above forms, the tables indicate and the notes verify that while the associations primarily follow one there can be traced the modifying effect of the other; "The nature of the words was influenced by a fringe of consciousness which contains the other starting-point," M.'s comment runs, while J. adds, "The other helped in the mental picture but had no special significance." There are no experiments in the first group where this partial fusion took place throughout the series, and there are only six examples where it took place before the sixth word and continued to the end. While there are no cases here of alternating partial fusion, yet it sometimes happens, that is, the predominating influence is sometimes transferred from one to the other. The modifying influence may reach such a degree at any point in the series that the identity of each point of departure of the association is lost and there is total fusion, resulting in one idea or one train of thought. For example, 'rel' and 'mem' gave 'realm' for V.; for Ht. 'fef' and 'tuz' gave 'fezz'; for M. they gave 'fez'; for Ro. 'fuz.'

Of the twenty-one experiments where one starting-point was followed exclusively, there were two cases for the first and nineteen for the second. There were but two cases of alternating groups, a third one being unique in that after the first three words there was a persistent alternation between two apparently independent series of inter-related words and inter-related visual images. There were three cases of complete fusion throughout the series and nine where total fusion started before the fifth word and continued.

In a number of cases the initial starting-points were forgotten before the end of the series and the subjects were sometimes unable to recall them.

As the subjects were kept ignorant of the characteristic persistence of either starting-point in forming separate associations, and of the tendency towards fusion, as well as of the preponderance of influence of one over the other, it is to the tabulated results that we must look for quantitative measurement of these. 76.2% of the associations are due to the independent influence of one or the other of the starting-points; 23.9% are due to the combined influence; for the first starting-point we have 18.7%, for the second 57.5%. In 15.3% of all the associations the combined influence was the result of total fusion; in 4.1%, the result of a union where the first starting-point predominated in influence; in 4.4%, the second predominated. The following graphic representation indicates the order of influences for the first pair of syllables. The table gives the results for the ten pairs.

TABLE I. TWO NONSENSE SYLLABLES SPOKEN

Time--50 seconds.

Characters: | represents the influence of the first, - the second, + complete fusion, α fusion with the first predominating, β fusion with the second predominating.

1) Taf--Coz

M. | - - - - | | | | | - - - - F. - - | | + + + + | | | + + β β β β β β H. - | | | | | | | | | | | | | | | R. | - - - - - - - - - - - - - - - K. - - - - - - - V. | | - - - - | - - | + + + + Ro. | | | - - - - - - - - - - - - - - - - Bl. - - - - - - - - - - - - - - - By. | - | | | - - - - - - - - - - - - - - - - - - - Bs. - - - - - - - - - - - - - - - Ht. - - | | | | | | | | | | | | |

Starting-points:

(1) Taf--Coz. (2) Cim--Bef. (3) Yud--Zid. (4) Sof--Deb. (5) Naf--Tam. (6) Fef--Tuz. (7) Sar--Nef. (8) Sek--Lub. (9) Hov--Bes. (10) Rel--Mem.

| - + α β

M. { 20 69 26 4 3 { 16.5% 56.6% 21.2% 3.3% 2.4%

F. { 28 22 37 0 39 { 22.2% 17.5% 29.4% 30.9%

H. { 39 98 0 11 0 { 26.3% 66.2% 7.5%

R. { 9 102 32 0 0 { 6.3% 71.4% 22.3%

K. { 47 58 1 14 0 { 39.1% 48.3% .7% 11.9%

V. { 30 79 54 0 0 { 18.4% 48.4% 33.2%

Ro. { 32 111 0 5 0 { 21.4% 75.1% 3.5%

Bl. { 5 131 15 15 0 { 3.1% 78.9% 9.0% 9.0%

By. { 43 102 23 15 22 { 16.8% 52.6% 11.7% 7.7% 11.2%

Bs. { 9 72 31 0 6 { 11.4% 58.5% 25.3% 4.8%

Ht. { 32 58 20 0 0 { 29.0% 52.8% 18.2%

Totals{294 902 239 64 70 { 18.7% 57.5% 15.3% 4.1% 4.4%

Number of subjects, 11; number of sections, 10; number of experiments, 110; number of associations, 1569.

GROUP II

Two concrete nouns with apparently equivalent connotation were pronounced. In order that the subject get no clue as to the preponderance of one starting-point over the other, the nine sections were given at irregular intervals in connection with other experiments.

The tables show few cases of fusion, there being but one case in the eighty experiments where all the associations were the result of the combined influence of both starting-points. There was no other instance where complete fusion took place before the sixth association and continued throughout, and only six cases where any form of fusion took place in the eighty experiments within the first five words (7.6%), while in the one hundred and seven experiments of Group I there were twenty-five (23.9%).

The tables show a tendency which was noted when two syllables were used and which is emphatically brought out here, namely, it is the position or sequence which determines which of the two equivalent starting-points shall produce the greater influence. When the two starting-points are given in immediate succession, it is the second which predominates in influence. This preponderance may be clearly demonstrated by the tables alone, though there are many notes which show "the greater influence of one" (the second). A. seems to have realized this when he wrote, "While the first was uppermost the second hovered in subconsciousness, resting content, knowing that it would have its turn soon." In corroboration the tables give--abstracting from all cases of fusion--268 words (23.2%) for the first, 721 words (62.3%) for the second. In fifty-five of the eighty experiments the last had full control at the end of the series. Taking the cases where one was followed exclusively throughout the series, starting between the first and the sixth named association, we find eight for the first and twenty-nine for the second. This greater influence is again to be found in the cases of partial fusion; the first predominating in 2.6%, and the second in 4.3%.

The results of both groups show that the starting-points tend in a high degree toward independent influence, and also that such a method of presentation is one of sequence rather than simultaneity, as the two words show unequal influence. The former must in a manner be reproduced to become a point of departure for associations, while the latter acts directly. Is it then a condition of mind that when similar impressions are presented the one in the presence of which consciousness is reacting directly has a greater influence in arousing associations than one which is just past? This we are forced to conclude is the case, but the proof of the conclusion will be supplemented by later results.

TABLE II. TWO WORDS SPOKEN

Time--50 seconds. Characters--same as Table I.

(1) Library--River

M. | + + β - - - - - - - - - - - - F. | - - - - - - - - - | | | | H. - - - - - - - - - - - - - - - - Ro. | | | | - - - - - - - - - J. | - | | | | - - - - - - - - - - - - - - - Bl. | | - - - - | | | | | | | | | - - / By. - - - - - - - - - - - - - - - - - - - - - - Bos. | - - - - - - - Bur. - + - - - - - - - - - - -

(1) Library--River. (2) Hat--Road. (3) Newspaper--Medicine. (4) Bicycle--Drum. (5) Theatre--Magazine. (6) Store--Church. (7) Soldier--House. (8) Ship--Boy. (9) Furniture--Tree.

TABLE II.--continued

| - + α β

M. {32 66 26 0 13 {23.4% 48.2% 18.9% 9.5%

F. {32 67 13 5 1 {27.2% 56.7% 11.0% 4.3% .8%

H. {59 79 5 0 0 {41.3% 55.2% 3.5%

Ro. {44 85 0 0 0 {34.1% 65.9%

J. {21 121 8 5 0 {13.5% 78. % 5.3% 3.2%

Bl. {22 83 6 0 1 {19.7% 74.1% 5.4% .8%

By. {16 97 18 14 23 { 9.5% 57.8% 10.7% 8.3% 13.7%

Bs. {26 64 3 7 1 {25.8% 63.5% 2.9% 6.9% .9%

Bur. {16 59 8 0 10 {17.2% 63.4% 8.6% 0 10.8%

Total {268 721 87 31 49 { 23.2% 62.3% 7.6% 2.6% 4.3%

Number of subjects, 9; number of sections, 19; number of experiments, 80; number of associations, 1156.

GROUP III

For two syllables and like words the second strongly predominates in awakening associations; will the same be true when two simple outline pictures are shown in the same order? The following results show in attestation of the above conclusion percentages remarkably similar to those of the words. There are 23.7% for the first and 61% for the second; 84.7% of the associations show no sign of fusion and only 15.3% for fusion and the different forms of partial fusion, the second still holding the ascendency. When the two pictures did fuse, R. tells us, "All the associations were more elaborate pictures than when mere words were given."

A comparative study, inadequate though it is, offers a partial parallelism between the predominating memory type and stimulation. The subjects who are preëminently of the visual, K., visual motor, M., H., V., and visual lingual motor, Ht., find the pictures more suggestive than the syllables, while those of the pure motor, F. and By., reverse the order and find that the syllables offer more numerous and "more vivid" (F.) associations.

Common experience, however, immediately shows the limitations of reaching conclusions, when considering prolificacy as criteria of suggestiveness, inasmuch as starting-points which are abundantly rich in associations tend to produce so many points of departure that they tend to inhibit one another. Again there is to be considered the kind of associations, those of the syllables being of a very elementary character and in serial form.

A few experiments were given in which colored slips of paper were used as starting-points. These proved very suggestive for the subjects. Also a few tones were given, but these were soon discontinued, as other tones, which could not be recorded, were frequently suggested. The sentence which forms a very satisfactory starting-point where one is used could not be used to an advantage where several were given. As we are here interested in the mutual influence of the starting-points, our remaining study will be confined to a quantitative and qualitative variation of the forms used in the previous groups.

TABLE III. TWO PICTURES SHOWN

Time--50 seconds. Characters--same as Table I.

(1) Boy Rolling Hoop--Blacksmith

M. | | | - - - - - - | | | | | F. | | - - | + + H. - - - - - - - - - - - - R. - - - - - - - - - - - - - - - - K. - - - - - - - - - - - - - - - - - V. - - | | + + + + + + + + + + + + + + S. - - - - - - - - - | | | | | | | By. | - - - - - - - - - - - - - - - - - - - - - Bs. - - - - - - - - - Ht. | | - - - - - - - -

(1) Boy Rolling Hoop--Blacksmith. (2) Old Man With Umbrella--Bird House. (3) Carpenter--Mower. (4) Children Playing--Boy with Basket. (5) Horse--Dog House. (6) Shoemaker--Fisherman. (7) Little Girl--A Chicken. (8) Boy--A Sheep.

| - + α β

M. {41 65 2 0 1 {37.6% 59.7% 1.8% .9%

F. {11 42 21 0 5 {13.9% 53.2% 26.6% 6.3%

H. {37 52 11 26 9 {27.4% 38.5% 8.1% 19.3% 6.7%

R. {41 58 8 0 15 {33.4% 48% 6.5% 12.1%

K. {32 82 0 0 0 {27.1% 72.9%

V. {12 95 21 3 5 { 8.8% 70% 15.4% 2.1% 3.7%

S. {36 44 0 0 0 {45% 55%

By. { 8 111 12 0 13 { 5.4% 77.2% 8.3% 9.1%

Bs. {13 40 7 0 0 {21.7% 66.7% 11.6%

Ht. {22 72 7 0 4 {20.9% 68.6% 6.7% 3.8%

Totals {253 661 89 29 52 { 23.7% 61% 8.2% 2.7% 4.4%

Number of subjects, 10; number of sections, 8; number of experiments, 76; number of associations, 1084.

GROUP IV

The previous experiments suggest the conclusion that that starting-point in the presence of which consciousness is reacting exerts a greater influence than one just past. The proof of this anticipated result is supplemented by subsequent consectary data.

It is necessary to test the outcome when the nature of the impressions is varied and the starting-points are given simultaneously. An outline picture and a word may be so given. These experiments verify from a different standpoint the statement that the picture establishes itself more permanently and is more influential, there being 47.8% of the associations produced by the picture alone, 14.8% by the words alone, 25% of a fusion of the two, and 12.4% of a uniting influence with the first predominating. The subject R. is an exception in that she favors the word when not favoring a combination of the two, a fact which I am unable to explain except to add that the subject stated that the æsthetic pleasure connected with the picture was sufficient to inhibit the associations. K. has 100% for the pictures, which is partially explained by the fact that she is a remarkable visualizer who reproduces all situations in visual terms.

TABLE IV. PICTURE SHOWN AND WORD SPOKEN

Time--50 seconds. Characters same as Table I.

(1) (a rabbit)--Table

M. | - - | | | | | | | + - - - - R. + + + + + + + + + + + + + + + + + + K. | | | | | | | | | | | | | | | S. | | | | | - | | | | | J. | - - | | | | | | | | | - - - | | | | | | | | | Ht. | - | | | | | | | | | | | | - - - -

(1) (a rabbit)--Table. (2) (horse)--Book. (3) (boy)--Dish. (4) (duck)--Iron. (5) (dog)--Paper.

| - + α β M. { 38 9 16 6 0 { 55.0% 13.1% 23.3% 8.6%

R. { 1 20 55 0 0 { 1.3% 26.3% 72.4%

K. { 65 0 0 0 0 {100.0%

S. { 29 15 3 10 0 { 50.8% 26.3% 5.3% 17.6%

J. { 75 19 14 9 0 { 64.2% 16.3% 11.9% 7.6%

Ht. { 17 7 30 33 0 { 19.5% 8.1% 34.5% 37.9%

Totals {225 70 118 58 0 { 47.8% 14.8% 25.0% 12.4%

Number of subjects, 6; number of sections, 5; number of experiments, 30; number of associations, 471.

GROUP V

The character of the starting-points determines which shall be followed in a simultaneous presentation; can the character of two immediately successive starting-points be so varied as to overcome the factor of temporal difference? The few experiments in this group show an attempt to shift and measure this influence, selecting as variants words of general and specific character, the first being a term of everyday experience with the subjects, of wide connotation, and therefore presumably rich in associations, and the second, which had the advantage of position, one of more limited and specific connotation, thus in a measure making it possible to test the stability of the second in establishing itself as a word-suggester. For the words "Gymnasium--Stamens" the second still strongly predominates. There is a slight increase in the amount of fusion (24.5% in all), the explanation of which would probably lie in the fact that the first is more assertive than before and offers more elements common to the two. There was one case of total fusion, one case where fusion started before the sixth word and continued, no cases of alternating fusion, four cases of partial fusion, five cases where the second was followed throughout, no case where the first was followed throughout, five cases of alternation between the first and the second. In 75.5%, the influence was exerted without fusion. Eight of the nine subjects favor the specific term.

TABLE V. GENERAL AND FAMILIAR WORD--SPECIFIC WORD SPOKEN

Time--50 seconds. Characters--same as Table I.

(1) Laboratory--Rake

M. | | | | | - - - - - - - - - - - - - - - - F. | - - - - - - - - - H. + + + + + + + + + + + V. | - - - - - - - - - - - - - | | | | | S. - - - - - - - - - - | | | | Bl. - - - - - - - - - - - - - - - - - By. | - - | | | | | | | | | Bs. Ht. - - - - - - - - J. | | β β β β β β β β β β β β | | | | |

(1) Laboratory--Rake. (2) Experiment--Butterfly. (3) Gymnasium--Stamens.

TABLE V.--continued

{ 22 25 4 6 0 M. { 38.6% 43.9% 7.0% 10.5%

{ 4 19 0 0 0 F. { 17.4% 82.6%

{ 4 25 13 0 0 H. { 9.5% 59.6% 30.9%

{ 7 38 2 0 7 V. { 12.8% 70.7% 3.7% 12.8%

{ 11 34 0 0 0 S. { 24.4% 75.6%

{ 0 36 0 0 0 Bl. { 100.0%

{ 11 16 7 0 9 By. { 25.5% 37.1% 16.5% 20.9%

{ 0 9 12 2 Bs. { 39.1% 52.2% 8.7%

{ 10 11 10 Ht. { 32.3% 35.4% 32.3%

{ 17 8 15 0 13 J. { 32.1% 15.1% 28.3% 24.5%

{ 86 221 63 8 29 Totals { 21.2% 54.3% 15.4% 1.9% 7.2%

Number of subjects, 10; number of sections, 3; number of experiments, 27; number of associations, 407.

GROUP VI

What modification of influence takes place when an auditory impression in the form of a word of abstract nature (the pure abstract words are taken later) is given in comparison with a concrete noun? The tables indicate that we are now able partly to overcome the disadvantage of first position by the advantage of concrete content. 24% of the associations are the result of total fusion; there are eight cases of total fusion throughout the series, and seven where total fusion took place before the third word and continued. There are two cases where the first was followed throughout, and but one where the second was followed exclusively. The tables indicate that there were very few words. Some of the subjects claim, "These words did not seem equally rich in associations. I was not at all conscious of the one while the other was in consciousness." (H.) The most characteristic feature here was, as Br. also indicates, "Great amount of rivalry at the beginning of the series"; while Bs. states, "There was a lot of confusion and a feeling of groping for words." Br. adds later, "For some seconds association seemed obstructed. Then by an effort the process was started which followed an involuntary course. A kind of confused presence of both words." Another subject adds, "There was a long blank after the words were said in which both words were balancing off in the fringe of consciousness and the mind expectant, passively waiting for an association to turn up. The hesitant period seemed marked by an attempt at a synthesis of these two words in some way."

SUBJECT XIX

Lamp--Justice.

| chimney lamp. | white lamp chimney. | yellow white chimney. | flame yellow. | nickel lamp plus the other words. - scales justice. - purple robe justice.

As the tables indicate, there was frequently a strong tendency here for the abstract terms to fuse. As H. noticed, "It does not tend to call up associations of its own stripe, but in some way becomes concrete."

TABLE VI. TWO WORDS: CONCRETE--ABSTRACT SPOKEN

Time--15 seconds. Characters--same as Table I.

(I) Desert--Hate

A. | | | + + + M. | | | | - - F. + + + H. | - | | | | | | | Ro. | | - - | | | | Bl. + + + + + + + By. | | | + + + + + + Bs. - | | | | | Br. | | | + +

(1) Desert--Hate. (2) Lamp--Justice. (3) Pen--Fatigue, (4) Gate--Fear. TABLE VI--continued

| - + α β A. { 10 2 6 0 0 { 55.5% 11.2% 33.3%

M. { 7 8 10 0 0 { 28.0% 32.0% 40.0%

F. { 3 8 5 0 0 { 18.7% 50.1% 31.2%

H. { 17 16 1 0 0 { 50.0% 47.1% 2.9%

Ro. { 15 11 0 0 0 { 57.7% 42.3%

Bl. { 4 15 7 2 0 { 14.2% 53.5% 25.1% 7.2%

By. { 4 21 8 0 1 { 11.8% 61.8% 23.5% 2.9%

Bs. { 5 1 15 0 2 { 21.8% 4.3% 65.2% 8.7%

Br. { 11 8 2 0 0 { 52.3% 38.1% 9.6%

Totals { 76 90 54 2 3 { 33.8% 40.0% 24.0% .8% 1.4%

Number of subjects, 9; number of sections, 4; number of experiments, 34; number of associations, 225.

Reversing the order by placing the concrete noun second, it gains in influence. We are told by the subjects at this point, "The choice seems to be determined by the concreteness of the word." (H.) "The abstract soon exhausted itself as a word-suggester." (Ro.) There was fusion in 24% of the associations of the first group, and 22.2% in the second. There are six cases where the second alone prevails, two for the first starting-point and four for the second. There are eleven cases of fusion beginning before the sixth word and continuing throughout the series. There is much partial fusion, with the second predominating in influence.

The results again emphasize the fact that the influence is transferable, also that normally the second has the advantage; furthermore, they illustrate the preponderance of the concrete word as a starter of associations, and that the abstract term when it exerts an influence tends to fuse rather than persist in having separate associations; all of which shows that concrete terms produce more vivid impressions than abstract ones, and would, when it is possible to use them, be of direct aid to the learner.

TABLE VII. TWO WORDS: ABSTRACT--CONCRETE SPOKEN

Time--15 seconds. Characters--same as Table I.

(1) Honesty--Tide

A. - | - - - - M. - - - - | | | | F. | - - - - - H. | - - - Ro. | | - - - - Bl. | - - - - - - By. - - + + Bs. - - - - - - - J. - + + + + + +

(1) Honesty--Tide. (2) Skill--Coal. (3) Terror--Sky. (4) Refined--Flag.

| - + α β A. { 2 7 11 0 4 { 8.4% 29.2% 45.8% 16.6%

M. { 4 11 10 1 5 {12.9% 35.5% 32.3% 3.2% 16.1%

F. { 2 8 9 0 0 {10.5% 42.2% 47.3%

H. { 8 11 1 3 0 {39.2% 43.5% 4.1% 13.2%

Ro. {10 13 3 0 0 {38.4% 50.0% 11.6%

Bl. {15 10 0 0 0 {60.0% 40.0%

By. { 2 20 3 0 0 { 8.0% 80.0% 12.0%

Bs. { 1 11 6 0 2 { 5.0% 55.0% 30.0% 10.0%

J. { 6 16 6 0 0 {21.4% 57.2% 21.4%

Totals {50 107 49 4 11 {22.6% 48.4% 22.2% 1.9% 4.9%

Number of subjects, 9; number of sections, 4; number of experiments, 36; number of associations, 221.

Increasing the disparateness by making the one a proper name and the other a pure abstract noun, we find the name dominates consciousness, almost to the exclusion of the abstract term. The tables confirm the conclusions that the abstract term, even when given the advantage of position, exerts little influence, for in the first group of eighteen experiments of two hundred and eighty-four associations there are 13.4 times as many associations for the first as for the second, or two hundred and fifteen words for the first (75.8%) and sixteen (5.5%) for the second.

Reversing the order, the burden of influence swings back to thirty-eight (12.8%) for the first and one hundred and fifty-six (52.7%) for the second with an amount of fusion increased to ninety-nine (33.4%).

TABLE VIII. TWO WORDS: PROPER NOUN--ABSTRACT NOUN SPOKEN

Time--50 seconds. Characters--same as Table I.

(1) Lowell--Liberty

M. | | | | α α α | | | | | | R. | α α | | | | | | | | | | | | | K. | | | | | | | | | | | | | | J. | | | + + + α α α α α α α α α α α α α S. | | | | | - | | | | | | | Ht. - - | | | | | | | | | | | | | |

(1) Lowell--Liberty. (2) Roosevelt--Fidelity. (3) Eliot--Integrity.

| - + α β

M. { 29 3 2 7 0 { 70.7% 7.4% 4.8% 17.1%

R. { 30 0 17 2 0 { 61.2% 34.7% 4.1%

K. { 43 2 0 0 0 { 95.5% 4.5%

J. { 42 0 8 13 0 { 66.6% 12.7% 20.7%

S. { 40 2 0 0 0 { 95.2% 4.8%

Ht. { 31 9 3 1 0 { 70.4% 20.4% 6.9% 2.3%

Totals {215 16 30 23 0 { 75.8% 5.5% 10.7% 8.0%

Number of subjects, 6; number of sections, 3; number of experiments, 18; number of associations, 284.

TABLE IX. TWO WORDS: ABSTRACT NOUN--PROPER NOUN SPOKEN

Time--50 seconds. Characters--same as Table I.

(1) Individuality--Lincoln

M. | + | + + + + + + + + + + + + R. - - - - - - - - - - - - - - - - K. - - - - - - - - - - - - J. | - - - - - - - - - - - | | | | | | | | | | S. - - - - - - - - - - - - - - - Ht. - - - - - - - - - - - - - - - - - -

(1) Individuality--Lincoln. (2) Brevity--Webster. (3) Justice--Hanus.

| - + α β

M. { 4 19 25 0 0 { 8.3% 39.6% 52.1%

R. { 5 31 12 0 0 {10.4% 64.5% 25.1%

K. {14 31 0 0 0 {31.2% 68.8%

J. {13 11 44 0 0 {19.1% 16.1% 64.8%

S. { 2 40 0 0 0 { 4.8% 95.2%

Ht. { 0 24 18 0 3 { 53.4% 40.0% 6.6%

Totals {38 156 99 0 3 {12.8% 52.7% 33.4% 1.1%

Number of subjects, 6; number of sections, 3; number of experiments, 18; number of associations, 296.

GROUP VII

Occasional experiments of the previous groups indicated an abnormal influence of those ideas which were accentuated by the special interests of the subjects. We ask therefore: What is the relation of the newly aroused associations to the present content of consciousness? May the present content of consciousness be so varied as to reënforce or inhibit the characteristic influence of the individual impressions successively presented? In order to test this, words or a sentence were pronounced just previous to the presentation of the two starting-points related to them as follows; in the first divisions the preparatory words lead to the second, in the second divisions to the first, in the third divisions to neither, in the fourth divisions to both, and in the last we have a slight change in the marginal setting and a sentence leads to the second.

The following are examples of the first and second divisions:

Quartz--Granite--Shale. HORN--SLATE. Dog--Sheep--Horse. SQUIRREL--TELEGRAM.

SUBJECT XVIII

--slate slate --slate ledge (Col.) slate --Great Falls, Mont. quartz, granite, shale --geology quartz, granite, shale, G. F. --Will B. geology --State University Will B. --Sexton James State University --Mrs. J. Sexton James --C. S. J. Mrs. J. --Indian School C. S. J. --Fort Shaw School Indian School --Mrs. C. Ft. Shaw School --Mrs. E. Mrs. C. --Seattle, Wash Mrs. F. --Miss M. Seattle, Wash. --Everyman Miss M. --Cousin V. Everyman --theatre Cousin V.

SUBJECT XII

| hunt squirrel (and dog) | pasture sheep, horse, and hunt | yard squirrel and pasture | Harvard yard | Freshman Harvard | themes Freshman and squirrel | hunting themes, dog, " | George hunting | squirrel George | creatures squirrel | animals creatures | activity " animals | agility " " | grace " " | cuteness " " | pets " " | civilisation creatures | vs. cats " | enemies cats

That one starting-point establishes itself more firmly, and offers more dominant associations with an increased degree of suggestiveness is intimated by various expressions of the subjects who state, "the preparatory words called forth associations connected with this or that starting-point"; or "there was a felt fusion of all"; "a summation of all"; or "the preparatory words had an influence throughout"; "they strengthened this or that starting-point"; "they affected one and not the other"; and many similar expressions.

The influence of these preparatory marginal settings was also indicated very often by the nature or kind of words in the series.

Turning to the tables which collectively represent in a graphic and quantitative form the notes of the subject, and which are in harmony with the above, we find they demonstrate that there is a very intimate and definite relationship. When the state of mind immediately preceding the moment of the formation of the associated series is conditioned by the preparatory words leading to the second, the amount to which that starting-point dominates consciousness in arousing associations is greater than in any previous case where the words of like nature are pronounced.

With the preparatory words leading to the second starting-point, we have 11.5% of the influence for the first, 69.2% for the second; with them leading to the first, 58.2% for first, 17.2% for second; with them leading to neither, 40.6% for first, 44.2% for second; with them leading to both, 16.1% for first, 54.8% for second; with a preparatory sentence leading to second, 12.9% for first, 59.7% for second.

A few experiments were made where one word was given as a setting and four or eight starting-points were shown, but the starting-points were so numerous that they tended to confuse the subject's introspective account.

TABLE X. TWO WORDS SHOWN: THREE PREPARATORY WORDS SPOKEN LEADING TO THE SECOND

Time--50 seconds. Characters--same as Table I.

Corn, Wheat, Oats. (1) MOB--HAY

M. + + + - + + - - - - R. - - - | + + α + β - - - - - K. - - - - - - - - - - - - - - S. + + + + - - - - - - - - - - - - J. + + + + + + β β - - - - - - - Ht. - + + + + β β β β β β β β

(1) Corn, Wheat, Oats. MOB--HAY. (2) Cloud, Mist, Dew. ORATION--CANOE. (3) Turtle, Fish, Frog. KEY--NET. (4) Quartz, Granite, Shale. HORN--SLATE. (5) Geometry, Plane, Rectangle. CASE--CUBE.

| - + α β

{ 10 41 17 0 7 M. { 13.3% 54.7% 22.7% 9.3%

{ 15 37 21 1 4 R. { 19.3% 47.4% 26.9% 1.2% 5.2%

{ 0 55 0 0 0 K. { 100.0%

{ 8 64 4 0 0 S. { 10.5% 84.2% 5.3%

{ 17 69 6 0 2 J. { 18.1% 73.4% 6.3% 2.2%

{ 2 45 17 0 8 Ht. { 2.8% 62.5% 23.6% 11.1%

{ 52 311 65 1 21 Totals { 11.5% 69.2% 14.4% .2% 4.7%

Number of subjects, 6; number of sections, 5; number of experiments, 29; number of associations, 449.

TABLE XI. TWO WORDS SHOWN. THREE PREPARATORY WORDS SPOKEN LEADING TO THE FIRST

Time--50 seconds. Characters--same as Table I.

Dog, Sheep, Horse. (1) SQUIRREL--TELEGRAM

M. | | | | | | | | | | | | | | R. | | | | | | | | | | | | | | | K. | | | | | | | | | | | | | S. | | - - - - - - - - - - J. | | | | | | | | | | | | | | | | | | | Ht. | | | | | | | | | | - - - - - - - -

(1) Dog, Sheep, Horse. SQUIRREL--TELEGRAM. (2) Pineapple, Banana, Orange. FRUIT-STAND--ELECTRIC LIGHT. (3) Justice, Truth, Beauty. CHARITY--PHOTOGRAPH. (4) Gold, Copper, Silver. DIME--SKULL.

M. {28 3 15 0 8 {51.8% 5.5% 27.9% 14.8%

R. {16 9 23 0 11 {27.2% 15.3% 38.9% 18.6%

K. {56 2 0 0 0 {96.5% 3.5%

S. {36 24 0 0 0 {60.0% 40.0%

J. {38 4 24 1 7 {51.4% 5.4% 32.4% 1.4% 9.4%

Ht. {37 20 0 0 0 {64.9% 35.1%

Totals {211 62 62 1 26 {58.2% 17.2% 17.2% .2% 7.2%

Number of subjects, 6; number of sections, 4; number of experiments, 24; number of associations, 362.

TABLE XII. TWO WORDS SHOWN. THREE PREPARATORY WORDS SPOKEN LEADING TO NEITHER

Time--50 seconds. Characters--same as Table I.

Botany, Statue, Postal. (1) CHOCOLATE--BANNER

M. | | - - - - - - - - - - - - - | R. - | | | | | | | | | | | | | | | | | | | | K. | - | | | - | - | - | - | S. | | - - - - - - - - - - - - - - - - J. | | | | | | | - - - - - - - - - - - - - - - Ht. - - + + + - - - - - - - - - - -

(1) Botany, Statue, Postal. CHOCOLATE--BANNER. (2) Idea, Proposition, Syllogism. SILVER--GAME.

| - + α β

M. {3 28 0 0 0 {9.6% 90.4%

M. {22 1 2 0 14 {56.5% 2.5% 5.2% 35.8%

K. {28 5 0 0 0 {84.8% 15.2%

S. {13 20 0 0 0 {89.4% 60.6%

J. { 17 22 0 0 0 { 43.6% 56.4%

Ht. { 0 14 8 7 0 { 48.3% 27.6% 24.1%

Totals { 83 90 10 7 14 { 40.6% 44.2% 4.9% 3.4% 6.9%

Number of subjects, 6; number of sections, 2; number of experiments, 12; number of associations, 204.

TABLE XIII. TWO WORDS SHOWN. THREE PREPARATORY WORDS SPOKEN LEADING TO BOTH

Time--50 seconds. Characters--same as Table I.

Tree, Shrub, Grass. (1) TEA--FERN

M. + + α α α α α α α - - R. + + + + + + + + + + + + K. - - - - - - - - - - - - - - - - S. - - - - - - - - - - - - J. | - - - - - - - - - - - - - - - - - - - - Ht. + + α α α α α α | | | | |

(1) Tree, Shrub, Grass. TEA--FERN. (2) Fox, Wolf, Moose. DEER--DOVE.

| - + α β

M. { 2 12 4 8 0 { 7.6% 46.3% 15.3% 30.8%

R. { 0 0 25 0 0 { 100.0%

K. { 1 30 0 0 0 { 3.2% 96.8%

S. { 8 12 0 4 0 { 33.3% 50.0% 16.7%

J. { 8 26 0 0 0 { 23.5% 76.5%

Ht. { 8 12 2 6 0 { 28.5% 42.8% 7.2% 21.5%

Totals { 27 92 31 18 0 { 16.1% 54.8% 18.4% 10.7%

Number of subjects, 6; number of sections, 2; number of experiments, 12; number of associations, 168.

TABLE XIV. TWO WORDS SHOWN (AUTHORS)--WITH A PREPARATORY SENTENCE LEADING TO THE SECOND

Time--50 seconds. Characters--same as Table I.

"God's in His Heaven--all's right with the world."

(1) BURNS--BROWNING

M. + - - - - - - - - - | | | | | | R. | - - - - - - - - - - - K. + + + + + + + + + + + + + S. | - - - - - - - - - - - - - | | | J. | | - - - - - - - - - - - - - - - - - - Ht. - - - - - - - - - -

(1) "God's in His Heaven--all's right with the world." BURNS--BROWNING.

(2) "All that glistens is not gold." BYRON--SHAKSPERE.

(3) "Hitch your wagon to a star." SPENCER--EMERSON.

| - + α β

M. { 9 15 15 0 4 {20.9% 34.8% 34.8% 9.5%

R. { 1 47 0 0 0 { 2.1% 97.9%

K. {15 0 26 0 0 {36.5% 63.5%

S. { 6 34 0 0 0 {15.0% 85.0%

J. { 3 43 14 0 0 { 5.1% 71.6% 23.3%

Ht. { 0 18 13 0 0 { 58.1% 41.9%

Totals {34 157 68 0 4 {12.9% 59.7% 25.8% 1.6%

Number of subjects, 6; number of sections, 3; number of experiments, 18; number of associations, 263.

GROUP VIII

The aim here is to test the effect of interruption in the series of associations, and to throw further light on the relation of the series to the present content of consciousness when this content is a series of associations and the new content is a pronounced word which is to act as a point of departure for new associations.

There are three divisions. The time in all is fifty seconds; in the first a word of general connotation is given and after an interim of ten seconds a second more specific word is pronounced. In the second three similar words are given at an interval of fifteen seconds; the third four words with an interval of ten seconds. The following are examples of the first and third divisions:

SUBJECT III

Commencement--Sieve

commencement | college commencement | cap college | gown cap | boys commencement | confetti commencement

sieve - holes sieve - water sieve - flour sieve - space holes - concept space - Royce concept - time concept - eternity time - damnation eternity - Hamlet "Consummation," etc - Shakspere Hamlet

SUBJECT VII

Wax--Jug--Tar--Sod

wax | Charley wax | picnic Charley | horse Charley | saddle horse

jug - ink jug - clay jug - Hegel jug

tar / 'old tar' tar / ship 'old tar' / Bermuda 'old tar'

sod \ grave sod \ graveyard grave \ house graveyard \ church house \ music church \ white church

In all the experiments the subject simply knew that possibly more than one starting-point would be given. There was of course the conscious recognition on the part of the subjects that the pronounced words were starting-points, which would imply an attentive consciousness, but they were cautioned neither to favor nor inhibit the newly pronounced word nor an association in progress.

The notes are uniform in showing that often one, two or three words of the former association-series are written after the new word is pronounced. "The momentum," says F., "was great enough to carry the associations two or three words beyond the pronounced word"; while Bl. found "a tendency for the trend of associations to persist, though not strong enough to overcome the new influence." By.'s experience was slightly different. As stated before, he often wrote the word as it came into consciousness. "On hearing a new word it gets precedence over the next associations not yet formed, and there is considerable confusion and lost time unless the motor discharge of writing the pronounced word is permitted to have free expression." The tables verify the same, and also show that there are more associations during the first interval.

Does a former starting-point regain its influence? In the first division there are two cases where the first and second fuse, but no place where the first independently forms an association; in the second but one word for subject Bl. in "Quill--Bench--Chalk," and in the third not any. There was a small amount of fusion in all, since but two words are due to the combined influence of the first and third, five to the combined influence of the first and fourth, with three starting-points, and one to the combined influence of the first and third with four starting-points.

The train of associations is inhibited by a new starting-point which dominates in influence. No mention is made in any note that a former starting-point remains in consciousness for the series, but M. emphatically writes, "Absolutely no influence of the preceding word or words when the next is taken up"; and later, "As soon as the new one is pronounced the old word and the series it had brought up were immediately suppressed." Bl. comments, "How remarkable it is that each new word crowds the old trend of associations out and starts new ones"; and the graphic representation, one of which only is given here on account of lack of space, shows that there is no return to the original series.

The tables are indicative of the tendency. In the first division of the group there are three possible lines of fusion, in the second six possible lines, and in the third twelve possible lines, but we find only 13.2% for all forms of fusion in the first, 7.9% for the second, and 10.5% for the third. In the eighty-seven experiments of the series there is but one absolute return to the previous starting-point. (See Group IX, sec. 2, Bl.) The tables show the varying degrees of fusion, and while the percentages have little meaning, as there is a variable time-element, the numbers do show accurately the number of words and the relative and continued influence of each starting-point.

We conclude that, when the present content of consciousness is a series of associations, the newly given impression establishes itself sufficiently to inhibit the associations of the previous series.

TABLE XV. TWO WORDS--GENERAL AND PARTICULAR--SPOKEN

Time--50 seconds, with an interval of 10 seconds between 1 and 2. Characters--same as Table I.

(1) Commencement--Sieve

M. | | | - - - - - - - - - - - - - - F. | | | | | | + + + + + + + H. | | | | | - - - - - - - - - - - - V. | | | | | - - - - - - - - - - - - S. | | | | | | | | - - - - E. | | | - - - - - - - - - Bs.| | | - - - - - - - - - - - - - - Ht.| | - + + + + + + By.

| - + α β M. { 3 14 0 0 0 {17.6% 82.4%

F. { 6 7 {46.2% 0 53.8% 0 0

H. { 5 12 0 0 0 {29.4% 70.6%

V. { 5 12 0 0 0 {29.4% 70.6%

S. { 8 4 0 0 0 {66.6% 33.4%

E. { 3 9 0 0 0 {25.0% 75.0%

Bs. { 3 14 0 0 0 {17.6% 82.4%

Ht. { 2 1 6 0 0 {22.2% 11.2% 66.6%

By. 0 0 0 0 0

Totals {35 66 13 0 0 {31.4% 55.4% 13.2%

Number of subjects, 9; number of groups, 1; number of experiments, 8; number of associations, 114.

TABLE XVI. THREE WORDS SPOKEN

Time--50 seconds. Interval, 15 seconds.

Characters: | first, - second, / third, + partial fusion between first and second, α partial fusion between first and second with first predominating, β partial fusion between first and second with second predominating, γ partial fusion between first and third, δ partial fusion between second and third, ε partial fusion between first and third with third predominating.

(1) Gun--Bug--Jaw

M. | | | | | | | | - - - - - - / / / / / / F. | | | | | - - - δ δ δ H. | | | | | | - - - - / / / / / V. | | | | | | | | - - - - - - / / / / / / Bl. | | | | | | - - - - - / / / / / / / By. | | | | | | | | | - - - - - - / / / / / / Bs. | | | | | + + + + / / / / / Ht. | | | | | | | α | | | δ - - - - Ro. | | | | | | - - - - - - / / / / / /

(1) Gun--Bug--Jaw (2) Quill--Bench--Chalk (3) Hall--Moss--Leather

| - / + α β γ δ ε

M. {20 14 18 1 0 0 0 0 0 {37.8% 26.4% 33.9% 1.9%

F. {15 11 15 0 0 0 2 3 0 {32.6% 23.9% 32.6% 4.4% 6.5%

H. {19 11 15 0 0 0 0 0 0 {42.3% 24.4% 33.3%

V. {24 15 21 1 0 4 0 0 0 {36.9% 23.1% 32.3% 1.5% 6.2%

Bl. {15 9 9 0 0 0 0 0 0 {45.4% 27.3% 27.3%

By. {23 8 14 3 0 4 0 3 0 {41.8% 14.5% 25.5% 5.5% 7.2% 5.5%

Bs. {13 10 10 4 0 0 0 0 5 {30.9% 23.8% 23.8% 9.6% 11.9%

Ht. {23 11 10 1 1 2 0 1 0 {46.9% 22.5% 20.5% 2.0% 2.0% 4.1% 2.0%

Ro. {15 15 24 0 0 0 0 0 0 {27.8% 27.8% 44.4%

Totals {167 104 136 10 1 10 2 7 5 { 37.8% 23.5% 30.8% 2.3% .2% 2.3% .4% 1.5% 1.2%

Number of subjects, 9; number of sections, 3; number of experiments, 27; number of associations, 442.

TABLE XVII. FOUR WORDS SPOKEN

Time--50 seconds. Interval, 10 seconds.

Characters -- same as Table XVI, with addition of \ representing the fourth, ς partial fusion between first and fourth, η partial fusion between second and fourth, θ partial fusion between third and fourth, ☐ total fusion between first, second, third, and fourth, ι partial fusion between first, third, and fourth, κ partial fusion between second, third, and fourth, λ partial fusion between third and fourth, with fourth predominating, [ partial fusion between first, second, and third.

(1) Den--Nag--Cot--Fan

F. | | | | | | - - - - - - δ - - \ \ \ \ \ H. | | | | | - - - / / / / / / \ \ \ \ \ \ \ \ \ V. | | | | | - - - / / / / \ \ \ \ \ \ \ Bl. | | | | - - - - / / / \ \ \ \ By. | | | | | | - - - - - δ δ δ \ \ \ \ \ Bs. | | | - - / / / \ \ \ \ \ Ht. | | | | | - - - - - / / / / \ \ \ \ Ro. | | | | - - - - / / / / \ \ \ \ \

(1) Den--Nag--Cot--Fan (2) Tax--Fan--Map--Dog (3) Paw--Wand--Box--Mug (4) Bud--Car--Cub--Mat (5) Wax--Jug--Tar--Sod (6) Cur--Elk--Pug--Man (7) Rope--Wig--Ink--Grass

(Table on page 460)

GROUP IX

The aim here was to see if it were possible to have the first starting-point such that the conditions would be similar to the results obtained by using preparatory marginal settings, but include rather than inhibit the second starting-point. A review of the tendency toward mental combination in the former experiments suggested that the words be in the relation of the whole and part. The "part" was given the position of predominating influence in order to see to just what extent it would persist in combining.

| -- / \ + γ δ ς η θ

F. { 29 19 11 19 3 0 4 0 0 11 { 30.3% 19.8% 11.4% 19.8% 3.1% 4.2% 11.4%

H. { 31 18 24 25 3 0 0 0 0 8 { 28.5% 16.5% 22.1% 22.9% 2.7% 73%

V. { 28 28 21 50 0 0 0 0 0 0 { 22.0% 22.0% 16.6% 39.4%

Bl. { 18 17 17 29 0 0 0 0 0 0 { 21.6% 20.5% 20.5% 34.9%

By. { 36 22 9 24 2 4 3 0 6 2 { 32.1% 19.6% 8.0% 21.4% 1.8% 3.7% 2.6% 5.3% 1.8%

Bs. { 22 13 17 18 4 1 0 1 0 6 { 24.2% 14.3% 18.7% 19.8% 4.4% 1.1% 1.1% 6.6%

Ht. { 30 18 10 21 3 0 0 0 2 1 { 34.4% 20.6% 11.4% 24.2% 3.4% 2.4% 1.2%

Ro. { 24 22 16 48 0 0 3 0 0 0 { 21.3% 19.4% 14.2% 42.4% 2.7%

Totals{218 157 125 234 15 5 10 1 8 28 { 26.6% 19.2% 15.2% 28.5% 1.9% .6% 1.3% .2% .9% 3.4%

☐ ι β κ λ [

F. 0 0 0 0 0 0

H. 0 0 0 0 0 0

V. 0 0 0 0 0 0

Bl. 0 0 0 2 0 0 2.5%

By. 4 0 0 0 0 0

Bs. 0 3 1 0 5 0 3.3% 1.1% 0 5.4% 0

Ht. 0 0 0 0 0 2 2.4%

Ro. 0 0 0 0 0 0

Totals{ 4 3 1 2 5 2 { .4% .3% .2% .3% .7% .3%

Number of subjects, 8; number of sections, 7; number of experiments, 52; number of associations, 819.

In the light of the interpretation of previous facts and the subsequent results of this group of experiments, we are now in a position to conclude that, if the present content of consciousness on the reception of a new impression is such that the reactions are not antagonistic but reënforce each other, the second will not persist in independent influence, but will be rather included in and supplementary to the influence of the first, which otherwise would be less assertive. The results below show that while the first of the two starting-points has a decided disadvantage of position, and therefore has little influence in arousing associations, it here is responsible for 43.2% and the second for but 9.5%, while there is a combined influence of 47.3%, the first strongly predominating in partial fusion. There is but one case where the second was followed exclusively.

The explanation of this, extended to an hypothesis, would rest in the fact that each word has definite characteristic reactions and that the fusion of two words or lines of thought means that the motor accompaniments are such that they unite and reënforce each other, or that the one includes the other. There are few antagonistic impulses.

TABLE XVIII

TWO WORDS--WHOLE AND PART--SPOKEN

Time--15 seconds. Characters--same as Table I.

(1) Crowd--Man

A. | + + M. + + + + + + + + + F. + + + + + H. J. | | | | | | + + + S. - - - - V. + + + α α α α α α E. | + + + L. | | | | Bs. α α α α α α Br. | | | | | + +

(1) Crowd--Man (2) Ton--Pound (3) Moose--Horn (4) Engine--Whistle (5) Book--Page (6) Music--Octave

| - + α β

A. { 4 3 24 0 0 { 12.9% 9.6% 77.5%

M. { 13 0 27 2 0 { 30.9% 64.2% 4.9%

F. { 7 3 22 1 0 { 20.6% 8.8% 64.8% 5.8%

H. { 28 5 0 0 5 { 73.6% 13.2% 13.2%

J. { 21 1 17 6 0 { 46.6% 2.3% 37.8% 13.3%

S. { 25 12 0 0 0 { 67.5% 32.5%

V. { 4 2 16 10 0 { 12.5% 6.3% 50.1% 31.1%

E. { 11 3 9 5 0 { 39.2% 10.8% 32.2% 17.8%

L. { 14 6 4 0 0 { 58.4% 25.0% 16.6%

Bs. { 15 2 11 6 0 { 44.2% 5.8% 32.4% 17.6%

Br. { 25 0 9 9 0 { 58.2% 20.9% 20.9%

Totals { 167 37 139 39 5 { 43.2% 9.5% 35.9% 10.0% 1.4%

Number of subjects, 11; number of sections, 6; number of experiments, 62; number of associations, 387.

GROUP X

In the last group the words were called whole and part, although the reader no doubt observed that this was not always strictly the case, but rather that the more complex was more influential as a word-suggester than the simple and often tended to include it. Can this be demonstrated in another way? Attention was called in Group of Experiments IV (two pictures shown) that picture no. 3 (a man sawing wood), although first, had more than the normal number of associations, the explanation of which might lie in the fact that it contained more objects. We here attempt to test this by comparing a comparatively complex and a comparatively simple picture as starting-points. It has been clearly proven that the second picture to be presented has a decided advantage of position. The simple picture is here given the advantage of position. The notes show the first to be more important, while there is no suggestion of the first including the second, on account of there being less fusion than for words and more independent associations for the second starting-point; the averages are 38.5% for the first, 30.5% for the second, 31.0% fusion.

(a) Picture of Blacksmith shoeing horse (b) Picture of a Sheep. | shop (a) | nails shop | shoe shop and (a) | blacksmith shop and shoe and (a) | Longfellow blacksmith - pasture (b) - sheep pasture - lambs pasture and sheep - grass pasture and lambs - hillsides grass - brook pasture and hillsides | iron shop-shoe | hammering iron | soldier iron-hammering and first group | battle soldier | shoe battle | horse-shoe nail iron, hammering, soldier, battle | anvil horse-shoe nail

TABLE XIX. TWO PICTURES SHOWN

Time--50 seconds. Characters--same as Table I.

(1) (a) Blacksmith shoeing Horse (b) A Sheep

J. | | | | | - - - - - - | | | | | | | E. | + + - - | | | W. | - | - - - | | | | | | | V. | | + + + + + + + + + + + + + L. - | | | | | | | | Ht. - - α α | | | | - + + + + +

(1) (a) Blacksmith shoeing Horse (b) A Sheep

(2) (a) Girl and Boy (b) A Bird

(3) (a) Three Children (b) A Duck

(4) (a) A Sower (b) A Dog

| - + α β

J. { 30 23 9 18 0 { 37.5% 28.8% 11.2% 22.5%

E. { 28 10 2 0 0 { 70.0% 25.0% 5.0%

W. { 14 18 5 0 0 { 37.8% 48.7% 13.5%

V. { 6 11 37 0 13 { 8.9% 16.4% 55.3% 19.4%

L. { 27 4 1 0 0 { 84.3% 12.5% 3.2%

Ht. { 16 30 6 6 0 { 27.6% 51.8% 10.3% 10.3%

Totals {121 96 60 24 13 { 38.5% 30.5% 19.2% 7.6% 4.2%

Number of subjects, 6; number of sections, 4; number of experiments, 24; number of associations, 314.

GROUP XI

Three words spoken, one immediately following the other, are given in this group in order to test the span of consciousness and the influence of an immediate interruption, as one was given immediately following the other in such a way that all associations were checked until after the third starting-point. If we attempt to follow the initial starting-points we see they may disappear after the first few associations and reappear in the series; or each remain as initial starting-points for a few associations; or one alone control the whole series, while the others are present without influence; or there may be an alternation of independent influences; or an influence which shows a modifying effect of one or both of the other starting-points, which may reach such a degree that all fuse, and in so doing get an advantage over the single words. The starting-points are to a large extent disparate and there is very little fusion; we find only .5% total fusion and 18.7% partial fusion.

An example would be as follows:

SUBJECT VIII

FROG--ICE--TABLE

- snow ice | bench table | fish frog | pole fish | boat fish | line fish | bait fish | weeds fish - skating ice - sleds ice - coasting sleds - girls coasting

There were no cases of the total fusion of all the starting-points throughout the series, no cases of total fusion of the first and second, and two cases for the second and third. There were two cases of partial fusion of the first, second, and third throughout. The first was followed exclusively in two cases, the third in one. The first and third were followed intermittently in four cases, the second and third in eleven, the first, second, and third in thirty-five.

The second starting-point has the ascendency of influence. This is not due to habits formed in earlier experiments, as the two groups of experiments with two words were given in connection with other experiments. This group informs us that in immediate interruption the new impression has not sufficient power to establish itself more firmly than a past impression which is just past, and must in a manner be reproduced to start associations. The influence of the first has been somewhat destroyed, but the second is greater than the third.

TABLE XX. THREE WORDS SPOKEN

Time--50 seconds.

Characters--same as Table XVI, with the addition of [ representing total fusion between the first, second, and third, μ partial fusion between first, second, and third with first predominating, ν partial fusion between first, second, and third with first and third predominating.

(1) Frog--Ice--Table

A. / - - - - - M. - | | γ γ γ γ γ γ γ γ γ F. - - - - - / / | | | | / / / - - - H. | - - - - - | | | | / / / - - - J. | - - / / / / - + + + + + + + + Bl. - / | | | | | | - - - - By. | / / / / / / - - - - - - - - - - - - - - - - Bs. - | | | - - - / / Ro. | | - - / / / / / / / / / / / Br. - / / / δ δ δ δ δ δ

(1) Frog--Ice--Table (2) Key--Shoe--Knife (3) Desk--Park--Glove (4) College--Church--Cafe (5) Boston--Elevator--Lake (6) Skate--Book--Theatre (7) Bridge--Sleigh--Ticket (8) Gun--Lamp--Watch

| - / + α β γ δ [ μ ν A. { 26 32 30 0 0 0 0 0 2 10 0 { 26% 32% 30% 2% 10%

M. { 25 36 21 3 0 0 10 1 1 0 6 { 24.3% 34.9% 20.5% 2.9% 9.8% .9% .9% 5.8%

F. { 15 23 24 8 11 0 7 6 0 0 0 { 15.9% 24.5% 25.5% 8.5% 11.8% 7.4% 6.4%

H. { 31 68 31 0 0 0 0 0 0 0 0 { 23.8% 52.4% 23.8%

J. { 19 18 48 15 0 0 0 22 1 0 0 { 15.5% 14.6% 39.1% 12.2% 17.8% .8%

Bl. { 23 33 16 5 0 0 0 22 0 0 0 { 23.2% 33.2% 16.0% 5.0% 22.6%

By. { 13 56 48 15 0 1 17 2 0 0 0 { 8.7% 36.8% 31.5% 9.8% .6% 11.2% 1.4%

Bs. { 19 13 19 9 0 0 0 19 2 0 0 { 23.4% 16.0% 23.4% 11.2% 23.5% 2.5%

Ro. { 32 33 33 0 0 0 0 0 0 0 0 { 32.6% 33.7% 33.7%

Br. { 12 24 38 0 0 0 6 0 0 0 0 { 15.0% 30.0% 47.5% 7.5%

Totals{215 336 308 55 11 1 40 72 6 10 6 { 20.3% 31.1% 29.1% 5.2% 1.1% .9% 3.6% 6.8% .5% .9% .5%

Number of subjects, 10; number of sections, 8; number of experiments, 79; number of associations, 1060.

GROUP XII

In order to test the relative influence and to throw more light on the problem of immediate interruption four words were shown and the subject was directed to read from left to right. There were few experiments and most of the subjects were new. The results show that the third starting-point has a decided disadvantage, having an influence of only 7.8%. The second and fourth are almost equal, while the first is again less.

TABLE XXI. FOUR WORDS SHOWN

Time--50 seconds.

Characters--same as Table XVII, with the addition of ξ partial fusion between the second, third, and fourth with the second predominating.

(1) Gun--Car--Ink--Fan

M. \ \ \ - - - - - / / / / / / R. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ K. \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ S. - | / - - - - - - - - J. | | | - - - - - - - - - - - Ht. + + α α α | | | | | | | | |

(1) Gun--Car--Ink--Fan (2) Brain--Umbrella--Telephone--Chain (3) Book--Money--Hour--Chart

| - / \ + γ θ α κ ξ M. {14 6 9 4 0 6 0 0 0 0 {35.9% 15.4% 23.0% 10.3% 15.4%

R. {16 0 0 30 0 0 0 0 0 0 {34.8% 65.2%

K. { 0 35 0 16 0 0 0 0 0 0 { 68.6% 31.4%

S. { 5 15 7 4 0 0 0 0 0 0 {16.2% 48.4% 22.5% 12.9%

J. {13 13 4 15 2 0 0 0 0 0 {27.7% 27.7% 8.5% 31.9% 4.2%

Ht. {15 9 0 5 2 0 7 3 1 1 {34.8% 20.9% 11.6% 4.7% 16.3% 6.9% 2.4% 2.4%

Totals{63 78 20 74 4 6 7 3 1 1 {24.5% 30.4% 7.8% 28.5% 1.6% 2.4% 2.6% 1.4% .4% .4%

Number of subjects, 6; number of sections, 3; number of experiments, 18; number of associations, 257.

GROUP XIII

Four similar words were pronounced in immediate succession, the results of which show that we are correct in calling the above a case of succession and also to establish clearly and definitely that there is a difference of degree between immediate and postponed interruption. The third starting-point has again a decided disadvantage of influence and exerts no apparent influence in thirty-seven of the fifty-two experiments. The last starting-point exerts the greatest influence.

TABLE XXII. FOUR WORDS SPOKEN

Time--50 seconds.

Characters--same as Table XVII with the addition of ο partial fusion between first, second, and third with first and second predominating.

(1) Cow--Roof--Fence--Girl (2) Cathedral--River--Elevator--Newspaper (3) Cane--Harness--Box--Coat (4) Book--Snow--Rope--Stone (5) Wire--Flower--Horse--Paper (6) Gun--Wharf--Chair--Stove

| - / \ + λ δ η θ ☐ ι ο ε β M. { 35 3 6 32 0 3 17 0 9 0 0 0 0 0 { 33.4% 2.8% 5.8% 30.4% 2.8% 16.2% 8.6%

F. { 16 11 1 35 0 3 0 0 1 2 1 0 5 0 { 21.3% 14.6% 1.4% 46.6% 4.0% 1.4% 2.7% 1.4% 6.6%

H. { 21 38 11 30 0 0 0 0 0 0 0 1 0 1 { 20.6% 37.3% 10.8% 29.5% .9% .9%

V. { 16 16 0 36 2 2 9 0 1 0 1 0 0 0 { 19.3% 19.3% 43.3% 2.4% 2.4% 10.9% 1.2% 1.2%

Bl.{ 36 8 10 19 0 0 4 0 0 0 0 0 0 0 { 46.8% 10.3% 12.9% 24.8% 5.2%

By.{ 23 30 3 36 1 0 0 0 1 0 0 0 0 0 {24.5% 31.9% 3.3% 38.3% 1.0% 1.0%

Bs.{ 22 14 1 20 1 0 0 7 0 1 0 0 0 0 {33.3% 21.3% 1.6% 30.0% 1.6% 10.6% 1.6%

Ht.{ 9 19 4 31 0 0 0 0 1 14 0 0 0 0 {11.5% 24.3% 5.2% 39.8% 1.3% 17.9%

Ro.{ 26 23 6 16 0 0 0 0 0 0 0 0 0 0 {36.6% 32.4% 8.4% 22.6%

Totals {204 162 42 255 4 8 30 7 13 17 2 1 5 1 {27.2% 21.6% 5.4% 33.9% .5% 1.1% 3.9% .9% 1.8% 2.3% .3% .2% .7% .2%

Number of subjects, 9; number of sections, 6; number of experiments, 52; number of associations, 751.

We ask finally how far our results and notes point to a theoretical understanding of the mechanism of associations. Previous work, especially that of James, Cordes, Calkins, and Scripture, as well as the accumulated notes of my subjects, confirm that the transition may be made by means of total, partial, and focal recall, and that in partial and focal recall the prominent persisting elements are surrounded in the formation of a new idea by other new elements.

If a latent idea remains in the margin of consciousness and exerts an influence, which not merely modifies but determines the series of associations, and leads up to the focalisation of the latent idea, we have a case of predetermined association, which, when noted by investigators, has invariably become confused with mediate association. Here there is an element or group of elements, persisting in the margin of consciousness, which is gradually maturing and becoming focalised into groups of elements comprising an idea which ultimately dominates consciousness. In some cases three, four, and five ideas have been named before this takes place, and we have here a reversed form of association. Four subjects noted the experience on different occasions, and it is not to be confused with the common experience of apprehending the present contents of consciousness as part of a larger whole where we are conscious of its existence but not of what it is.

The notes further show that the common conscious elements may be predominantly visual, auditory, olfactory, gustatory, or kinæsthetic, or a complex or compound of these in character, while to this may be added an indication of the fact that the transition, incipient as it is, may in many cases be reduced to a condition which is in the last analysis one of the motor nervous system. Ht., for instance, finds that the words all pass over into innervations of the organs of speech and "are accompanied by the impulse to make the sound," stating later, "they hang on the tongue." The following is one of the series given which represents rather an extreme case, Taft, taffy, toffy; tough, rough, ruff; buff, bluff, tough; muff, duff, tuff. Br., who also gave a large percentage of verbal associations, finds that "some part of each word seems to linger on the tongue with motor sensations till the next comes." "I am subject," he adds, "more or less frequently to verbal automatism of this auditory incipient motor type." Ro., who has many auditory associations, reports "they are always accompanied by motor images, together with many associations." A changing of orientation is a common accompaniment, with statements of the feeling of the impulse to turn in various directions. For F., who is predominantly of the motor type, we have an example where the rhythmic ticking of a clock fades into the rhythmic watching of a boat rising and falling on the water.

The notes would seem to indicate that there is no idea without a motor fringe, and also that these elements of incipient impulses to movement may accompany the elements of transition, and are observed introspectively by the subjects. They are therefore data for psychology. Do they influence or direct the associations? In short, are they the processes which connect and which determine the associations?

F. states, "There seem to have been waves of motor sensations. Such waves may start with a word and carry one in faint mimicry through the whole succession of bodily sensations that one experienced in that event, and then may come a relapse until other stronger currents appear." Here we are face to face with the dynamics of association, the most fundamental and important problem of brain association. Have these phenomena of ideational images "acquired by contact a kind of magnetism which causes the one to attract the other and have, so to speak, become magnetic?" (Zanotti.) Or are they on the other hand independent of all force and "merely ideas of antecedence and sequence only?" (Mill.) While there is no mention of a magnetic force, the notes and results all show that the ideas are systematically conditioned in a way which cannot be explained by the contiguity of the objects. The motor elements play the deciding role. Ht. emphasises the influence of ideated movement when he writes, "Kinæsthetic. Slow regular tramping on snowshoes brought up the characteristic swing of arms, and therewith the idea (sensations of weight) of the stick (or stock) which I have generally carried on Norwegian snowshoes. Transition from Vermont to the Black Forest by association with snowshoeing in both places. Real sensations in play were free breath, movements in chest (kinæsthetic), fresh air (olfactory), cold (thermal), and emotion of emotional strength." Again, "Looking up at sun suggested general ideas of expansion of attention and with this breath comes the idea, breezes"; another subject adds, "A tendency to imitate the sounds of syllables and this leads on to a train of associations"; another, "A slight feeling of sudden changed impulse"; another, "A sort of motor after-image came back and took the foreground"; and F. goes further when he states, "Ideationally my hand wandered to the upper right-hand corner of the page, then suddenly the auditory image of 47 came up as if whispered to me." All of which indicate that some ideas at least depend for their entrance into consciousness upon motor reactions.

Passing to the more refined reactions expressed in emotions we find that they are not merely accompanying coloring influences, but also often actual determining factors. All of the subjects notice at some time a coloring atmosphere from an emotion, but others find that "the growing word is rather felt emotionally than definitely formulated," and we have "a nameless idea, largely feeling-tone" (Ht.); or the words may "all come as parts of a growing feeling, an indistinct though strong state of mind." (J.) The same subject observed, "The previous word may create a mood or feeling which in the main determines the associations; a group of words is dependent upon strong accompanying feeling--there is a summation and a discharge while the next word has been accumulating force" (J.), and we have a form of summation; or in other words, "a general mood accumulated while several words were in mind at once, then all dropped and another general feeling came to the front with an accumulation of other words." (F.) Here we have a typical example of constellation where all the words and ideas are implicitly present as a total attitude or disposition, the elements of which become successively focalised into a series of associated images. The last subject finds that "the emotional atmosphere often controls the associations." Indeed, it would seem that occasionally for some subjects this strong accompanying undercurrent of undifferentiated emotional feeling is capable of bringing about trains of thought independent of any logical connection. K. finds "the feeling to carry one on"; H. finds the "point of departure the interesting idea"; all find that the words change with the disposition, as may be verified by a study of the lists of associations.

We are forced to conclude that the impulses to movement or other emotional attitudes may act as determining factors in association, which extended to an hypothesis would mean that the mode of transition in the associated series is in the last analysis to be found in delicate incipient motor tendencies to action, the psychic concomitants of which are observable; that psychic states are both as to their unity and organisation consequences of motor reactions which are implicitly present as parts of a total reaction to the present situation. It is these motor tendencies to action which determine what idea shall enter consciousness. Just in so far as they become released they become prolonged, accentuated, and form a nucleus for the new idea. To speak of association independent of motor elements is merely to make an empirical classification of successive states of consciousness.

There remains a psychical phenomenon which must be satisfactorily accounted for before we go farther. An element of an idea, an idea or a series of ideas may occupy consciousness to the exclusion of others. If the second starting-point were not given, the associations would undoubtedly follow the given one. Inhibition must then be one form of "obstructed association," the inhibiting ideas being present to the exclusion of the inhibited. But are we thus forced to say inhibition is the "negative side of the association process," claiming that all ideas not in consciousness are inhibited, and thus being forced to conclude the conscious idea is inhibiting an unconscious idea, which cannot exist (by the very definition and presuppositions of psychology) until it is an object of consciousness. This would mean that content of consciousness and inhibition are identical. On the other hand, the notes and exemplifying facts of the tables show Dr. Breese's fallacious position when he concludes that "because, obeying the laws of association, the train of ideas takes one direction rather than another can hardly be considered sufficient ground to hold that the other possible train of ideas is inhibited." He has overlooked the possibility of two or more trains of associations having been started and the associations of one starting-point are excluded from entering the focus of consciousness by the direction of the given series. Inhibition would then be the negative side of fusion. The explanation must, as has already been demonstrated, be psycho-physical in character. If these impulses to action have actually been observed by the subjects we are justified in concluding that just as in physiological inhibition one action excludes another, so the correlative tendencies to movement of one idea exclude others.

By. observed that the image of the starting-point lingered and inhibited subsequent ideas. The implication here, from our previous reasoning, would be that not the ideational images, as such, but the physiological motor concomitants, persisted and excluded others, and this is why disparate terms give a "shock to the nervous system" (A.), "require different lines of expression" (A.); and "one has more momentum," as so many report. This would explain why the associations of a new starting-point inhibit the associations of a former one; for as the motor nervous impulses tend to work themselves out into action, the reaction of the previous impulse will be suppressed by those of a new impulse which enters, by the conditions of these experiments, an attentive consciousness. Thus the prepotent impulses to action are the conditioning factors in mental inhibition.

All this indicates that the basis of habit which has been the universal principle of explanation of associations is inadequate. As Münsterberg has pointed out, contrary to what we mean by habit, either idea may bring to consciousness the other, in a manner independent of the order of the original presentation. Extending our hypothesis to include the formation of associations, the conclusion will be that in order for two ideas to become associated they must be together in consciousness, each as parts of a total experience, a total attitude; the motor reactions of the ideas must be parts of a more comprehensive reaction which includes both as simultaneous correlated motor impulses: when, in future time, the reactions of the one are reëxperienced, there is a sequence of infinitely delicate and complex impulses to movement, and any tendency toward such reaction tends to reproduce the whole of which it is a part, as each reaction is more or less bound up in the integrity of the whole central nervous system.

FOOTNOTES:

DISSOCIATION

BY C. H. TOLL

The purpose of this investigation, of which the following gives a preliminary report, was to compare the tendency to associate by contiguity, with the tendency to associate by similarity.

In every series of stimuli to which one gives attention there is tendency to association by contiguity. But some similarity among certain elements of the series may produce a dissociation of the given elements into two series with some bond of similarity in each. This is a matter of common experience, as when you find you can read your newspaper and listen to your neighbors' conversation at the same time, understanding both, although the actual order in which the several words are perceived would form a meaningless mixture.

We may say dissociation is always accomplished by a tendency to association by similarity overcoming the constant tendency to association by contiguity. Study of the relative efficacy of the two may therefore be called a study of dissociation. The tendency to associate by contiguity might be measured in two ways.

First, when one attempts to learn a series in exactly the given order, the number of errors in the series as recollected may be taken as an inverse indication of the strength of association by contiguity. The three kinds of error possible in nearly all of the experiments were Omissions, Displacements, and Imperfections. All of these three have been tabulated. But the number of elements omitted seems considerably the most reliable as an indication of the degree of inadequacy of the associative tendency. The cases of displaced or imperfect elements are comparatively few: moreover, Displacements and Imperfections are not mutually exclusive categories. A single element may be both imperfectly recollected and wrongly placed in the recollected series. On the whole, it seems that the number of given elements which were omitted in the recalled series is the most positive and reliable of the errors. Our conclusions are based on the Omissions.

Second, when one makes no attempt to learn the series, simply giving attention to each element as it comes, and afterward lets the elements recur spontaneously, the number of cases in which a recollected element is followed by an element given contiguously may be taken as a direct indication of the strength of association by contiguity.

Tendency to association by similarity can evidently be measured in the same two ways, by counting errors when one purposes to learn the series as two groups of similar elements, and by counting sequences of similar elements when one avoids any effort to learn the series and recollection is spontaneous.

In the first seven experiments we used the first method. The errors made when the purpose is to associate by contiguity can then be compared with the errors made when the purpose is to associate by similarity, an equal number of series, given under the same conditions, and of identical character, being given in each case.

In the last four experiments we have used the second method. The number of sequences of elements given contiguously can then be compared with the number of sequences of similar elements.

Five subjects have coöperated in this, but the experiments were strictly individual, one observer being alone in the room with the experimenter. Each test lasted about an hour. As a matter of course, the results have been calculated for each of the five subjects and their agreements and disagreements have been carefully considered. But as this first report is to indicate merely the general tendency, we give here at first only the average of the five persons.

The experiments have varied as to the kind of elements used, the manner of presentation, the time allowed, and the manner of recording the recollected series. But throughout each experiment the series were of one identical type, while the individual elements were altered in each series.

In the experiments where the series were to be learned, some in the given order, some dissociated by similarity, it was found rather confusing to turn from one method to the other; so several consecutive series were learned by one method, and then several by the other, four alternations being made each hour to neutralize any effect of practice or of fatigue.

The series were of course different in kind in the several experiments, but were usually of eight or of ten elements. Half of this number had some distinct characteristic in common, the other half some other characteristic. In some experiments these elements were alternated, in some arranged irregularly.

In the first eight experiments the subject wrote down the elements recalled, as soon as the series had been given. In the last three the subject spoke the elements recalled.

In all cases where the first method of measurement was used, the time allowed for learning the series was made a little too short to permit of learning the series perfectly. Since comparison of the number of mistakes was our method, we naturally had to make sure there would be mistakes to compare.

The details of the several experiments were as follows:

(1) The elements were letters and numbers. They were about 12×8 mm. in size and were printed on white cards 15×30 mm.

Five letters and five numbers were placed, alternately, in a straight row on a sheet of white cardboard. The series was then exposed to the subject by turning up the small tin shutter of a screen that was clamped to the table-edge.

The time of exposure was measured with a stop-watch and was constant throughout the hour for each individual subject. Four seconds proved the best time for most of them, but in one case it was necessary to allow only three seconds. Twenty series were presented during each hour, ten for each method of memorizing. There were duplicates of all the numbers, and of eight letters, but not more than two of any element. Selection in forming the series was by chance. In dissociating, the letters were separated from the numbers.

As soon as the exposure was ended, the subject wrote down the elements recollected, trying to preserve their relative order. This recollected list was then copied beside the operator's record of the given series, so making the errors apparent.

(2) The elements were all letters, printed as before, and the alternate cards were placed half their length out of alignment with the original row.

The method of presentation, the length of exposure, the number of elements presented, etc., were as in no. 1. In dissociating, the letters on one level were separated from those on the other.

(3) The elements were all letters, printed as before, and five of the ten elements presented were placed out of alignment. But the disaligned cards were at irregular intervals and often in groups, and were only a quarter of an inch out of alignment. This order was varied each time, but without any system.

The other details were as in no. 1.

During this experiment I came to notice the effect produced by the natural tendency to learn the five elements of the dissociated series in a rhythmical form, thereby increasing the ability to retain them; while there appeared to be no natural tendency to apply any such inclusive rhythm to the ten elements of the series when learned in the given order. To counteract this effect the subjects were instructed to consider the series, when learned in the given order, as two consecutive series of five elements each, and to use the same natural rhythm in learning these as they did in the dissociating. But this correction was not made in the first two hours, nor very perfectly in the rest.

(4) The elements were all numbers, printed as before, five of the ten being placed a quarter of an inch out of alignment, and in irregular groups, precisely as in the last experiment.

The time was reduced to three seconds for some and two seconds for the others. Details of presentation were as described in no. 1.

This time all the subjects tried to neutralize the effect of the instinctive rhythm for the five-element series by learning the ten-element series in two groups of five elements each.

(5) The elements were all nonsense syllables, each consisting of a vowel between two consonants, printed on white cards 20×20 mm. Eight of these were placed in an even row on a sheet of white cardboard, and four of them were marked by laying a quarter-inch strip of blue paper over the bottom of the card. The serial position of the marked cards was irregular, and was altered each time.

Ten seconds was given to some subjects, eight to the others. Other details of exposure, etc., were as in no. 1.

In learning the series in the given order, the blue markings were ignored; but in dissociating, the marked and unmarked syllables were learned in separate groups.

There seemed to be no rhythmical tendency; but to be safe the subjects were instructed to learn the straight series in groups of fours.

Seven series were given to be learned in each method during the hour with each subject.

(6) The elements were one-syllable nouns, alternated with nonsense syllables, all spoken by the operator. The nonsense syllables were all different from those used in the preceding experiment: the nouns were ordinary words, and were so arranged as to avoid any obvious sequence or relation among them. Very few, if any, were used twice in one hour. Five nouns and five syllables were given in each series.

The elements were spoken at the rate of forty-six a minute, timed by a metronome which was muffled in a heavily padded box so that its sound was no disturbing factor. The speaker sat within three feet of the subject and enunciated as distinctly as possible.

Dissociation was performed as previously: in each hour eight series were dissociated, and eight learned in the given order.

(7) The elements were one-syllable nouns, spoken as before, alternated with nonsense syllables, printed on small white cards. The nouns were all different from those used in the previous experiment: the nonsense syllables were the same, but were this time printed, in letters 10 mm. high, on cards 40 mm. square. They were exposed by sliding them, one at a time, in front of an opening in a cardboard screen which was fastened to the table-edge.

The optimum rate for presenting the elements was found to be about forty a minute, measured with the metronome.

Five nouns and five nonsense syllables were given in each series. Eight series were given to be learned in the given order, and eight to be dissociated into separate series of nouns and of syllables.

(8) The elements were names of mammals, alternated with names of cities of the United States, all spoken. The names were all fairly familiar. Ten elements were given in each series.

The interval in reading was planned to be long enough for some appreciation of the meaning of the words, but not enough to permit mental repetition of the preceding elements. Any mechanical time-measurement was found impracticable.

The subjects were instructed to avoid any effort to memorize the series, simply receiving each element as given.

After the last element there was a pause of about two seconds, to decrease the mere sound-recollection of the last few elements. Then the operator repeated, in an altered tone, one of the given elements. The subject at once wrote down the first element that came to mind, then the next, and so on.

In the seven preceding experiments the set of series presented had been different for each subject, though of course identical in character. But in this experiment and the following ones the lists of words were identical as read to each subject. The same element was repeated for each. Sixteen lists were given.

(9) The elements were nouns. In each series five names of similar objects were alternated with five names of a different sort of objects, e. g., names of fishes with names of poets. All were read, as before. In each series new sorts of objects were chosen. The subject never knew what sort of words were to be given; the subjects agreed this was not a disturbing factor to them, and it obviated the tendency to think what words would probably be given, as is natural when the general character of the series is announced beforehand.

The subjects were instructed to be passive during the reading, and during the four-second pause that followed, avoiding mental repetition of the words. Then the operator gave a signal and the subject repeated aloud the words as they happened to be remembered. The words being numbered on the list from which they were read, the operator was able to record the words as fast as spoken.

The subjects were instructed to give the word which they found to be foremost after they had spoken the preceding one, rather than to try to repeat a group of words which usually appeared simultaneously at the first effort of recollection, but which usually faded while one of them was being spoken.

The same sixteen series, of ten elements each, were given to each subject.

(10) The elements were nouns, the ten presented in each series all being names of similar objects, e. g., flowers. Five were spoken, alternated with five printed on small cards which were shoved in front of a 10×10 cm. opening in a cardboard screen fastened to the table-edge. Cards were 40 mm. square, the words printed by hand, but carefully, in letters 10 mm. high.

A series was given in about 13 seconds, but the time was not mechanically measured; it was at a rate which some practice showed to give a fair time to comprehend each element.

As before, the subjects were told to be passive until, after a four-second pause at the end of the series, the operator gave a signal. Then the recollected words were spoken.

The class of nouns was different in each series.

(11) The elements were nouns. In each series five of some familiar class were alternated with five of some other familiar class. The classes were different in each of the twelve series given.

From this regular series of ten, five were chosen irregularly, and were printed on cards as in no. 9. The remaining five, of course also irregularly placed in the series, were spoken. This irregularity was different in each series. Thus some words of one kind were spoken, the rest printed; some words of the other kind were spoken, the rest printed.

The other conditions were exactly as in the last experiment.

* * * * *

A table for the individual subjects, indicating not only the omitted but also the displaced and imperfect objects would have, for instance, the following character: C indicates that the effort was made to associate by Contiguity, S by Similarity.

SPOKEN NOUNS, ALTERNATED WITH PRINTED NONSENSE SYLLABLES

Nouns Omitted Syll. Omitted Displaced Imperfect C S C S C S C S Turley 13 16 21 14 7 13 6 10 Emerson 4 5 26 16 7 4 4 13 Miss Kent 5 8 15 8 18 5 9 4 Flexner 4 6 10 9 7 3 8 16 Toll 8 7 8 2 10 12 8 3

Total 34 42 80 49 49 37 35 46

If we consider total results only, and among them only the omitted elements, we come to the following percentages. They give the percentage of the errors of omissions among the elements recalled.

1. Letters and numbers alternated C 26. S 10.8 2. Letters, alternatingly disaligned C 21.2 S 15. 3. Letters irregularly disaligned C 23.8 S 22.4 4. Numbers irregularly disaligned C 7. S 20. 5. Nonsense Syllables, irregularly marked C 27.5 S 27.5 6. Nouns and Nonsense Syllables alternated, spoken C 35. S 37.2 7. Nouns and Nonsense Syllables alternated, nouns spoken, syllables printed C 28.5 S 22.7

In the second group, experiments 8 to 11, not the errors of omission, but, as explained above, the different kinds of reproduced elements, had to be analyzed with special reference to the question whether a sequence linked two contiguous or two similar objects. In the following table the total number of recalled sequences is taken as basis and the different kinds of sequences are given in percentages of it. The elements themselves are described above. B means a break, that is, a sequence without similarity or contiguity.

8. Dissimilar elements, similarly presented S 45 C 28 B 28 9. Dissimilar elements, different kind in each series S 53 C 25 B 21 10. Similar elements, dissimilarly presented S 54 C 20 B 26 11. Dissimilar elements, dissimilarly presented S (Meaning) 27 C 7 B 8 S (Presentation) 13.

The results by the first method of measurement may be summarized as follows, though the first and third conclusions are weakened by disagreement among the individual subjects.

A. When the only dissociating factor is some slight unessential feature (a bit of color on the card, a slight disalignment), this similarity and contiguity are nearly equally efficient. No. 3 and no. 5.

As this unessential feature is made more striking (disalignment half a card-length), the strength of similarity increases, only three fourths as many errors being made in dissociation as in contiguous association. No. 2.

The case of no. 4 (all numbers) is of little or no value. The time allowed for learning had to be made short enough to ensure the appearance of some errors; perfect recollection would obviously give no basis for comparison. And the time had to be so short in this case (only two seconds for some of the subjects) that the additional eye-motions and adjustments necessary in dissociating took time enough to spoil the results.

B. When the only dissociating factor is in the meaning of the elements (letters and numbers), this similarity is stronger than contiguity, only one half as many errors being made. No. 1.

The results of no. 6 do not support this proportion, but its results are not consistent, while those of no. 1 are.

C. When both meaning and manner of presentation are combined as dissociating factors (nouns and nonsense syllables, seen and heard), this similarity is stronger than contiguity, only three fourths as many errors being made.

But this method of measurement is not well adapted to series of auditory elements, so this experiment is unsatisfactory. No. 7.

The results by the second method of measurement may be summarized as follows:

A. When the only dissociating factor is in the meaning of the elements (names of different sorts of objects), this similarity is stronger than contiguity, twice as many similarity sequences as contiguity sequences being recalled. No. 8 and no. 9.

B. When the only dissociating factor is in the manner of presentation (to sight and hearing), this similarity is stronger than contiguity, nearly three times as many similarity sequences being recalled. No. 10.

C. When both meaning and manner of presentation are dissociating factors, these similarities are much stronger than contiguity, more than four times as many similarity sequences being recalled.

D. When these two dissociating factors are opposed to each other: (1) Four of the subjects show similarity of meaning much stronger than similarity of presentation, from two to five times as many similarity-of-meaning sequences being recalled. (2) One subject is strongly and consistently otherwise, giving nearly three times as many similarity-of-presentation sequences. No. 11.

MOTOR IMPULSES

THE ACCURACY OF LINEAR MOVEMENT

BY B.A. LENFEST

The starting-point for our investigation was the observation of Woodworth that there is a certain rhythm in which a certain hand-movement is made with the maximum of exactitude, and which represents thus an optimum for the periodical discharge of the particular motor centre. Our question was whether this rhythm is a constant one for all parts of the body, or whether different groups of muscles produce the greatest exactitude in different periods; further, whether secondary factors, like complexity of movement, resistance by weight, fatigue, etc., influence this psycho-physiological optimum.

The investigation, however, showed soon the necessity to consider the whole problem of the accuracy of rhythmical linear movements, and the experiments are thus not always directly related to our starting-point.

There is very little material published that can be collected under the subject head, accuracy of voluntary movement, and still less when the enquiry is confined to the accuracy of straight lines or linear movements.

The most suggestive contribution is that of Dr. Woodworth on the accuracy of voluntary movement. He has collected consistently what can be found up to the date of his publication, and the reader is referred to pages 7-16 of his monograph for the most reliable collection of authorities.

It must be said, as we run over the list from Goldscheider on the threshold of perceptible movement, through the results of Hall, Hartwell, Loeb, and Delabarre on "bilateral asymmetry" and comparisons of right and left hands; consider Fullerton and Cattell in their suggestive results, and Münsterberg's studies of movements; and finally take the testimony of Bryan as to the growth of accuracy of movement in children, that the vast accumulation of material bearing on reaction time--and similar phenomena would be of more value if concerned more with the accuracy and less with the production or perception of movement.

A paper by Miss M. K. Smith, in the Philosophische Studien for 1900, with the title, Rhythmus und Arbeit, concerns the influence of rhythmical action upon the quality and quantity of work performed. The method was to commit to memory nonsense syllables and letters.

The results show a tendency to take up a certain rhythm, especially in the later results and after practice; easier memorizing if rhythm is present; motor reactions, as tapping, nodding, or swaying of body are noted frequently; the feeling of pleasure accompanies rhythmic reactions. While there are no data as to accuracy, there is suggestive matter bearing on the optimal rate and on the relations of compound and simple movements of the hand.

As far as the writer knows, he is the first to present systematic results as to the head and foot movement. The purposes of this enquiry may be briefly stated as

(1) the collection of a large body of facts, bearing on the actual and relative accuracy of straight-line movements possible with various parts of the body, such as hands, arms, head, legs, and feet;

(Something like 340,000 lines have been drawn and calculated.)

(2) to introduce certain variations in the conditions attending the production of ruled lines, such as

(a) to rule with the eyes opened and eyes closed, with other conditions the same;

(b) to change the rate of ruling or interval between the production of ruled lines; the rates chosen were 20, 30, 40, 50, 60, 70, 80, 100, 120, 140, 160, 180, and 200 beats per minute;

(c) to change the length of the normal or first line; the lengths used were 14, 10, and 1 cm.;

(d) to impose a weight on the ruling hand to either retard or accelerate the movement, choosing a weight of such magnitude that it would be perceptible, but would not have mass enough to cause pain or fatigue; 260 grams was used;

(e) to introduce a simultaneous movement of the free hand; i. e., the one that did not carry the recording pencil, of a similar character and extent but of opposite direction to the ruling hand;

(f) to record movements of both hands, of the head and of both feet;

(g) to conduct a series of experiments of similar character, as regards time-rate and extent of movement, to the series presented by Dr. Woodworth, with the idea of corroborating or disproving the results of his investigations; lines of 140 cm. were accordingly chosen;

(h) to conduct a series of experiments where the subject chooses his own rhythm or rate at which the easiest and best lines, subjectively speaking, could be ruled;

(i) to find the rates of respiration and pulse-beats and find the connection, if any, between them and the linear records.

(3) To examine, by variations of the number of lines ruled, the questions of fatigue and persistence of the memory-image; series of 50 lines for the first year and of 20 lines for the second year, were accordingly selected.

(4) To find the relations, if any, between constant errors and mean variations, so called.

THE APPARATUS

It is proposed to give the briefest possible discussion or explanation of the apparatus required for the investigation, it being desired at a later stage to enter into a comparison of the method adopted here with that of the only other investigation at all comparable to this one: the research problem of Dr. Woodworth, already referred to.

The underlying principle has been to avoid complication in apparatus, partly because of the delay and expense involved in working out, and making up elaborate schemes for apparatus, but mainly because of the advantage in duplicating this series of experiments, or of carrying on related investigations, to be derived from a choice of such parts, entering into the complete apparatus, as are at hand in any psychological laboratory, or that can be obtained and set up at small expense.

The use of smoked paper has been avoided, because a short preliminary series, using the usual smoked-paper records, was found to give no better results than did the method here adopted of ruling on white paper with a soft pencil, and the labor was thus considerably reduced.

To the objection that the pencil-ruling is more difficult, and involves more loss in friction and more complicated adjustments on the part of the subjects, only one of fourteen subjects admits that this is the case; and even if the testimony was unanimous as to the greater ease of production of the smoked records, it would be no reason for its adoption, since one of the first rules for all experimental work is uniformity of conditions, and this is equally well attained in either case.

The apparatus for free hand-movements and for the compound movements of both hands consists:

(1) Of an adjustable wooden rest (see Fig. A) with a base (a) about 40 × 60 cm. hinged to a vertically adjustable flat board (b), called the arm-rest, about 40 × 70 cm., and having on its upper edge two brass pins or plates (c) about 30 cm. apart.

The pencil is started from one of these pins, depending on the hand used, and moved until it comes in contact with a wooden rod that is held against the opposite pin and which is of the right length to give a movement of the pencil of 1, 10, or 14 cm., as desired.

The operator holds this rod in place for the first line ruled and then instantly removes it, so that the second and all later lines are ruled by memory of the first one, as closely in length to the first, or so-called normal line, as is possible.

(2) The apparatus for actuating and taking care of the paper.

This consists of two drums (d and d´, Fig. B) 20 cm. diameter by 40 cm. wide, mounted on suitable supports about 1 metre apart, and fastened to a table, with axes parallel.

The drum upon which the record is to be made (d) is adjusted close to the arm-rest, so that each ruled line will be carried down and out of sight before the next one is ruled, the pencil being held in the position (e); note that the arrow shows the direction of rotation.

The second drum (d´) is actuated by a motor (F) through a round belt (g), this motor being a clockwork type, with gear-changes and adjustable vanes for varying the speed, and having the power derived from a suspended weight (w).

The recording paper (h) transmits motion from (d´) to (d). This paper consists of a strip about six metres long by twenty-eight cm. wide, with one end pasted to (d), and then wound upon (d), leaving enough to be carried to (d´) and pasted to the latter. As the paper is unwound from (d), it is wound upon (d´), and, both to keep the paper tight and to prevent too rapid unwinding of (d), it is necessary to apply a friction-brake to the shaft of (d).

(3) A metronome, capable of being used for a range of 20 to 200 beats, and a stop-watch, to enable the operator correctly to time the subject, are in constant use.

The metronome is set in vibration and the subject is permitted to take his own time to start the ruling, the operator holding the wooden rod in place with one hand, while the other hand holds the stop-watch ready to start it the instant the subject's pencil is moved. There is thus a personal equation for the length of period, but this is of no consequence, as will be apparent when the method of calculation and the use of the planimeter is considered.

In the series of records with the weight, it is impossible to run the speed about 80 to 100 beats, unless the modification in apparatus shown in Fig. C is used; for the vibration of the string running from the hand to the weight around a pulley is violent enough either to throw the string off the pulley or cause the weight to jump so severely as to render the records useless.

This is entirely obviated by the given method of using a heavy weight acting with a small leverage (about 1 cm.) and thus moving only a short distance, so that it is capable of operating at the highest speeds with no perceptible shock or jump; the string is led to the hand or wrist from a grooved pulley of about 12 cm. radius, so the highest velocity of the weight is only about one twelfth that of the hand. This method makes it possible to carry the weighted records to the highest speeds.

This same method is used for the head and foot records, with the following additional apparatus; the string (Fig. C), shown leading to the hand, is led horizontally over to and around a similar large pulley on the opposite side of the table and either down to the foot or in a diagonally upward direction to the head; so that movements of the head or foot are faithfully recorded on the drum by means of a pencil held in a block of wood, this block of wood being fastened on the horizontal string in a suitable position for recording on the drum paper. The pencil is kept against the paper by a light spring or elastic band.

The foot is connected to the string by a stirrup that prevents any movement of the feet at all, unless the same is recorded by the pencil.

The head is furnished with a skull cap or harness consisting of non-elastic webbing and stiffened, where the string is attached, by a strip of sheet brass formed to fit the forehead or the back of the head, as the case may be. The object of the brass strip is to prevent a lost motion in the flexible webbing, that is found troublesome otherwise.

It will be evident, then, that the weight is continually acting as an accelerating or retarding influence in all records for head and feet, but it is not considered objectionable, for it is a constant throughout the series.

The other plan would require a circuit of cord leading in both directions from the head or feet in a complete circuit, and would cause in the opinion of the writer too much complication of apparatus.

The pulse-beats were taken by the stop-watch and wrist method so familiar to the physician, while the respiration results were obtained by the usual tambour apparatus for registering the chest expansion upon smoked paper.

THE METHOD OF CALCULATION

Suppose that the drums have been set in rotation and that the paper is unwinding from (d) and being wound on (d´), Fig. B, and suppose that the subject has ruled series of 20 to 50 lines, as may be desired, regulated by the stop-watch in the hands of the operator. The records will appear much as Fig. 5 under the planimeter discussion, there being for each speed one normal line to start and a series of lines following and intended to be of the same length as the normal line. A series of records, then, consists of 13 records of 20 or 50 lines, each running from 20 to 200 beats per minute, the complete series having not less than 260 and not more than 650 lines.

It should be added that the operator holds a pencil-point on the end of each normal line just after the record of 20 or 50 lines is made and turns the drum (d), thus marking a line nearly perpendicular to the ruled lines and at the average or normal distance from the starting-point; an absolutely correct record would show all ruled lines ending on this line.

The calculation of this series of records by the ordinary method of measuring each line, adding the lines of the series, averaging for the constant error, and repeating the operation in a slightly different form for the mean error or mean variation is of such enormous labor for an extended investigation as to be beyond the capacity of one or of several students; it is fortunate that the planimeter is at hand to be employed in averaging each series, and this instrument has therefore been selected as overcoming this difficulty.

It is desirable to consider the method employed by Dr. Woodworth to overcome this danger of excessive computation, and it will now be subjected to a critical and comparative examination.

He says, page 19 of his monograph on the Accuracy of Voluntary Movement, that the subject's sole duty was to make the present line equal to that immediately preceding, and the width of the slot was so adjusted that the subject could see only the line just ruled. After discussing certain matters of memory and its relation to the memory-image, in the attempt to support this changing normal plan, he confesses, on page 20, that this device is advantageous in much simplifying the most tedious part of the graphic method, that of computation.

While this is undoubtedly true, it needs careful scrutiny before adoption, for, on the same page, he says that one source of error in the method of making each line equal to the preceding one is that the different movements in the same series are not comparable, but the positive constant error is cumulative in its effect, and the normal tends to become longer and longer.

Some relation between this source of error and such a record as shown on page 29, Fig. 2, is evident, for, while it should be noted that this cumulative effect is peculiar to a series of lines for one speed, it has further a tendency to produce overruling at all speeds, and the natural result is to increase the error unduly and unnaturally for the higher speeds or as the speed increases, because there is then less time for the discrimination and choice that will tend to shorten the ruled line. It may be predicted, then, that Dr. Woodworth's method will show a slight lengthening of normal between lines at slow speeds and a much greater one at high speeds, the effect being to introduce a variable factor that would have no existence were a better plan adopted. The computation required for the average error is simple, being dependent only on the first and last lines of a series, and it is suspected that this very simplicity has led to its adoption and the consequent neglect of certain serious sources of error.

He tells us, on page 20, that the constant and variable error may well be isolated and studied separately, but indicates that they must "somehow" be considered combined as nature has made them; that is, analysis is desirable, but the synthetic method is more scientific.

This investigation will present data suggesting that

(1) Such a curve as that on page 29 of his monograph is not a characteristic one and relations of length of ruled line, as well as effects of weight, make it impossible to apply Weber's law or even the law of Fullerton and Cattell in the way proposed by Dr. Woodworth.

(2) There is no relation, mathematical or other, between constant and mean errors, and they not only may be but must be isolated and studied separately, if an investigation is to be conducted in the interests of scientific exactness.

It will be necessary to reject the method of Dr. Woodworth if the most reliable results are desired, in which case the planimeter is a necessity.

The theory of the planimeter cannot be developed at this place; every physicist and engineer is acquainted with it. The writer believes he was the first to apply the planimeter to the calculation of results from psycho-physical data for averaging both mean and variable errors. More than 340,000 lines were involved, each demanding two measurements. The best type of planimeter for general use and the one used here is the Amsler adjustable-arm form.

In Fig. D is shown a record taken at twenty beats per minute that will both explain the method of computation and show how the planimeter has been used to find the constant and mean errors.

Hylan-20 beats. L.H.E.c.-2-13-Ό1.]

The record, as made and ready for computation, is not provided with the line cd or with the dotted lines that connect the ends of the ruled lines. The line ab is drawn by turning the drum of the apparatus with a pencil held at the end of the normal or left-hand line af, which was here 100 mm. long.

The tracing-point of the planimeter being placed at a, a reading is taken, which was in this case 1486; after following with the tracing-point the dotted path to g and returning, via gb and ba, a second reading is taken, which was 1248; subtracting gives 238, which should be read 2380 square mm. for the area of the space agba; dividing by the distance ab, in this case 119 mm., gives the average height, which is + 20.0 mm., the plus sign suggesting that the distance thus found, which is the constant error for the series, be laid off in addition to or beyond Fa.

This being done, a line cd is drawn parallel to and 20.0 mm. from ab, as the mean line of constant errors.

To find the mean error of the series a slightly different method is necessary.

Place the tracing-point of the planimeter at c and read vernier, giving 1916; follow the dotted path from c to h, the straight line from h to i, the dotted path from i to k, the straight line from k to l, the dotted path from l to m, the straight lines from m to n and n to g, the dotted path from g to m, the straight line from m to l, the dotted path from l to k, the straight line from k to i, the dotted path from i to n, and the straight line from h to c, when a second reading is taken, which was in this case, 1806. Divide the difference of these two readings, 1100 mm., by the length of cd, 119 mm., and the result is 9.1 mm., or the mean error (mean variation).

It will be noted that this method gives the sum of the errors from the mean line cd; that is, the same result would be obtained if the tracing-point were (1) carried from c around all the area below cd, and this area were calculated as before; (2) carried from c around all the area above cd and the area measured as in other cases; and (3) these two results added and averaged.

To apply the method for ab, or constant error computation, to cd should give equal readings at c or a 0 mean error, a result evidently incorrect in the record selected.

After averaging results by the planimeter, the collection of data has been arranged by months; the record for one month only can be presented here, but the method of tabulation is the same throughout.

Each figure given for N, M, c and v, in the accompanying typical table for the month of May, 1904 (pages 495-499), is the average from 20 or 50 lines, ruled as already shown, Fig. D.

RESULTS

It is necessary to observe that the limits of space imposed on the writer preclude all but the barest outline of the deductions to be drawn from the investigation, and to this fact is due whatever of dogmatism is inherent in the argument; for it is manifestly impossible to present all the material, and the writer asks, then, the indulgence of the reader when he claims to have impartially examined and presented the evidence.

HAND MOVEMENTS

Simple movements Lines 14 cm. long.

TYPICAL SERIES FOR THE MONTH

Key. v = mean error. R.H. = right hand. R.F. = right foot. E.O. = eyes open. si. = simple motion N = normal line. Unit = 1 mm. L.H. = left hand. L.F. = left foot. E.C. = eyes closed. co. = compound motion. M = mean line. b = beats per minute. c = constant error.

See Beats per minute. Day. Subject. Key. 20 30 40 50 60 70 80 100 120

6 Hylan. N 10 10.5 11 11 10 10 10 12 10 L.F.E.O. M 16.1 13.5 12.1 13.8 13.1 10.0 10.0 11.2 12.5 c +6.1 +3.0 +1.1 +2.8 +3.1 0.0 0.0 -0.8 +2.5 v 2.8 3.5 1.9 0.9 1.6 4.6 2.0 2.7 1.2

140 160 180 200

12 10 11 11 12.0 6.1 11.0 14.21 0.0 -3.9 0.0 +3.2 1.0 1.3 2.1 4.0

Hylan. N 10 11 10 12 10 12 11 11 11 L.F.E.C. M 8.9 10.5 10.9 11.4 13.1 10.3 11.8 13.1 13.9 c -1.1 -0.5 +0.9 -0.6 +3.1 -1.3 +0.8 +2.1 +2.9 v 2.7 6.4 2.2 1.4 2.7 1.6 2.3 1.6 4.9

11 11.5 11 11 15.8 11.0 15.0 17.8 +4.8 -0.5 +4.0 +6.8 2.9 2.3 2.3 2.5

George. N 10.5 10 10 10 10 10 10 10 11 L.F.E.O. M 14.6 9.7 7.1 6.4 7.7 8.3 8.0 10.0 11.0 c +4.1 -0.3 -2.9 -3.6 -2.3 -1.7 -2.O 0.0 0.0 v 2.1 1.6 1.8 2.3 0.8 0.7 1.4 1.7 0.8

10 10 11 11 9.0 10.0 8.0 12.5 -1.0 0.0 -2.0 +1.5 1.0 1.3 1.6 0.4

George. N 10 9 9 10 11 9.5 9 10 8 L.F.E.C. M 11.7 9.2 10.2 12.6 13.2 11.5 12.2 6.6 5.0 c +1.7 +0.2 +1.2 +2.6 +2.2 +2.0 +3.2 -3.4 -3.0 v 2.0 2.9 2.5 0.7 0.5 1.0 0.8 2.0 4.1

10 9 9 10 8.0 7.6 8.0 11.0 -2.0 -1.4 -1.0 +1.0 3.1 2.5 2.1 3.7

Moore. N 10 10 10 10 11 11 11 10 10 L.F.E.O. M 15.7 18.8 17.7 16.7 18.3 18.5 16.5 15.6 16.0 c +5.7 +8.8 +7.7 +6.7 +7.3 +7.5 +5.5 +5.6 +6.0 v 3.5 3.8 1.1 0.8 3.6 2.1 2.7 0.6 2.4

10 11 10 10 16.2 16.3 16.4 18.8 +6.2 +5.3 +6.4 +3.8 0.5 3.5 2.2 3.8

N 10 10 10 10 11 10 10 9 Moore. M 19.6 15.3 15.3 14.3 14.9 13.5 6.7 13.4 L.F.E.C. c +9.6 +5.3 +5.3 +4.3 +3.9 +3.5 -3.3 +4.4 v 2.6 2.5 2.4 0.4 2.0 6.4 0.7 4.1

9.5 11 11 10 10.5 20.0 14.3 14.5 16.9 17.9 +10.5 +3.3 +3.5 +6.9 +7.4 2.6 3.8 1.1 3.1 2.8

N 10 10 9 10 10 9.5 10 10 9 Angier. M 13.6 12.5 11.6 11.6 11.5 11.7 13.0 12.9 R.F.E.O. c +3.6 +2.5 +2.6 +1.6 +1.5 +2.2 +3.0 +2.9 v 2.9 2.3 1.8 2.3 1.7 1.6 2.0 2.7

11 10 10 10 10 12.2 12.2 13.5 22.1 16.5 +1.2 +2.2 +3.5 +12.1 +6.5 3.3 1.5 1.5 6.7 1.5

N 10 10 10 9.5 10 9 10 10 Angier M 11.4 7.7 13.8 8.3 11.4 11.5 11.0 10.3 R.F.E.C. c +1.4 -2.3 +3.8 -1.2 +1.4 +2.5 +1.0 +0.3 v 4.8 1.3 3.0 2.6 1.4 1.7 1.8 1.7

140 beats

9 10 10 11 10 10 9.9 12.8 16.7 17.7 12.4 12.0 +0.9 +2.8 +6.7 +6.7 +2.4 +2.0 1.8 1.4 1.7 4.2 2.1 1.7

N 11 10 10 9 10 11 10 10 Huggins. M 12.5 8.5 12.6 9.7 9.7 16.6 15.7 18.7 R.F.E.O. c +1.5 -1.5 +2.6 +0.7 -0.3 +5.6 +5.7 +8.7 v 3.6 2.6 2.7 2.8 3.3 3.2 3.3 3.6

10 11 11 11 10 13.6 15.8 9.3 18.3 14.3 +3.6 +4.8 -0.7 +7.3 +4.3 3.1 2.7 2.9 3.8 3.0

N 11 8 10 10 11 10 10 10 Huggins. M 6.5 10.5 10.7 8.8 12.6 12.9 12.2 22.3 R.F.E.C. c -4.5 +2.5 +0.7 -1.2 +1.6 +2.9 +2.2 +12.3 v 1.4 2.1 1.8 1.6 3.7 3.1 2.0 4.1

10.5 10 10 10 10 9.9 21.8 12.0 12.5 16.3 -0.6 +11.8 +2.0 +2.5 +6.3 1.4 5.0 2.3 3.6 1.8

N 9 9 9 9 9.5 8 7 10 13 Lenfest. M 12.7 11.9 11.1 11.3 12.1 8.7 11.6 12.6 R.F.E.O. c +3.7 +2.9 +2.1 +2.3 +2.6 +0.7 +4.6 +2.6 v 3.7 2.4 1.9 2.3 2.8 0.2 2.8 2.1

10 10 9 8 10 9.0 8.6 11.7 7.0 11.8 -1.0 -1.4 +2.7 -1.0 +1.8 2.0 2.1 3.2 1.4 4.1

N 12 11 11 10 10 10.5 11 11 Lenfest. M 14.9 14.7 12.2 13.5 12.6 11.9 12.6 8.7 R.F.E.C. c +2.9 +3.7 +1.2 +3.5 +2.6 +1.4 +1.6 -2.3 v 3.3 2.2 3.5 3.5 3.2 3.0 2.6 1.7

11 11 10 10 10 9.1 8.9 8.4 14.1 6.9 -1.9 -3.1 -1.6 +4.1 -3.1 2.4 2.8 2.9 4.1 2.3

N 12 11 11 11 11 10 11 10 George. M 8.3 9.0 8.6 8.8 8.7 7.0 7.4 7.9 R.F.E.O. c -3.7 -2.0 -2.4 -2.2 -3.3 -3.0 -3.6 -2.1 v 5.0 3.3 2.3 0.1 3.3 2.9 2.4 3.1

11 10.5 11 10 10 6.8 9.7 9.0 5.8 11.0 -4.2 -0.8 -2.0 -4.2 +1.0 1.5 4.6 2.9 2.9 3.9

N 12 11 12 11 11 11 10.5 11 George. M 7.0 7.7 14.1 8.3 10.3 8.4 8.7 7.8 R.F.E.C. c -5.0 -3.3 +2.1 -2.7 -0.7 -2.6 -1.8 -3.2 v 2.5 2.6 2.4 3.7 2.9 2.3 1.6 2.1

10 10.5 11 10 10 6.4 10.0 9.0 7.9 7.3 -3.6 -0.5 -2.0 -2.1 -2.7 1.9 3.7 2.9 1.8 3.8

N 11 11 11 11 12 10 9 11 Moore. M 14.5 16.6 16.5 9.0 17.0 10.7 10.4 11.9 R.F.E.O. c +3.5 +5.6 +5.5 -2.0 +5.0 +0.7 +1.4 +0.9 v 1.7 2.2 1.9 1.8 1.3 2.2 2.6 2.2

10 10.0 10.5 11 9.5 11.5 12.5 12.5 15.1 12.6 +1.5 +2.5 +2.0 +4.1 +3.1 2.2 2.5 2.5 1.4 0.8

N 11 10 11 11 11 10 10 9 Moore. M 14.3 13.2 15.9 12.3 17.9 10.0 11.3 15.9 R.F.E.C. c +3.3 +3.2 +4.9 +1.3 +6.9 0.0 +1.3 +6.9 v 1.8 2.8 1.6 2.1 1.7 2.4 3.2 2.2

11 11.0 10 10 10 14.8 15.5 13.4 11.8 12.2 +3.8 +4.5 +3.4 +1.8 +2.2 2.7 1.4 2.2 1.8 2.2

N 11 11 10 11 10 10 10 11 16 Angier. M 14.3 13.5 9.3 14.4 14.1 9.5 15.1 12.7 R.F.E.O. c +3.3 +2.5 -0.7 +3.4 +4.1 -0.5 +5.1 +1.7 v 1.7 1.9 2.0 2.4 2.0 1.1 2.4 2.8

10 11 10 11 11 11.1 12.6 13.7 11.0 16.6 +1.1 +1.6 +3.7 0.0 +5.6 2.0 2.0 2.7 1.3 3.5

N 11 10 10 9 10 10.5 10 10 Angier. M 14.4 4.8 2.5 3.6 5.8 6.7 8.0 11.0 R.F.E.C. c +3.4 -5.2 -7.5 -5.4 -4.2 -3.8 -2.0 +1.0 v 3.2 2.7 2.3 3.6 2.7 2.3 1.3 2.8

11 10 11 11 11 13.5 11.1 13.9 10.1 11.0 +2.5 +1.1 +2.9 -0.9 0.0 2.8 1.4 2.4 1.8 2.2

N 97 97 98 97 98 99 97 98 Huggins. M 107.6 107.8 100.4 114.2 98.0 110.3 89.4 102.5 L.F.E.O. c +10.6 +10.8 +2.4 +17.2 0.0 +11.3 -7.6 +4.5 v 5.9 0.6 4.8 6.5 5.3 6.7 6.4 5.5

95 98 99 99 100 99.2 100.8 106.4 101.5 108.9 +4.2 +2.8 +7.4 +2.5 +8.9 7.8 6.0 4.7 7.0 12.7

N 97 98 100 97 95 95 99 99 Huggins. M 102.8 104.3 115.1 101.4 94.1 102.0 106.4 93.4 L.F.E.C. c +5.8 +6.3 +15.1 +4.4 -0.9 +7.0 +7.4 -5.6 v 5.8 8.6 10.8 9.9 7.9 10.4 8.1 6.8

97 98 99 99 100 87.7 98.7 108.4 97.7 110.8 -9.3 +0.7 +9.4 -1.3 +10.8 9.4 5.8 6.9 5.3 6.1

N 11 11 10 10 10 10 10 10 20 Lenfest. M 16.1 13.5 11.6 10.9 10.9 14.2 11.7 10.5 L.F.E.O. c +5.1 +2.5 +1.6 +0.9 +0.9 +4.2 +1.7 +0.5 v 2.8 3.0 3.0 3.0 1.2 1.9 1.0 1.6

11 10 11 11 10 12.4 11.1 12.2 8.7 12.9 +1.4 +1.1 +1.2 -2.3 +2.9 3.2 5.1 1.9 1.7 4.2

N 12 11 10 10 10 11 10 10 Lenfest. M 20.4 16.8 11.6 11.2 11.8 14.0 9.0 9.3 L.F.E.C. c +8.4 +5.8 +1.6 +1.2 +1.8 +3.0 -1.0 -0.7 v 2.5 3.6 3.0 1.9 1.6 2.0 3.0 1.2

10 10 11 11 10 6.2 10.0 8.5 5.1 8.3 -3.8 0.0 -1.5 -5.9 -1.7 1.9 0.3 1.8 1.7 0.8

N 98 97 94 98 98 97 98 98 George. M 104.0 100.4 102.6 99.3 105.1 111.1 101.9 100.5 L.F.E.O. c +6.0 +3.4 +8.6 +1.3 +7.1 +14.1 +3.9 +2.5 v 6.4 11.3 9.6 7.7 5.3 8.0 3.7 5.6

97 96 98 98 97 96.4 102.0 97.4 94.2 95.8 -0.6 +6.0 -0.6 -3.8 -1.2 5.7 5.7 9.1 7.8 3.6

N 98 97 94 97 97 98 98 97 George. M 93.6 81.2 94.7 92.7 104.2 99.3 93.4 89.6 L.F.E.C. c -4.4 -15.8 +0.7 -4.3 +7.2 +1.3 -4.6 -7.4 v 9.1 8.9 5.5 4.6 5.8 6.9 4.9 6.6

96 98 98 99 100 96.4 93.0 87.3 101.8 87.4 +0.4 -5.0 -9.7 +2.8 -12.6 7.8 4.8 7.7 5.6 5.7

N 97 99 98 97 96 96 97 98 Moore. M 106.1 106.9 105.2 103.0 104.7 108.6 102.7 106.9 L.F.E.O. c +9.1 +7.9 +7.2 +6.0 +8.7 +12.6 +5.7 +8.9 v 4.3 5.6 4.6 5.2 8.0 6.0 7.4 6.9

99 98 99 99 99 110.9 105.7 111.2 99.9 93.0 +11.9 +7.7 +12.2 +0.9 -6.0 4.8 9.0 4.7 11.4 1.8

N 96 97 97 97 96 97 96 97 Moore. M 119.2 79.8 87.3 79.6 81.6 91.2 96.5 106.2 L.F.E.C. c +23.2 -17.2 -9.7 -17.4 -14.4 -5.8 +0.5 +9.2 v 16.1 6.7 11.6 7.2 7.5 4.1 4.0 6.6

99 98 97 98 99 100.4 91.9 104.0 88.6 80.9 +1.4 -6.1 +7.0 -9.4 -18.1 4.0 6.1 5.4 6.6 8.2

N 98 97 95 96 96 97 99 97 23 Lenfest. M 105.7 110.8 97.4 93.8 97.8 100.2 99.9 92.9 L.F.E.O. c +7.7 +13.8 +2.4 -2.2 +1.8 +3.2 +0.9 -4.1 v 9.3 1.5 3.6 5.1 3.6 4.4 2.9 3.3

97 98 98 98 99 86.1 96.6 91.9 80.9 83.0 -10.9 -1.4 -6.1 -17.1 -16.0 6.7 5.0 8.2 6.0 7.8

N 96 94 96 96 96 94 96 97 Lenfest. M 111.3 95.0 98.3 101.7 101.7 110.1 98.8 79.0 L.F.E.C. c +15.3 +1.0 +2.3 +5.7 +5.7 +16.1 +2.8 -80.0 v 7.1 5.1 5.8 2.7 5.4 7.8 12.1 3.7

97 97 100 100 100 88.9 72.7 88.4 69.7 80.8 -8.1 -24.3 -11.6 -30.3 -19.2 5.6 8.0 7.4 8.4 7.0

N 96 97 96 96 98 98 97 97 Huggins. M 113.5 105.7 105.5 97.7 92.5 98.4 92.5 102.0 L.H.E.O. c +17.5 +8.7 +9.5 +1.7 -5.5 +0.4 -4.5 +5.0 v 4.2 2.4 3.6 5.0 3.5 3.1 4.2 5.8

95 98 98 99 97 95.0 112.3 104.5 108.7 95.7 0.0 +14.3 +6.5 +9.7 -1.3 4.7 6.4 5.8 7.1 4.4

N 98 97 97 97 97 99 99 97 Huggins. M 103.5 82.8 86.1 87.8 87.7 94.2 92.1 94.5 L.H.E.C. c +5.5 -14.2 -10.9 -9.2 -9.3 -4.8 -6.9 -2.5 v 8.4 6.7 5.4 3.5 6.7 5.4 6.0 5.8

97 97 98 98 96 99.9 111.5 101.3 113.7 95.1 +2.9 +14.5 +3.3 +15.7 -0.9 4.8 13.3 4.7 9.6 2.6

N 96 95 96 96 96 97 96 27 Lenfest. M 101.6 93.4 91.2 90.0 97.5 93.2 93.4 R.F.E.O. c +5.6 -1.6 -4.8 -6.0 +1.5 -3.8 -2.6 v 6.2 5.1 3.4 2.8 5.5 4.5 3.7

96 99 102 101 100 89.8 97.5 88.6 96.8 66.2 -6.2 -1.5 -13.4 -4.2 -33.8 5.5 3.6 4.3 4.9 8.1

N 96 96 97 97 97 98 98 Lenfest. M 103.8 101.2 94.0 100.3 96.4 101.2 105.0 R.F.E.C. c +7.8 +5.2 -3.0 +3.3 -2.6 +3.2 +7.0 v 6.2 7.4 4.5 3.8 3.5 4.7 2.4

95 98 100 99 100 78.0 98.5 88.8 85.1 65.8 -17.0 +0.5 -11.2 -13.9 -34.2 4.7 4.3 7.0 6.9 5.9

The records are averaged for nine subjects, three of them being left-handed. For the right hand we find, for mean error, a reduced error with visual control.

For constant errors, a similar result is apparent; when following the eyes-closed curve one may note a large negative error 20-50 beats, and a similar but larger positive error 70-160 beats, with a falling to a negative error again at 200 beats.

This may be interpreted to mean a groping for the correct length of line at the lower speeds when some time for reflective processes is allowed, and an inhibitory effect on the motor discharge; later the speed prevents this discrimination, and introspective testimony goes to show that a mental conception of a barrier, beyond which one cannot carry the pencil, is set up and kept more or less constant through the help of the joint and muscular sensations. It would follow, then, that this muscular stop is overestimated where reflection is not possible.

Finally, the falling-off of the length of the line is probably due to physical inability to rule a line of the full length of 140 mm. at 200 beats per minute and an examination of some individual cases confirms this opinion, for the lines may be started some distance away from the origin apparently in order to end them at the correct point.

Curve inclinations are upward, for mean errors, with visual control, while the eyes-closed records show no increase in error for the increased speeds.

For constant errors, with visual control, there is a similar inclination downward for both hands, with a 0 error at about 120 beats. It should be noted that this opposite tendency in mean and constant errors suggests that they should be kept separate in all computation.

The left-handed subjects have much better control of their left hand than have the right-handed subjects, and they may dispense with visual control to a large extent.

On the other hand, for right-hand records we find much the same increase in irregularity and error for both left- and right-handed subjects; they all must depend on visual control for reduction of errors.

It follows that the non-visual control exerted by the left-handed subjects on the right hand is as good or as great as for the right-handed subjects; while they have the hand in which they may be expected to excel under much better control.

It is not intended to present this as an argument for teaching left-handedness, but it is certainly suggestive when considering the question that ambidexterity be taught in early life.

It should be noted that two of the three left-handed subjects might be expected, because of special training, to show marked manual dexterity, while only one of the four right-handed subjects has had special training along this line.

No extended discussion is appropriate here as to the question of what portion of this extra ability of the left-handed subjects to react accurately is due to practice and habit, i. e., is automatic, and accomplished without reference to the sensory motor by-path to the cerebral cortex; and on the other hand, as to whether the direct sensory motor path via spinal cord or medulla is not cut off entirely.

For 140 mm. averages and free motion, we find in general

(1) a reduced error and greater uniformity of result at all speeds where visual control is added, in the case of both mean and constant errors and for all subjects;

(2) the mean errors for visual-control records show a rise along a line whose equation is approximately y = px, or the equation of a straight line, where p is an undetermined constant.

On the other hand,

(3) the eyes-closed mean errors show no increase or decrease in value during the entire series;

(4) the constant errors for visual-control records show a drop from positive errors to negative errors, along a line whose equation is approximately y = qx or the equation of a straight line, where q is an unknown constant, somewhat less in value than p in the case of mean errors; the constant error becomes 0 at about 120 beats;

(5) the eyes-closed constant errors follow the same equation for left-handed subjects, using the left hand, but all other cases suggest a curve of the parabolic form, having 0 constant errors at 60 and 180 beats and being convex upward.

Considering individual records for 14 cm.

A general survey of the charts suggests certain irregularities that call for explanation, for there will be sudden large increases in errors, that are explicable on the hypothesis that the subject has temporarily lost control of the moving hand, that is, that fatigue is to be noted.

While the purpose of the investigation has been to allow no lines to be ruled while the subject was conscious of any such feeling, there being a pause of any desired length to permit time for rest, it is to be noted that a considerable amount of recorded data as to fatigue shows that it is an unconscious or subconscious phenomena.

Further, the series of records have been arranged to occur from 20 to 200 beats and never in the reverse order, because of subjective limitations, so it is reasonable to expect that during the period of twenty minutes to one half an hour required for a series of records, there will be lapses of volitional control entirely beyond the ken of the subjects. It is to this cause rather than to pure chance that the results will be attributed. With this exception, the individual records show close agreement with their average.

The results obtained from a consideration of free hand-movements of 1, 10, and 14 cm. length are:

For 14 cm. lines

for the average of nine subjects:

The mean errors,

(1) increase with speed for eyes open;

(2) do not change in error with speed-change for eyes closed;

(3) visual control reduces errors for right hand, but does not for the left;

(4) right-handed subjects alone gain from visual control.

The constant errors,

(1) decrease with speed in visual cases;

(2) increase with speed to the middle and then reduce to 200 beats for eyes closed;

(3) left-handed subjects are more accurate for the left hand and can dispense with visual control;

(4) all subjects need visual control for the right hand;

(5) left-handed subjects show less error throughout for non-visual.

For the individual cases,

the mean errors

(1) show evidences of loss of control or fatigue for some speeds, and the average results are confirmed.

The constant errors show that the average deductions are confirmed.

Lines 10 cm. long:

Averages for seven subjects as regards mean errors have especial interest for l.h.e.c. records, which alone show a rise of error with speed-increase.

Noting that the records are overwhelmingly averages of right-handed subjects (six to one), it is of interest to examine this record.

We may say, then, that for right-handed subjects, the voluntary control for the right hand is not much improved by the introduction of visual assistance; it is more marked for speeds of 100 beats or less than for the high speeds. And in the latter case, it is under 10%; but when the left hand is considered, a marked gain or 40 or 50% is apparent, when the eyes are used except for the two lowest speeds.

As far as it is possible to offer any hypothesis from the few facts tabulated, it may be said that right-handedness implies a high development of muscular control, but slightly improved by the introduction of the visual element, as far as the right hand is concerned; but for the left hand muscular control comparable to the right-hand control can be obtained only with visual control; in short, I fail to find evidences of cross-education, where the visual element is absent, nearer than about 50% of the mean error.

No clearly marked gain through visual control can be pointed to in the case of constant errors; there is, to be sure, a slight gain in steadiness and error-reduction, where eyes help in the case of both hands, but not 5% in magnitude of the difference noted with mean errors.

Individual records show fatigue-points at 40 to 80 beats and again above 140 beats, but there is no perceptible loss of control during a series of lines ruled at only one speed.

It is apparent, when comparing with the 140 mm. records, that there is no physiological reason why the subjects may not rule the full length of a 10 cm. line at 200 beats, and the limit of movement for high speeds is probably between 10 and 14 cm.

The constant errors are in general positive and only Me. shows a tendency to underrule lines at high speeds.

For 10 cm. lines,

for the average record,

(1) the mean-error curve is horizontal for l.h.e.o., but otherwise rises with speed-increase;

(2) visual control is a 10% gain for right hand and 40 to 50% for the left, so right-handedness is prominent for the eyes-closed series;

(3) the constant-error curve for r.h.e.c. rises, but all others show reduction of error as rate increases;

(4) visual-control gains are not over 5%.

The individual curves show,

for mean errors,

(1) marked-fatigue points for l.h.e.o;

(2) r.h.e.o. curve is horizontal, but all others rise;

(3) l.h.e.o. curve is about the same as r.h.e.o., but there is more loss for non-visual series with left hand.

For constant errors,

(1) r.h.e.c. curve rises, but all others are horizontal;

(2) the eyes reduce errors especially for the left hand;

(3) overruling is prominent even at high speeds, for there is no evidence that the lines are shortened at high speeds.

Lines 1 cm. long:

Averages are made for nine subjects, three being left-handed.

The eyes are effective in reducing mean errors and to a less extent for constant errors.

A noteworthy feature of the constant-error record is that the errors are positive with one exception, that of 20 beats with l.h.e.c. and even this curve jumps rapidly above the 0 line.

In all cases the motor discharge is of sufficient magnitude to cause overruling in cases of normal lines of one cm. The assistance afforded by the eyes is not marked.

Certain evidence of an introspective character, that most of the subjects offer, is to the effect that, "when I would do good, evil is present with me"; that, where there is a decided feeling that the muscular limit, if such a term be permitted, is exceeded, yet the subject's will-power is not sufficient to inhibit the overruling; there is a more or less vivid conscious error in the 10 mm. series for the hands.

In regard to the relation of mean and constant errors, there is more close uniformity than with the 140 mm. lines, but it is to be noted that there is no comparison to be drawn between maximum or minimum points; for example, at 100 beats the minimum points for r.h.e.o. agree closely, but the maximum constant error matches the minimum mean error at 100 beats for r.h.e.c.

We cannot predict, then, that a subject capable of closely ruling to the normal will be able also to rule each line of the same length as the rest of the series, or vice versa.

As in the case of mean errors in general note that subjects show a less constant error and more regularity for their more dexterous member; it is not true for the left hand that for left-handed subjects visual control is a hindrance for accurate work; otherwise the same gain, by use of the eyes, is to be noted for the rest of the records.

Individual records show close correspondence with the average of results, and the latter may be considered fairly representative.

Almost the whole series shows the constant error positive, the most consistent example being for J. with l.h.e.o.; this tendency to overrun the 1 cm. lines is consistently uniform and has been elsewhere commented on, so it may be left with the observation that the log shows that the subjects were frequently conscious of this overruling, but confessed inability to correct it.

Only in the case of Y. for the three left-handed subjects and for W. among the six right-handed men does the left hand show less mean error than the right hand, and all other cases show such an interweaving of curves as to render it difficult to perceive any advantage that the more dexterous hand possesses on the score of accuracy.

For constant errors:

For individual cases it is to be noted that for the left-handed subjects J. is better for the right hand, Le. is indifferent, and Y. prefers the left hand; while three of the six right-handed subjects prefer the left hand and one is indifferent; thus giving still further proof that a more dexterous hand is a fiction on the score of the right or left-handed theory, when accuracy of straight-line movement is to be considered.

For 1 cm. lines, for the average, note for the mean errors:

(1) visual control reduces errors;

(2) errors increase for eyes open, but decrease for eyes closed, as speed increases when considering right hand, but left-hand errors are constant;

(3) as left-handed subjects are better for r.h.e.o. than right-handed subjects, but not for eyes closed, it is suggested that visual control equalizes differences in the subject's less trained hand.

For constant errors:

(1) visual control reduces errors;

(2) curves are horizontal in all cases;

(3) all errors are positive, showing consistent overruling;

(4) as visual control of the left hand is a gain for right-handed but a hindrance for the left-handed subjects, the more practised hand is probably able to dispense with visual control, and depend largely on the muscular sense.

Mean and constant errors are not comparable. For the individual cases we find a corroboration of the above and for mean errors: more dexterous hand does not excel, and evidence against ambidexterity is conflicting; for constant errors: overruling is consciously done.

Constrained hand-movements for lines 14 and 1 cm. long and for the weight, both accelerating and retarding the movement, are to be next considered.

Constrained motions are of two general types as examined by the writer. Series of the records for the hands were taken at 140 mm. and 10 mm. bases with a weight hung on the finger or fingers of the hand under investigation; in one series the weight acted as a pull or accelerating effect on the ruled line and in the other series the weight was imposed as a retarding effect, tending to restrain the movement of the hand.

This weight was in all cases 260 grams, this weight being chosen as of sufficient amount to have a perceptible effect, but not large enough to cause feelings of pain or fatigue in any case.

The average for seven subjects, three being left-handed, is as follows:

In general, the mean errors for the right hand are less, and less variable as compared with the left hand. The left-hand records are very close to the corresponding right-hand curves, especially the portions of the eyes-closed records 20 to 120 beats.

This may be said for both mean and constant errors. In general, mean errors are reduced, and curves are more nearly straight lines when the weight is added; also the weight reduces constant errors, and gains more regular records at all speeds. It is to be noted as a point of unusual interest that there is no apparent shortening of the line ruled when the weight is hung on the hand, for the negative errors are less, not more when the weight is applied.

In general, then, the imposition of a weight that will be small enough not to cause pain or fatigue shows that both mean and constant errors are reduced; that the amount of error is less variable over the range of speeds used; that the records show no retarding effect, but that the subject is both able to move the hand just as far as without the weight, and do it with much greater accuracy.

Individual records for 14 cm. and weight-retarding show a marked reduction in both mean and constant errors, and a less marked gain in uniformity in every case. This tends to confirm the introspective opinion of W. subject that the imposition of a retarding weight tends to reduce errors of both classes and to cause greater steadiness.

It should be added that there is evidence of an occasional letting-go of voluntary control, so to speak, resulting in a large increase in mean error, as already pointed out, or a large increase in negative constant error, as shown on all individual records, and it would seem then that the matter of cortical control is more vital and indispensable for the restricted movements.

The effect of weight-retardation on visual records is to reduce the error and steady the ruling of the less dexterous hand to a much more marked degree than for the well-trained hand.

In the l.h.e.c. records the lack of corrective effect of visual control is marked, as in the case of free movements, but the dip in the curve at 30 to 70 is not noted in the free ruling and should be considered as a distinct shortening due to weight-retardation before discriminative processes have oriented the subject.

Without considering the accelerating weight-records in detail note that:

The effect of weights (less than that necessary to cause pain or fatigue), either tending to accelerate or retard motions, is to reduce both mean and constant errors and to render more uniform or more uniformly increasing or decreasing such errors, except in the case of l.h.e.c., where constant errors are greater positively with the weight-pulling and greater negatively with the weight-retarding, than for free motions; that is, the effect of the weight is natural, and shows no signs of inhibition in this particular case.

There is no such marked fluctuation in error for the pull-records as was noted for the weighted curves, and it is further noted that the individual pull-records are more bunched or consolidated about some mean than are the free-movement curves. This suggests that the accelerating weight is a decided help for accuracy and regularity, and it would seem to call for less voluntary control than for either of the other movements.

Further, as the effect of pulling weights is to equalize the accuracy of movement of the hands, the hypothesis is proposed that weights either accelerating or retarding the movements of the hand tend to equalize their accuracy or to promote ambidexterity as far as accuracy of straight-line mean errors is concerned.

L.h.e.c. rise for Ha., are horizontal for J., W., and Y., and slope downward for the other subjects, the net effect being a slight downward slope. The loss of accuracy and regularity when the visual sense is inhibited is to be noted in every case, it being especially marked for Bo., Li., and W.

As compared with the r.h.e.c., there is not sufficient evidence to lead to the conclusion that the right hand is a more accurate member than the left, but on the contrary the left-hand record for non-visual control is lower for both weighted series than is the right-hand curve. Contrasting this with the eyes-open records for free and weighted movements, the visually aided results show a greater accuracy and regularity for the right hand.

This leads to a proposition that the greater dexterity on the line of accuracy, of one hand, that is the right hand for right-handed subjects, and the left hand for left-handed subjects, is a matter of visual control and is in no sense due to the muscular sense or to automatic action, for without eyes we are ambidextrous as far as accuracy of linear movements is concerned; the proposition needs careful scrutiny in application to the general question, but is held to be correct within the range of experiments.

We are tempted to extend this matter somewhat in the following way, by saying that there is no evidence deducible from this research that there is hereditary preponderance of activity or accuracy of one hand or one leg (as shown later) over its mate, and the baby is brought into the world with an equal capacity of accuracy of both members.

It is, then, an evolutionary matter, not racial but individualistic, and right-handedness or left-handedness is largely a development after birth. Our system of education is responsible for the over-development of one hand, and such a case as that of Dr. Anderson of the Yale University Gymnasium, who in class demonstration cannot instantly tell which hand is being used to actuate the chalk at the blackboard, is the normal symmetrically developed man.

The school reform for ambidextrous training is radical enough, but seems a logical conclusion of the argument. Apologies are appended for driving the argument beyond the limits of the investigation, but it is hoped that the enquiry is at least suggestive.

For 14 cm. lines,

weight-retarding movements:

For the average of nine subjects:

The weight reduces errors and promotes regularity in the case of both mean and constant errors, nor does it tend to cause underruling, save in the case of left-handed subjects for l.h.e.c. records. There is a gain, in general, when the visual factor is introduced.

For mean errors,

(1) right-hand curves are horizontal, while the visual records show increasing error and l.h.e.c. a reduction of errors;

(2) the right hand gives slightly better results;

(3) note that l.h.e.c. record is equally good for free or weighted movements.

For constant errors,

(1) r.h.e.o. and l.h.e.o. curve downward, while both non-visual curves slope upward;

(2) the left hand seems equally efficient, as compared with the right hand.

For individual cases,

note (1) fatigue-spots are more numerous than for free movements, especially for the left hand;

(2) weight reduces both mean and constant errors and to a less extent even records.

For mean errors,

(1) visual control reduces errors;

(2) the weight tends to equalize the accuracy of the right and left hands.

For constant errors,

(1) there is no general testimony showing shortening of lines at high speeds;

(2) the less trained hand is more helped by the weight, especially for non-visual work.

The evidence for right- and left-handed subjects is inconclusive, and we cannot finally say that the more trained hand is capable of greater accuracy.

Weight-accelerating movements:

The average of seven subjects:

The accelerating weight reduces mean and constant errors, and improves regularity of curves, except for l.h.e.c. constant-error record. There is some evidence that a pull causes overruling, while a retarding weight causes underruling, but there are exceptions enough to warrant care in finally accepting this statement. Visual control with accelerating weight reduces error more than the weight acting alone.

For mean errors,

(1) weight reduces errors for r.h.e.o. and l.h.e.c. as compared with free-movement records, while the other two curves are inconclusive;

(2) visual sense helps in accurate ruling;

(3) non-visual records are not reduced, as a rule, from the results of free motion.

For constant errors,

(1) the accelerating weight tends to greater accuracy, with an exception for the-non-visual records.

No testimony of marked importance is to be noted in comparison of right-handed and left-handed subjects; the more trained hand shows greater accuracy in some cases, but fails to excel in others; so the data is inconclusive.

For individual cases we find:

(1) the acceleration records are more accurate and regular, and present fewer lapses than the free or retardation results, suggesting greater ease with weight assisting;

(2) visual control is prominent throughout, and evidence shows that this sense is the greatest factor in the predominance of the more trained hand; the non-visual records should and do show no marked difference in the hands;

(3) a weight tends to equalize accuracy of hands;

(4) the overruling effect of weight is over-corrected in some cases for constant error of low rates.

Constrained movements of 1 cm.:

The average is of seven subjects, three of them being left-handed:

With weight-retarding movement, there is no reduction of mean error with visual control of right hand, but there is with the left. Constant errors show little reduction for either hand with eyes open.

The facts would seem to warrant the hypothesis that, for the left hand, a movement uncontrolled visually, whether restricted by a weight or not, can be made with greater accuracy, when time is permitted for discriminative and reflective processes and visual-control results in about the same error whatever the speed, while the right-hand motions show no such evening effect of visual control with the weight-records or even reduction of error; the free movement, however, does show a reduction of error.

A general statement may be deduced that, for lines of 10 mm. in length, there is no difference in either mean or constant errors, when a weight is imposed to cause retardation, provided the weight is not large enough to cause pain or fatigue.

By separating the averages for right- and left-handed subjects, it may be further said that:

Visual control is not efficient to reduce the error and no particular gain in regularity can be noted. The left-handed subjects show, for the left hand, much better results without visual control as far as the free motion is concerned.

While somewhat contradictory, it may be stated that constant errors are reduced by the weight addition, and there is some evidence leading to the belief that the ruled line is shorter when the weight acts as a retarding influence.

INDIVIDUAL RECORDS

Considering lines 10 mm. long with a retarding weight:

A glance over the seven individual records shows some considerable increase in both constant- and mean-error irregularities, as compared with the free-motion curves, as well as in actual errors; there are distinct losses of volitional control for both classes of errors, especially at or near the ends of the series.

There are cases of very low mean error to be found on all records, where the value is 1/4 mm. or less, and, while the same phenomenon is found with free motion, it is more marked here and occurs more frequently; in most cases it seems as a drop from errors of larger values rather than a gradual matter, as if the subject realized the large error and exerted unusual volitional control to correct and produce a very accurate record, but found that the attention needed was beyond his will-power, as shown by the immediate lapse of accuracy.

There is an indirect confirmation of this view from the introspective testimony of the subjects.

The visual element steadies but does not reduce mean errors when weight is retarding.

The general shape of curve for eyes closed is downward 20-40 beats, and rising for the rest of the series; it is less regular, but more accurate than the visual results.

The fact that constant errors are mostly positive leads to a denial of any inhibitory effect of the retarding weight.

For 1 cm. lines,

weight-retarding movements:

For the average of seven subjects we find:

(1) visual control does not improve accuracy or regularity as in free movements;

(2) a retarding weight tends to make errors constant whatever the speed-rate;

(3) the testimony goes to show that the free-movement records are more accurate than the retardation ones.

Mean errors are:

(1) no more accurate and perhaps less regular, when the weight is imposed;

(2) right- and left-handed subjects are equally accurate.

Constant errors:

(1) the more dexterous hand is superior for coördinations requiring accuracy;

(2) ruled lines are slightly shortened in some cases;

(3) weight-records do not give more accurate results as compared with free movements.

For individual records we find:

(1) retarding weights increase errors and irregularity;

(2) fatigue-points are more marked and frequent than for free movements.

Mean errors,

(1) the visual factor is of some value, but the testimony is varied; right hand for increased regularity only, and left hand for greater accuracy only;

(2) curves are horizontal or reducing with speed-increase.

Constant errors,

(1) the more dexterous hand coördinates better;

(2) all errors are positive;

(3) visual control helps only for regularity;

(4) curves are horizontal or rising.

With weight-accelerating movements, the average record shows a sudden rise in mean error at both ends, not in evidence with free or retardation results.

In general it is to be noted:

(1) that the visual element is of no value for reducing the error, and of little value for promoting regularity;

(2) that the pull-records are closely comparable to the free-motion records, and the accelerating influence of the weight is imperceptible;

(3) that the pull-records are more regular and closer to the free-motion curves than are the weighted records, especially at the ends of the left-hand curves.

For constant errors:

It is more in accord with the facts to say that the imposition of a weight tends to reduce the constant error, and this is more marked when the weight acts in pulling or to accelerate the motion.

Comparing with the weighted curve, we find the same general type of rising curve, similarly located, and the same is true when compared with the free-motion curve. Constant errors are reduced, but slightly, and visual control is rendered nil, when the weight acts either to accelerate or retard the movement, and of the two, the accelerating effect is more marked, as reducing errors and promoting regularity.

There is no appreciable tendency for the weight to reduce the ruled lines when retarding motion, nor is the weight as accelerating, able to extend the line beyond the point set in the free motion.

When contrasting averages from right- and left-handed subjects it may be said:

As compared with free motions there is a slight reduction of error and irregularity more marked with the left-handed subjects, but a general close correspondence of results.

The question is now appropriate, why should the right-handed men show a reduced error for speed-increase, while the left-handed subjects show the reverse? Bearing in mind that the right hand is the more dexterous or better trained in the former case, it may be suggested that the order of record from 20 toward 200 beats is such as to cause more accurate results at the upper limit, in spite of the fact that less time is allowed for discrimination and adjustments; on the other hand, left-handed subjects have much less advantage of practice and habit in their use of the right hand, and will show the predominance of error, when the ruling is too rapid for careful discrimination.

It becomes a struggle between automatism, or semi-automatism, on the one hand, and discriminative processes on the other.

Visual control is not an advantage in the case of accelerating weight, and the large reduction in error with visual control for the free movements is not evident with weighted motions.

For the left-handed subjects we find that the eyes-closed record shows closer work than does the eyes-open curve; it is lower and nearer the line of 0 error; in this respect, it shows the same effect as with the free-motion curve, and to a less extent as for the retardation weight-record. The accelerating record is, however, more accurate and regular than either of the other curves.

It will be clear, then, as observed, that for constant errors, visual control tends to reduce errors and steady records whatever the speed-increase, as far as right-handed subjects are concerned, but this effect is not noted for left-handed subjects using the right hand, and, with their left hand, visual control is a disturbing element.

Further this erratic effect of visual control is less marked but clear when the weight acts as a retarding factor, but is much more noticeable for the free-motion record.

Individual records show few lapses of control for either errors.

The bulk of the evidence is that the weight imposition, whether acting as a retarding or accelerating influence, is effective in rendering the results more accurate and regular, though at least one subject exhibits the opposite effect for the accelerating weight.

The left hand is better for J., Le., and W., but is less regular for all subjects, save Le. and W., showing again a somewhat complex mass of testimony, from which we may conclude that the right hand is the more accurate member for right-handed subjects, and to a much less extent the left hand is preferred by the left-handed subjects.

Visual control is to be noted as effective for accuracy and regularity, except for Ha., where the curves closely intertwine, and for J., where the eyes-closed record is much better.

Weight-accelerating movements:

For the average of seven subjects we find

visual control is of doubtful advantage, for left-handed subjects, but shows a clearly marked reduction of error for right-handed subjects.

Mean errors are:

(1) similar in all respects to free-movement results;

(2) acceleration-curves are closer to free-movement results than are retardation records;

(3) the more trained hand shows reducing error for speed-increase, while the other hand shows increasing errors, because of superiority of practice-effects over the native tendency to increase error as speed-rate rises, for the more dexterous hand alone.

Constant errors:

(1) there is no tendency to overrule, as compared with free movements, when weight acts to accelerate movements, for there are even cases of lines being shortened with accelerating weights;

(2) a weight seems to negate the results of visual control, as a rule.

For individual records we find:

(1) fatigue-points, for the right hand only, are to be found in a few cases;

(2) weight promotes regularity and accuracy;

(3) visual control is effective only for reducing variations of error;

(4) the better trained hand is the more accurate in the records, to a slight extent;

(5) there are evidences of semi-hypnotic or dreamy states in the non-visual series.

COMPOUND MOTIONS

Series of records were taken at 100 mm. and 10 mm. bases for the hands, with what is called compound motion. This consisted in an additional movement of the hand that was not ruling with the pencil, in a similar manner, as regards the amplitude and general character of the motion, but in an opposite direction.

For example, suppose the left hand is ruling a 100 mm. line outward, or to the left; coincident with this movement would be a similar motion of the right hand outward or to the right. The origin of both motions, or the starting-ends of actual and imaginary ruled lines, was optional, it being desired to bring out the effect of such additional motion, as little complicated as possible with restrictions, as to its position or extent. Actually this distance varied from about 10 mm. where both motions were outward to 600 mm. for inward motions.

A comparison of such compound motions with single-hand records shows in general the following:

For 10 cm. lines:

The case for mean errors may be summed up by saying:

(1) left-hand records are less accurate and regular than the right-hand curves;

(2) visual control reduces error and irregularity in all cases, but is more marked with the left hand;

(3) errors increase with speed-increase;

(4) compound-motion records show little increase in error or irregularity, as compared with the simple motions.

For constant errors:

No marked peculiarities are to be noted, but in general,

(1) left-hand records are less accurate and regular;

(2) visual control reduces errors and irregularity;

(3) errors reduce with increase of speed, except for compound motion uncontrolled visually;

(4) compound-motion errors are not much greater, nor is the irregularity increased.

INDIVIDUAL CASES

100 mm. hand with compound motions.

A glance at the charts shows for individual records a few examples of inhibition of voluntary control for both constant and mean errors, it being much more marked in the case of mean errors.

These lapses of control appear for constant errors for A. with l.h.e.c. at 160 and 180 beats, for mean errors for G. with l.h.e.c. at 200 and with r.h.e.c. at 180 beats; for Le. with l.h.e.c. at 50 and with r.h.e.c. at 160 beats; for A. with l.h.e.c. at 200 beats; thus giving evidence that the visual element has a steadying effect, and that the left hand is less reliable save for Le.

There seems reason for contending that the compound motion can be carried out, as arranged, without loss of accuracy or regularity on the part of the ruling hand, and further that the subjects are pretty generally apt to react to a given stimulus within certain rather narrow limits of accuracy.

The evidence is here pretty conclusive that the right-handed subjects, as a whole, show greater accuracy by about 25% for the more dexterous hand; but it will be wise to consider the individual cases on this point.

Greater regularity and accuracy for the right hand is attained by all right-handed subjects, while the preference of Le. for the left hand is clear but much less definite.

For individual cases:

The evidence again is fairly well marked that the more practised hand will give a better account of itself even when visual control is not called on.

The results for compound movements of the hand for 1 and 10 cm. lines are summarized as follows:

It should be kept in mind that the compound records were in all cases taken in connection with a duplicate series of lines for one hand, and called simple movements. These simple movements correspond with the free-movement records that have been considered already.

The purpose has been to bring out the modification of results that a compound movement introduces, rather than to bear heavily on intrinsic phenomena, i. e., comparison is deemed more important.

For lines 10 cm. long:

Average of seven subjects:

We find for mean and constant errors:

(1) left-hand records are less accurate and uniform;

(2) visual control increases accuracy and regularity, especially for the left hand;

(3) there is an increase as the speed increases for mean errors, and a decrease for constant errors;

(4) compound movements are practically as accurate and regular as the simple ones for constant errors.

For individual cases:

(1) a few lapses of control or fatigue-spots more marked for mean errors with the l.h.e.c., for the visual sense steadies ruling, and the left hand is less reliable;

(2) compound and simple records show close agreement;

(3) the more trained hand reacts more accurately, and with greater regularity;

(4) non-visual records show a cautionary shortening of line at low speeds, and another at the upper limit, the latter being due to physiological limitations.

For 1 cm. lines:

As far as averages are considered:

We may say, then, for mean errors:

(1) that visual control is effective for reducing errors, and increasing steadiness in both sets of records, being more marked with the less trained hand, the left;

(2) only in the case of the l.h.e.c. curves is the movement of the free hand noted as appreciably affecting the accuracy or steadiness of the record.

(3) eyes-closed records in general show a considerably greater error at 20 beats that practice rapidly reduces up to 40 to 60 beats.

We may note for constant errors:

(1) all errors are positive and confirm the earlier deductions on this point;

(2) visual control reduces error and improves steadiness of record;

(3) the free hand-movement does not affect either the accuracy or uniformity of results;

(4) errors do not increase with speed.

For individual records:

Comparing the individual cases of simple and compound movement, there is no particular reason for concluding that the compound movement is a disturbing influence as far as the records of all subjects are concerned, save possibly the lapse of G. at 20 beats, and on the other hand a case of greater accuracy and evenness for compound movements for Mo. with r.h.e.c. constant error.

It is, then, possible to extend the conclusion of the 10 cm. records, and say that both lengths of lines are ruled with a fairly constant limit of error, whether the movement be simple or complicated by movement of the free hand.

Individually there is testimony in favor of the gain in accuracy with visual control for Hu., Hy., and Le., while the crossing of curves for the other subjects shows that there is no difference in eyes-open and eyes-closed results, the general conclusion being in favor of the value of the eyes for accurate results.

For mean errors the right hand is more efficient in the case of A., Hu., Hy., and Me., while the reverse is the case for the rest, and the evidence goes to suggest that greater accuracy can be attained with the more practised hand.

For constant errors the right hand is more accurate in the case of A., G., Hy., only; Me. and Mo. are equally accurate with the hands, and the rest show a marked preference for the left hand, the evidence being thus conflicting, pointing to the theory of ambidextrous development on the lines of accuracy.

L.h.e.c. records are horizontal for all subjects except Hu., Me., and Mo., who show an upward slope to the curve. Evidences of visual control as giving greater accuracy are noted in general above 70 beats and individually for Hy. and Mo., only, the remaining records being so intertwined that no difference can be noted, all suggesting that the eyes are of but little assistance when the left hand is considered. The right hand is preferred with eyes closed by A., Hu., and Le., while four right-handed subjects testify that the less trained hand is more accurate.

The testimony here seems conclusive as pointing to a denial of the current notion as to the greater accuracy of the right hand for right-handed subjects, and of the left hand for left-handed subjects, and further suggests that visual control is a large factor in the supposed superior excellence of the hand mentioned.

SUMMARIZING

For lines 1 cm. long:

Average of seven subjects:

It may be said that:

(1) visual control reduces both mean and constant errors, especially for left hand;

(2) errors are constant whatever the speed;

(3) constant errors are positive showing overruling in all cases;

(4) there is no disturbance created by the second-hand movement, save for r.h.e.c. mean errors, where the accuracy is less for the compound records; this is probably due to the fact that this record shows the least evidence of voluntary control, and is thus most subject to disturbances;

(5) there is a marked reduction of mean error 20-50 beats, probably due to practice.

For individual cases note:

(1) fatigue-spots for non-visual mean errors only;

(2) the equality of result of both types of movements is noted for all cases;

(3) the non-visual right-hand records for some subjects are more accurate;

(4) the more trained hand is not, as a rule and subject to exceptions, the more accurate one, especially for the non-visual records; and

(5) there is evidence that the superior accuracy of the right hand for right-handed subjects is largely a matter of visual control.

HEAD-RECORDS

There was no attempt made to differentiate the visual element because the very movements of the head prevent the full use of the eyes; as a matter of fact, the subject's attempt to make use of the eyes and the aid is more marked at slow speeds and upon facing the apparatus. It is to be noted here that the visual element, as reducing the error at low speeds, is equally marked whether the eyes are directed toward the recording pencil or not. This raises an interesting question as to the direction the eyes must take for the optimal result; must the eyes be fixed on the moving pencil, on its immediate surroundings, or may they wander at will about the surrounding objects?

My own introspective testimony, corroborated by others, who have acted as subjects for this investigation, is that the eyes are most effective when gathering spatial relations in a gross way, and it may be expected that the effects of visual control as reducing errors will be equally efficient, whether the recording pencil be screened or visible, provided it be possible to bring on the retina objects that are grouped about the centre of attraction, the pencil, but not in its immediate neighborhood.

The records show that there is underruling at the higher speeds because of physiological limitations; but this shortening is greater for the backward movements, for the position of the subject is such as to lead to greater uncertainty as to the exact length of ruled line, and it is probable that a cautionary or inhibitory feeling is the cause of this shortening beyond what will be clearly due to inability to perform the desired movement.

Further, visual control is effective, in the case of constant errors, in lengthening the ruled lines at high speeds, and thus reducing the negative constant error.

While the muscular control of the head is a constant, whether the movement be forward or backward, it is less effective for constant error reduction when the head is moved backward. Consequently, while the backward and forward curves are fairly well in correspondence, there is some reason for offering the proposition that either the eyes are of assistance in forward movements to reduce mean errors at high speeds, and they are of no such value for backward movements, or the muscular control of the platysma myoides, trapezius and associated muscles of the neck group is more nearly perfect for movements of the head forward than for backward motions, the latter being to my mind the better hypothesis.

The results for head-movements for lines 1 and 10 cm. long are summarized:

For lines of 10 cm. length:

Average of six subjects:

For mean errors:

(1) the curve for head-forward and head-backward closely corresponds to l.h.e.c. record; the errors increase by 50% with increase of speed-rate, suggesting that

(a) visual control is negligible, as far as seeing the moving pencil is concerned;

(b) control of head for forward equals that for backward movements.

For constant errors:

(1) there is underruling at high speeds because of the usual physiological limitations, and this is more marked for head backward results, suggesting that

(a) spatial relations are obtained, when the apparatus is visible, that tend to correct underruling, or

(b) an extra inhibitory effect, due to lack of knowledge of spatial relations, is added to the normal physical shortening and the subject moves the head a less distance than is naturally possible; or

(c) the muscular control is less complete for movements of the head backward.

For individual cases we find:

(1) fatigue-lapses are less in magnitude than for the hands, because the head-movement can be only a fraction of the forearm-movement;

(2) mean errors increase and constant errors decrease with speed-rise;

(3) similarity of individual head-forward and head-backward curves is suggestive, taken with the fact that no typical form of curve is to be found;

(4) head-backward constant errors are greater and less regular in all cases, suggesting that the eyes, in head-forward records, by getting spatial relations, are more efficient.

For lines 1 cm. long:

Average of six subjects:

For mean errors note:

(1) the head-backward records are less regular than the head-forward ones, and rise a little with speed-increase, showing visual assistance for accuracy or better muscular control for the forward movements or both;

(2) the constant errors show shortening of ruled lines at high speeds a little more marked for the head-forward results;

(3) there is constant overruling.

Individual cases suggest:

(1) fatigue-spots are apparent, especially for head-backward movements;

(2) errors do not increase with speed;

(3) the movements of the head forward are under better control.

FOOT-RECORDS

10 cm. records show that

(1) the eyes are of no assistance as to increasing accuracy but help in promoting regularity of error;

(2) a shortening of ruled lines with speed-increase is noticeable, and is probably due to the usual physiological reason;

(3) the feet are capable of less accurate motion than the hands, but show better results than the head;

(4) mean errors increase but constant errors decrease with speed-increase.

Individual records show:

(1) less violent fluctuations of errors in all respects than do the results of head or hands, for vertical foot-movements are of less extreme extent than are arm- or head-motions;

(2) that for visual control with mean errors, no foot is the more accurate, and there is no reason to believe that the feet are unequally educated.

1 cm. records show, as far as mean errors are concerned, that:

(1) visual control is of no value as either reducing actual errors or as effecting greater regularity;

(2) Errors for foot-movements are no less, but considerably more regular than for head-motions;

(3) errors for hand-movements are more regular, and only 50% of the results for either head- or foot-movements;

(4) all curves are horizontal;

(5) there is no appreciable advantage as to accuracy or regularity that can be attributed to either foot. The evidence goes to show that the subjects are ambipedalous, if it be permitted to coin such a word.

In general, we find that, as far as constant errors are concerned,

(1) visual control does not help to reduce actual errors or promote uniformity;

(2) errors for foot-movements are less than the head records, and but little greater than the hand results, while the regularity for the feet is comparable to the hand, and much greater than for the head;

(3) all curves are horizontal;

(4) there is no particular advantage that either foot has over the other either as to accuracy or regularity.

The evidence is that the subjects were ambipedalous, as far as ability to reach a certain point equally well by either foot is concerned. The popular notion has been to the contrary, and it is a point of considerable importance to note the last point.

For example, in kicking, as developed by football trainers, it is commonly assumed that the right foot for right-handed subjects should be developed, and the opposite foot for left-handed men. Or again, in the case of a person lost in the woods and walking in a circle, it is observed that right-handed persons will turn to the left; probably because of the pace of the right foot being slightly longer than the left. My reply to this evidence will be that the data herein presented is for vertical movements of the foot, starting from the floor in every case, the subject being seated in a chair.

On the other hand, it is an entirely different movement, calling for a much different and greater muscular control in the case of kicking or walking that must be considered. For this reason the evidence, while conclusive within its range, is not offered as more than suggesting that the feet are equally well trained for the usual adjustments, and only an exhaustive investigation covering all possible foot-movements will settle the question.

The result for foot-movements for lines 1 and 10 cm. long is here summarized.

For lines 10 cm. in length:

Average of seven subjects:

Note in general that

(1) l.f.e.c. mean error is most erratic, while the same curve is the most accurate, as far as constant errors are concerned;

(2) the left foot mean and constant errors are slightly greater than those for the right foot, for visual records;

(3) mean errors increase and constant errors reduce with speed-increase;

(4) the visual sense improves regularity, but does not reduce errors;

(5) there is a physiological reason for the shortening of lines at high speeds;

(6) the feet are more under control than the head, but less than the hands.

For individual records:

(1) fatigue-lapses, all for non-visual, are less numerous and of less magnitude than for the hands and head, for the vertical movement of foot is likely to be of less extent than that of head and hands for the particular motion required here;

(2) there is no foot capable of being called more accurate than its mate;

(3) the eyes appear to be of no value for reducing or regulating errors for foot-movements.

For lines 1 cm. long:

Average of six subjects:

It may be said in general that

(1) the visual sense is valueless for promoting accuracy or regularity of curve;

(2) errors of foot-movements are more regular and, for the constant errors, more accurate than for the head-records;

(3) errors of foot-movements are less and less regular by 50% as compared with records for the hands;

(4) errors do not increase or decrease with speed-changes;

(5) the feet are equally accurate.

For individual results:

(1) fatigue-lapses and cases of large error-increases are noted in a number of subjects, both for visual and non-visual records;

(2) further, evidence is available as to the indifference to visual control;

(3) no preference for either foot is to be discovered.

The results for individual choice of rhythm.

In this series of records, the metronome was dispensed with, and the subject was permitted to react as he desired, taking the speed preferred because of ease, pleasure, or other reason.

Records were obtained for six subjects for feet, head, and hands, both single-hand and double-hand movements, all for lengths of line 1 and 10 cm. The charts for individual choice were plotted for a comparison of speeds rather than for accuracy.

It was noted for the hands:

(1) that every subject reacts more rapidly with the left hand;

(2) the eyes had little effect as to changing the speed-rate;

(3) single and double hand-movements were equally rapid.

Some subjects, as A., react more rapidly for the shorter lines, though no clearly marked evidence of this speed-increase is to be noted.

For the head, the results for both eyes opened and closed show the impossibility of separating the optimal or preferred rate of speed on the score of visual assistance or because of direction of head-movement.

There is a close agreement of the subject as to his best speed, and this is independent of special conditions; for example,

A. selects 50-57 beats per minute for 1 cm. and 48-68 for 10 cm.; G. has a preference for 61-66 and 56-71; Hu. rises to 103-125 for 10 cm. and selects 68-82 for 1 cm.; Le. 52-55 for 1 cm. and 45-52 for 10 cm., and so on.

We may say, then, that free rate-choice for head-movements results in a selection of some rate of speed that is not affected by the visual sense or direction of movement, and is strictly individualistic, covering a range of 50-130 beats per minute, and not increasing as the amplitude of movement is reduced.

Turning to individual choice of speed-rate, for the feet it will be seen that

(1) the non-visual records closely correspond as to chosen speed, and there is a less close correspondence of visual speeds;

(2) the visual records are ruled at a lower rate in some cases, but A., G., and Mo. show little difference;

(3) there is a tendency to speed up as the series progresses;

(4) the shorter lines are ruled with greater speed as a rule, though G. and Le. fail to show this phenomenon;

(5) The left-foot records show a higher speed-rate for all cases.

Among many interesting points that cannot be examined in this connection, such as relation of voluntary choice of rate to the main line of metronome records as regards accuracy, the fact of the higher rate of ruling for the left hand and foot stands most prominent.

Whether a record of head-movements to right or left, or other devices to compare the sides of the body or to contrast arm and leg speeds, will bear out this testimony is as yet unknown, so that the writer prefers to announce the result and not now fit theory to data. It may be said that the records were taken in reverse order and rearranged, as regards right and left foot or hand, and, in addition, the initial foot-movement varied with the subject, some being right and some left.

We ask finally: Is the time in which the greatest exactitude is produced, the same for every group of muscles; that is, has every motor apparatus the same natural rhythm? and: Is this natural rhythm a constant rapidity for all motor nerve-centres or does it depend upon the complexity and character of the movement?

The comparison will fall first on the averages and finally on the individual records.

The hand-movements show the following results:

Constant errors for 14 cm.:

For simple and weight accelerating and retarding motions, there is a close agreement about 120 beats for the minimum error for visual and right-hand non-visual records; left-hand non-visual records are spread more, but will also average the same.

For 10 cm. simple and compound movements the visual minimum errors are at 180-200 beats, while with the eyes closed the results are grouped about 60 beats; one record, that for l.h.e.c., has two minimum points at 60 and 180 beats, the latter being clearly a crossing of the 0 error-line, because of physiological limitations.

For 1 cm. simple and weighted minimum errors are grouped between 20 and 60 beats, while the simple and compound group show less regularity and a tendency to group minimum errors at 100 beats.

The head-movements show for both 1 and 10 cm. lines a minimum error at 180-200 beats, there being, however, one exception at 100 beats for 10 cm. head-backward movements.

The foot-movements show minimum errors at 80 beats for the right foot, and 180 and 100 beats for the left foot, visual and non-visual respectively.

Bearing in mind for a moment the individual choice records, there seems here a suggestion that the left foot is capable not merely of higher speeds, but of minimum errors at the higher rates as compared with the right foot.

No such differentiation of the hands can be discovered, however.

Mean errors:

For the hands:

For 14 cm. for simple and weighted results we find that the right-hand and left-hand eyes-open minimum errors are at 180 beats, but the non-visual left-hand minimums are at 30 beats.

For 10 cm. simple and compound records we find all minimum errors are between 160 and 200 beats.

For 1 cm. simple and weighted results there is a scattering of minimum errors from 20 to 200 beats, with a heavy preponderance at 200, and the same is true for the simple-compound series.

The head-movements minimum errors are at 40 beats without exception.

The foot minimum errors are distributed from 20-30 beats for the left foot to 160-180 for the right.

It is thus evident that each group of muscles and each motor centre has its own optimum, and that the conditions of complexity, resistance, etc., influence greatly the accuracy of the periodic movement impulse.

FOOTNOTE:

THE MOTOR POWER OF COMPLEXITY

BY C. L. VAUGHAN

A. COUNTING OF SIMPLE AND COMPLEX VISUAL OBJECTS

If every sensory stimulus has a motor reaction, then a simple figure perceived in any way ought to produce a somewhat different response from a more complex figure similarly perceived. Of course if only one figure of each kind is given it is difficult to measure in any way this difference, since it is so small. But we might make it measurable by multiplying the process. Therefore I have cut out a row of similar figures in a strip of cardboard and on another strip another series of a different pattern. Now if these rows are counted figure by figure each figure has a certain motor effect which influences the speed of counting, so that the time of counting (measured by the chronoscope) should give some indication of the comparative motor power of the figures in question.

In the accompanying illustration nine cards of various patterns are shown. Cards 1, 2, and 3 are comparatively simple patterns while 4, 5, and 6 are comparatively complex, Card 6 having the added complication of different kinds of figures on the same card. Cards 7, 8, and 9 form another group, Card 7 having the same letter throughout, Card 8 having letters composing a sentence and Card 9 a series of the letters, mostly consonants, mixed promiscuously. In order to prevent the subject from knowing the exact number, and thus, perhaps, bring in another influence at the end of the row, most of the different cards have different numbers of figures, but this difference is not great and some cards have the same number. The subject usually forgets, from one experiment to the next, the number on each card.

At first the experiment was performed with the figures in a straight row, instead of in the broken line which is seen in the illustration. In counting the straight rows, the observers found it hard to keep the place in the line. A subject would become confused and count some spot twice or else he would omit it altogether. Furthermore this disturbance was found to be much greater with some figures than with others, with Card 1, for example, more than with Card 2. Therefore the device was adopted of diversifying the line, both by placing some of the figures above and some below the line and by making the distances from one figure to the next, different in the different cases. And in order to prevent the subject from associating any peculiar turn in the line with a certain number counted, it was decided to have the arrangement on the different cards different. But it was still necessary to have the intervals between figures about the same in all the cards, and therefore the row was divided into sections of six figures each and these sections were used as units, variously arranged, in constructing the other rows. For example the first unit of Card 3 is the same as the second of Card 4. Sometimes this six-figure unit is turned end for end or upside down, and thus, though the same spaces are used, the cards appear dissimilar.

The subject would be seated at the table with one hand resting lightly on the key which sets the chronoscope in motion, his eyes raised so that the table in front of him is not seen. One of the cards would then be put in the proper position in front of him (always the same), and he is told that all is ready. He looks down at the card and as soon as he begins to count the first figures in the line he presses the chronoscope key, and when he has reached the end of the line he releases the key. The time for the operation is then noted. The whole series of cards is thus gone through. An extra card of which no record is taken is used for the first few tests so that the subject may be in the proper state when the first test to be noted down is taken. Also the order of the series is changed from one experiment to the next, each card taking its turn at being first and last. It was hoped in this way to distribute among the different cards the effects of practice and fatigue, and also to guard against any expectations on the part of the subject as to the character of the next card.

The subject is told to count as fast as he can, with a reasonable feeling of certainty as to his correctness, the main object being to have a uniform principle, in counting the different series. Wrong counts were excluded, but later on the same cards given again so as to keep the tables even. Subjects were not allowed to count the figures by groups, but one by one. At first a certain amount of difficulty was found in the fact that subjects in counting would repeat the numbers to themselves, and as they seemed to be retarded by this, especially in those numbers whose corresponding names have 2 or 3 syllables, the result was that we were getting the speed with which subjects could count the numbers from 1 up to 38 or 39 and this would be practically the same whatever the figure. But all the subjects were finally trained merely to think the number, or at least to have as little vocal adjustment as possible. When this was done the subject no longer felt that it was the speed with which he could count that was being measured but the rate at which he could take in the different figures on the card, one at a time.

Between three and four hundred tests were made of the counting of the figures on the nine cards, the work being divided among seven subjects, though not in exactly equal amounts. Since the number of figures on the different cards are different, I have found the time it takes to count one figure by dividing the total time by the number of figures on a card. The following table shows the average time taken by each subject for one figure on each card, time given in thousandths of seconds. A.M.V. stands for average mean variation.

A. A.M.V. B. A.M.V. C. A.M.V. D. A.M.V. E. 1 279.69 11.47 186.87 13.22 247.62 14.89 193.08 12.38 262.77 2 270.60 12.47 180.55 11.88 249.21 18.00 190.51 11.82 257.56 3 274.43 9.87 180.89 11.57 247.59 15.51 192.07 7.87 259.96 4 286.82 12.47 190.39 12.56 255.20 16.78 200.53 10.72 267.11 5 290.29 11.89 195.41 12.36 262.27 19.73 199.89 9.27 271.06 6 293.06 12.21 185.33 11.51 275.40 18.13 199.20 7.92 264.59 7 273.32 15.54 192.23 13.73 229.26 19.30 185.60 9.83 265.56 8 279.77 13.86 185.23 12.69 246.66 18.86 193.39 9.81 277.52 9 285.09 11.97 197 04 12.28 269.96 19.95 186.30 9.05 259.72

A.M.V. F. A.M.V. G. A.M.V. 20.20 217.00 12.32 442.63 36.51 16.03 195.00 9.50 431.00 24.39 17.41 191.50 11.03 434.83 24.28 20.31 226.40 29.11 445.71 14.58 20.86 233.50 20.46 459.17 18.92 15.00 220.80 14.92 432.17 26.87 19.20 189.20 30.31 402.13 21.34 14.93 220.70 21.79 388.77 27.48 14.27 210.34 11.71 419.57 18.22

A, B, C, D, E, F, G are the different subjects, and 1, 2, 3, 4, etc., refer to the cards with the different patterns. It is seen at a glance that great differences exist between the rates with which the different subjects count. Subject G had much fewer tests than the others, and thus, not having as much training, his average is higher in comparison than it would be had he had the same training.

Now if we compare the counting of the first three or relatively simple patterns with that of the next three or comparatively complex ones, we notice at once that the simple figures are almost invariably counted in less time than the complex, there being only two exceptions. B counts 6 a little faster than 1, and G counts 6 faster than 1 and 3. Even these apparent exceptions are easily explained. As noted already, subjects are much more apt to lose their place in counting certain cards than others. This is especially true of Card 1 even after the line is broken. Now Card 6 is arranged on a different plan from the others, for it has many kinds of figures on it. This is a great help in keeping one's proper place in the counting of the series, and since wavering between two figures is avoided, the series is counted more rapidly. But B is the most rapid in counting, of all the subjects, and it is natural that any differences in the ease of keeping place should show themselves here, since the more rapid the counting the easier it is to lose the proper position. This cannot be said of G, who is a slow counter, but on the other hand it may be noted that he had only a few cases, and at first the ability to keep one's position is much less than after considerable experience. So in Cards 6 and 1 there are two conflicting principles, degree of complexity and tendency toward confusion of position. Of course both these principles are present in all the other cards, but they reach a maximum in 1 and 6, in 1 extreme simplicity with difficulty in keeping place, in 6 extreme complexity with ease in keeping place. Card 1, it will be seen, is with nearly all subjects a little slower than 2 and 3, while 6 is generally faster than 4 and 5.

Therefore it would seem that the apparently small exceptions are not real exceptions, but variations due to the presence of other factors than mere differences in complexity of the figures used. In observing the averages for 7, 8, and 9 we see that as a rule 7 is fastest, 8 next, and 9 the slowest. The tables are not quite so regular as for the cards just given. B and G count 8 faster than 7, and E counts 9 faster than 7. The most of these cards have on them 36, 37, 38, or 39 figures. Card 8 has 43 letters. The subjects report that the last three on this card are counted much faster. They know, as soon as they reach 40, just how many there are, and it is hard to keep from counting the rest in a group. Otherwise they do not feel any difference in counting Cards 8 and 9. Arranging the letters in words does not affect the speed of counting, so far as they can see, for in counting they do not notice the words at all.

When we average the records of all the subjects giving equal weight to each subject, though the number of tests may be different with the different men, we get the following table. Time given in thousandths of seconds.

(1) 261.38 (2) 253.49 (3) 254.54 (4) 267.46 (5) 273.08 (6) 267.22 (7) 248.19 (8) 256.01 (9) 261.15

It is seen, from looking at this table, that all divergences from the general rule have stopped. Cards 1, 2, and 3 each take less time than any of the 4, 5, 6 group, and 7 is faster than 8 and 9. So the evidence seems very strong that it takes longer to count complex than simple figures. Should one object that the difference is extremely small, a few thousandths of a second, and that thus a slight error in one test might invalidate the result, we reply that the time which is given is the time in which we count just one figure of the given pattern, and that thus of course the difference between counting two different figures must be very small. Moreover there has been a remarkable agreement of the tests taken at different times. It is not a case of finding 1, 2, and 3 counted faster one day and 4, 5, and 6 counted faster the next, but 1, 2, and 3 are counted faster nearly every time. Occasionally 1 will take longer than one of the 4, 5, 6 group. And extremely seldom is there a case where the average of 1, 2, and 3 is not less than that of 4, 5, and 6.

The experiment seems to have proven that it takes a longer time to count a row of complex figures than a similar row of simple figures. The complex figure exercises a retarding effect upon the eye as it sweeps along. There is a greater amount of sensory stimulation, consequently a greater amount of motor excitement. This motor excitement does not act in harmony with the motor activity which impels the eyes along, but has a somewhat antagonistic effect. The eye is held more by the complex figure; it is a greater effort to withdraw the gaze to look at the next figure. A certain interest, as we say, on the psychological side tends to hold one to the figure looked at. This interest is greater (other things being equal) the greater the complexity of the figure. The nervous processes involved in counting, though admittedly in very small degree, are thus inhibited by the complexity of the figure and act more slowly.

B. REACTIONS TO SIMPLE AND COMPLEX OPTICAL IMPRESSIONS

Since the preceding experiments seem to show that reactions on optical impressions are different according as the figures are more or less complex, it would seem that we ought to be able to measure by graphic methods the reactions to visual fields of varying grades of complexity and in this way to demonstrate their different motor powers.

A Porter kymograph was used on which to register the reactions. Resting on the top of the drum, and revolving with it, was a circular band of white paper, upon which were pasted the different figures to be observed. A screen was placed in front of the kymograph, thus concealing the figures; but at their level was a little square window in the screen, which, when the eye was placed in the proper position, allowed the subject to see one of the figures but nothing more. A few inches in front of this window was an eye-rest which kept the eye properly placed. A tambour received the movement from the subject and communicated it to a straw which made a scratch on the smoked paper which covered the drum.

The figures used in this experiment form two series, one, composed of geometrical figures, varying in complexity from a circle to a very complex figure consisting of many overlapping squares, triangles, etc., and the other composed of colored figures varying in complexity from a simple square of one color to a very complex mixture of various colors. The area of the visual field is about the same in all cases,--an inch square. The geometrical figures were formed of black lines on a white background. The figures used are shown in the accompanying illustrations.

The subject would be seated in front of the screen, his eye at the eye-rest a few inches in front of the window in the screen, and the forefinger of the right hand on the tambour, which is to the right of and behind the screen, and thus not seen while the eye is at the rest. Then as the drum revolves and brings a figure in front of the window, the subject observes this figure carefully, and when it is all in the field of vision he presses down with his forefinger, thus producing a curve on the drum surface. He tries to make the same finger-movement every time, whatever the figure at the window may be. But his attention is not to be too much taken up with the making of the movement, for he must be closely observing the figure. If he looks at the figure until he observes its characteristics clearly and then turns his attention from this to the finger-movement, it is evident that the optical sensation would not have much effect upon the movement. The movement must be performed while his interest in the figure is highest. Now, after a little practice, any one can accustom himself to make a certain definite movement in about the same way every time, and he can then agree that he shall make this movement as a reaction to a given stimulation. Then when the stimulus comes he makes the movement without any longer thinking of the character of the movement. It has become, to a certain extent, automatic and can look out for itself.

This is the state into which I have tried to get my subjects. Their whole attention is to be taken up with the seeing of the figures in the window, and to these figures they are to react as automatically as possible. Thus, though finger-movements are usually voluntary, all the capricious character of voluntary action will be removed here, and if the stimulus is the same in all cases, the reaction tends to assume the form of a uniform movement. There is, then, a chance to see the influence of different optical stimuli upon this action.

Six different geometrical figures were seen at each revolution of the drum and six reactions given by the subject. Between figures a white surface would occupy the field of vision. The simple and complex figures were distributed so that the subject never knew what kind of a figure would come next. The purpose of the experiment was kept as much as possible from the knowledge of the subjects; but some, knowing my general problem, surmised quite correctly my main object here.

Ten revolutions were made at each sitting, thus causing the subject to react ten times to each figure. Then a new drum paper was taken and the case with the colored figures placed upon it. This had five colored figures, and ten revolutions were made also in this case. Thus, in all, in any one day, the subject would make one hundred and ten of these finger-movements.

Since we have in all these experiments tried to find out in the different figures merely differences in the amount of the reaction, and not differences in the character of the reaction, we shall keep up this method here. Now a stronger reaction makes a higher curve, and since the drum is all the while revolving, and since the higher the curve, other things being equal, the longer it takes, the stronger reaction will also make a wider curve. So it would seem that if we wish to observe the differences in the amounts of reaction the most natural course to pursue would be to measure the heights and widths of the curves we have registered. This accordingly has been done.

In our discussion of these measurements let us, then, first, take up the curve heights, and of these, those of the geometrical figures which we call U, V, W, X, Y, Z. The height is measured from a base-line [drawn by revolving the drum after the subject has taken his finger from the tambour] to the highest point reached. These measurements are taken from two hundred reactions to each figure, divided among seven different subjects.

Heights of Curves U V W X Y Z Subject A 6.83 6.68 6.59 6.55 6.63 6.79 B 8.64 7.26 6.41 7.79 6.39 9.75 C 6.67 6.55 6.73 6.85 5.87 8.53 D 21.35 21.26 21.46 21.90 21.33 21.31 E 16.13 15.77 15.17 15.85 15.29 16.08 F 16.90 16.97 16.14 16.52 15.81 17.91 G 11.42 11.32 11.39 11.48 11.06 11.10

87.94 85.51 83.89 86.94 82.38 91.48

Average 12.56 12.26 11.98 12.42 11.77 13.07

Arranged in order of height of curve Z U X V W Y 13.07 12.56 12.42 12.26 11.98 11.77

If we put the figures in the order of strongest reaction for the different subjects we get the following table:

Subject A U Z V Y W X B Z U X V W Y C Z X W U V Y D X W U Y Z V E U Z X V Y W F Z V U X W Y G X U W V Z Y

It is seen from these results that, although the subjects differ, the height of the curve varies directly with the complexity of the figure. The order of the figures, which we get by measuring the height of the curves and then putting that figure with the highest curve first, with the next highest second, and so on, is exactly the same order in which we should put them if we were asked to put the most complex first, the next second, and so on. Though the individual subjects may vary somewhat from this rule, when they are all grouped together there are no exceptions.

The variations of the reactions with the different subjects may be shown very clearly in the following way, where the different figures are in the left-hand side arranged in order of descending complexity. "1st place," etc., refer to the order of arrangement of the figures by the different subjects as shown in preceding tables. Thus, Z, 3 times, 1st place, means that three subjects have in the average a higher curve for Z than for any other figure.

1st place 2d place 3d place 4th place 5th place 6th place

Z 3 times 2 times 0 times 0 times 2 times 0 times U 2 times 2 times 2 times 1 time 0 times 0 times X 2 times 1 time 2 times 1 time 0 times 1 time V 0 times 1 time 1 time 3 times 1 time 1 time W 0 times 1 time 2 times 0 times 3 times 1 time Y 0 times 0 times 0 times 2 times 1 time 4 times

One can see at a glance from this, how, as the figures decrease in complexity, they take their position further on in the series. If a diagonal is drawn from the upper left-hand corner to the lower right, it will pass through or near the larger numbers in the table, thus showing that the figures belong in the ordered series in the places already shown.

Next in order let us take up the measurements of the widths of curves for the same geometrical figures which we have been considering.

Widths of Curves in mm.

U V W X Y Z Subject A 20.83 20.59 20.93 21.22 20.21 21.89 B 11.18 10.77 10.46 10.31 9.92 10.79 C 4.28 4.43 4.10 3.78 4.95 4.70 D 21.08 19.36 18.33 18.75 18.17 21.09 E 14.22 13.85 13.40 13.56 11.96 14.13 F 17.00 15.26 15.92 16.52 14.52 16.47 G 5.25 5.19 5.30 5.08 5.11 5.37

93.84 89.45 88.44 89.22 84.84 94.44

Average 13.40 12.78 12.63 12.75 12.12 13.49

Order Z U V X W Y 13.49 13.40 12.78 12.75 12.63 12.12

If as before we take the orders for the different subjects, we get the following table:

Subject A Z X W U V Y B U Z V W X Y C Y Z V U W X D Z U V X W Y E U Z V X W Y F U X Z W V Y G Z W U V Y X

Here, as before, in the case of the heights, it is seen that though the order is different with the different subjects, yet the general tendency is to place the most complex figures first and the simplest last. The most simple figure Y never comes in front of the fifth place except with subject C, who places it first. This exception may be ascribed to the fact that this subject, on account of his going away, did not have so many tests. In fact only one day's work of 10 reactions for each figure is recorded, and it is but natural that some variations from the standard should occur in his case.

If now, as before, we investigate where each figure occurs in the series for the different subjects we get the following table:

Times in 1st place 2d place 3d place 4th place 5th place 6th place Z 3 3 1 0 0 0 U 3 1 1 2 0 0 V 0 0 4 1 2 0 X 0 2 0 2 1 2 W 0 1 1 2 3 0 Y 1 0 0 0 1 5

Here we again see the large numbers on a line from the upper left-hand to the lower right-hand corner.

Thus we get the following order from the geometrical figures as measured by the height and width of the curves:

Height Z U X V W Y Width Z U V X W Y

The only difference, it is seen, is that the positions of V and X are reversed in the two series. Such a change would on our principle be fairly likely to occur, since V and X are figures near to each other in complexity and the motor effects are very similar.

In the same manner, the following tables show the reactions to the colored figures of different grades of complexity. And first, as before, is the table of the heights of the curves for the different subjects, given in millimetres. The numbers given represent the averages of all reactions made. We will call the figures, for the sake of reference, L, M, N, O, P.

L M N O P Subject A 5.75 6.01 5.90 5.82 5.74 B 6.72 5.56 6.35 7.53 4.94 C 10.92 10.90 10.76 10.52 10.99 D 25.49 25.42 26.23 25.89 25.52 E 20.63 20.82 20.37 20.55 20.30 F 15.67 15.23 15.15 15.98 14.51

85.18 83.94 85.26 86.29 82.00

Average 14.20 13.99 14.21 14.38 13.67

Order arranged as before in a descending series according to height of curve:

O N L M P 14.38 14.21 14.20 13.99 13.67

This is exactly, as I should judge, the order of the complexity of the figures reacted to.

The arrangement by the individual subjects is as follows:

Subject A M N O L P B O N L M P C P L M N O D N O P L M E M L M N P F O L M N P

We see that individual differences are stronger here than in the geometrical figures, but that the same tendency to react more strongly to the complex is present in nearly every case. This can be brought to the eye more clearly if we observe the table in which is shown the position of the different figures in the series of the different subjects.

Times in 1st place 2d place 3d place 4th place 5th place O 2 1 2 0 1 N 1 2 0 3 0 L 0 3 1 2 0 M 2 0 2 1 1 P 1 0 1 0 4

M here presents the principal exception, coming too often in the first place.

Finally we give the tables for the widths of the curves for the colored figures; and first the table of the averages of all the subjects for all the figures:

L M N O P Subject A 25.76 24.27 25.06 24.77 23.14 B 9.42 9.49 9.06 9.84 8.11 C 5.85 5.35 5.62 5.80 5.24 D 13.68 13.53 13.18 13.26 13.38 E 22.06 21.37 22.50 22.17 20.44 F 16.30 15.08 16.65 16.76 15.13

93.07 89.09 92.07 92.60 85.44

Average 15.51 14.85 15.35 15.43 14.24

Order, arranged in a descending series according to width of curve:

L O N M P 15.51 15.43 15.35 14.85 14.24

Here the order is not just the same as we got from a measurement of the heights. The three complex figures have changed places somewhat, but there is no exchange of a simple and a complex.

The arrangements by the individual subjects are as follows:

Subject A L N O M P B O M L N P C L O N M P D L M P O N E N O L M P F O N L P M

The three complex figures have different places with different subjects, but very seldom is a simple figure found among the complex, or vice versa.

This can be seen easily from the following table:

Times in 1st place 2d place 3d place 4th place 5th place L 3 0 3 0 0 O 2 2 1 1 0 N 1 2 1 1 1 M 0 2 0 3 1 P 0 0 1 1 4

A theoretical word may close our report.

The growth of biology and physiology has tended to show that there is no break in the nervous mechanism. The stimulus goes to the brain and out through motor channels to muscles, glands, etc. The nervous current does not wait in the brain for the permission of the mind to leave on its journey to a muscle nor does it need mental reënforcement. The nervous current as a whole is a unity. The nervous system is a physiological instrument for producing the appropriate reaction to a certain stimulus. In the unicellular organism there is no nervous system, but the protoplasm receives the stimulus and produces the reaction. As we go up in the animal series a differentiation is seen to be present in the organism. Some parts are more concerned with the receiving of stimuli and others with the approach toward or withdrawal from the stimulating object. There is a division of labor. The nervous system is developed as a means of rapid communication between the different parts, but this communication is a physiological one. The stimulus sets up a chemical action in the sensory organ which is transmitted along the nervous path to the motor organ which is caused to react. As we ascend the animal series the differentiation becomes greater and greater, and consequently the means of communication must become more and more complex. So trunk lines are formed which lead to a centre, and from this centre again go out main lines which divide and subdivide until the muscles are reached. The centre acts as a kind of automatic switch-board.

Accepting such a view of the nervous system it must be granted that different stimulations would produce different reactions. It was my aim in the experimental work which has been described to show that this is true. And while much work has already been done in showing that different kinds, or different amounts of stimulation produce differences in reactions, it seemed important to demonstrate also that mere differences in the complexity of the stimulus bring about differences in the reaction. So the experiment of counting figures of different complexity was entered upon, and we found that it took longer to count figures the more complex they were in spite of the fact that the act of counting seems always the same. The question is how must the fact that counting becomes slower and slower, as the figures become more complex, be interpreted?

When we count a row of figures, the eyes do not move along at a regular uniform rate, but make a quick jump from one figure to the next, halt a moment, make another jump, and so on. Now, I think the principal difference comes in with the figures of different complexity in the time the eye halts at each figure. The halt is longer the more complex the figure is. It is well known that any visual object which stimulates the retina is brought by a reflex movement of the eye to the place of clearest vision. Of two objects stimulating the eye at the same time, the more pronounced one will produce the reflex and will hold the eye longer than a weaker stimulus. Similarly here, the more complex figure produces a stronger reflex and holds the eye longer than the simple figure. This is repeated at every figure in the series.

The complex figures have more features about them, all of which by way of the retina and optic nerve are represented in the cortex and thus more cortical cells are involved, which in turn produce a stronger stimulation of the muscles which move the eye in the proper way to see the figure, and thus the eye is held more strongly by the complex than by the simple figure.

Again in the second experiment, the subject reacts more strongly to the complex as shown already in explaining the first experiment and for the same reasons. It might be said that in looking at the colored figures, e. g., that since the same amount of retina is stimulated, the reaction ought to be the same. But we may presume that the complex figure, on account of the different shapes and contrasts on its surface, will more variously affect the same amount of retina and that the nervous currents sent to the cortex will, many of them, be stronger than those from the simple figure and will thus cause the cortical cells to be more strongly excited, or by a process of irradiation the stimulation will spread to adjoining cells and thus finally more cells be stimulated. However this may be, the amount of discharge into motor cells is certainly greater and the muscular reaction, therefore, also greater.

The interesting side of our results is thus given in the fact that we have here two activities--counting with highest speed and making hand-movements of certain length--which are performed every time with exactly the same intention and with the subjective impression of equal result, and which yet show marked differences according to the complexity of the psycho-physical stimuli. It is a new contribution to our knowledge of the independent motor power of ideas.

ANIMAL PSYCHOLOGY

THE MUTUAL RELATIONS OF STIMULI IN THE FROG RANA CLAMATA DAUDIN

BY ROBERT M. YERKES

I. ANIMAL BEHAVIOR AND THE SENSES

Since the behavior of an animal is conditioned by its senses, it is extremely important that the comparative psychologist should have accurate and detailed knowledge of the sense-impressions received by his subjects. Knowledge of the so-called "special senses" does not suffice for the satisfactory description of behavior, for there are several other kinds of sense-data of equal or even greater importance than those of the five special senses. As investigation of the subject progresses the banefulness of the notion that all sense-experience is summed up in "the five special senses" becomes more and more evident. For the comparative psychologist the senses are not five, six, eight, or ten, but as numerous as are the kinds of sense-data which condition animal activities. There can be no doubt that many of the lower animals are largely dependent upon senses which are not included in the conventional special sense list. The contention that certain organs which are commonly recognized as sensory in function, the cristæ acusticæ of the ear, for instance, are merely reflex control organs, has little weight in this connection, for every sense-organ is part of a motor control mechanism, and so far as we are able to judge from available evidence each has as an accompaniment of its functioning a mode of sensation. If there are two kinds of peripheral organs in connection with the afferent nerves, namely, those whose functioning has sensation for an accompaniment and those in which motor control is the sole phenomenon, it is high time that the fact were definitely known.

Even a thoroughly accurate knowledge of the general condition of the senses in a particular phylum, genus, or species may be of trifling value in the study of the behavior of a given individual, for within these groups the state of development and relative importance of a sense may differ strikingly. Herrick, in his admirable investigation of the sense of taste in fishes, has rendered comparative psychology an important service by showing that even a highly developed sense may be of markedly different value in the associative life of different species. The cat-fish, according to Professor Herrick's observations, obtains its food primarily by the aid of taste-impressions, the hake by the aid of touch, and the sea-robin chiefly by means of vision. All three of the senses mentioned are possessed by each of the fishes, yet their values differ so widely that an understanding of the habits and associative processes of any one of the species would be impossible except in the light of just such facts as Herrick has discovered. Clearly, then, we must know the relative importance of the various sense-impressions received by an animal before we can discuss its behavior or psychic characteristics intelligently.

Furthermore, if behavior is to be serviceably described in terms of stimuli and physiological conditions it is necessary first of all to recognize that an animal responds to a situation, not to any one independent and isolated stimulus. Every situation, to be sure, may be analyzed into its component simple stimuli, but the influence of each and all of these stimuli is conditioned by the situation. Too often in our accounts of an animal's behavior we name some one stimulus as the condition of the reaction and entirely neglect the situation, without which the stimulus would have been of quite different value to the animal. For any given stimulus other external and internal stimuli constitute an environment. The complete description of a reaction demands knowledge of all the stimuli which enter into the situation and of their mutual relations of interference or supplementation. A frog which in its native habitat and undisturbed by an unusual situation would react violently to the light touch of a stick may give no sign of reaction to the same stimulus when a human being stands nearby. The influence of the tactual stimulus has been changed entirely by the simultaneous appearance of visual, olfactory, and possibly still other sense-data (man). The animal reacts not to the touch alone, but to this stimulus as part of a certain situation. The general effect of a situation we often speak of as excitement, timidity, etc. These are words for which must be substituted in our accounts of animal behavior accurate descriptions of the situations. Experimental studies prove that an animal must become thoroughly accustomed to the general situation in which it is to be observed before the influence of any particular condition can be studied to advantage.

Only a few of the important reactions of an animal to either external or internal stimuli are visible to the casual observer, and many of them can be detected only by the employment of indirect methods. Frequently the lack of a visible motor response to a new situation is good evidence of a fundamentally important reaction. The death-feigning opossum, crustacean, or insect, truly reacts by becoming motionless. As Whitman has shown in the case of the leech it is as hazardous to judge of the degree of sensitiveness of another animal solely on the basis of our own as it is to maintain that lower animals possess only the senses which are ours. Varied, indirect and delicate methods are necessary in the investigation of the senses, as the results of the experiments to be described below help to prove.

It is my purpose in this paper to call attention to and emphasize the importance of studying stimuli in their mutual relations of interference and supplementation. This I shall do by presenting the results of an investigation of the behavior of the green frog. I shall discuss briefly, first, the sense-data received by the animal, their relative importance, significance, and mutual relations, and, next, the phenomena of reënforcement and inhibition.

II. THE SENSORY REACTIONS OF THE GREEN FROG

The following sensory reactions have been observed in the frog, but most of them have not been studied with care: olfactory, temperature, visual, tactual, equilibrational, and auditory. It is my purpose to investigate each of these senses in such fashion that we shall know the receptive capacity of the animal. Thus far I have completed only the work on auditory reactions, but the chemical and temperature senses and vision will be discussed in similar fashion later.

At present there is little known concerning the chemical senses. Unpublished observations made by Mr. Sherwin in this laboratory indicate the existence of olfactory sensitiveness to camphor, iodine, and several other strong stimuli. The reactions to the stimuli were slow, however, and there is no reason to believe that the sense of smell is of great importance to the animal. Of taste barely more is known than that it is present.

There is marked sensitiveness to variations in temperature, as I have demonstrated by preliminary test experiments, but the limits, distribution, and significance of this sensitiveness remain to be investigated. I am not aware that the existence of temperature spots has been determined. In connection with a study of the reactions of frogs to light Torelle discovered that the animals suddenly become inactive and usually attempt to bury themselves when brought into a temperature of 8° to 10° C. This reaction is prompt and definite; its value to an animal which hibernates is evident, yet one would scarcely anticipate the suddenness and regularity with which it occurs.

My studies of habit-formation and reaction-time have revealed the importance of vision in the life of the frog. Perception of movement appears to be of far greater value to the animal than perception of form or color. The spectral colors are discriminated in all probability, for the animals react very differently to those of the blue end than to those of the red. According to Torelle blue is preferred to red. There is evidence that red has a higher stimulating value than blue, and the apparent avoidance of red in Torelle's experiments may be due to this fact. None of the work with which I am familiar demonstrates that the suspected color-reactions are due to stimulation of the eye. They may be due to stimulation of the skin, for Parker has shown that the reactions of Rana pipiens to light are due to stimulation of the sink as well as of the eyes, or they may even be due to intensity instead of color.

The tactual-auditory sense series is better known and also, it would appear, better developed than the chemical series. A large portion of the body surface of the green frog is keenly sensitive to mechanical stimulation, and Steinach by measurement of electrical changes in the nerves of the skin has discovered the existence of "touch spots." His method, which is ingenious, promises to be of considerable value in the objective investigation of the senses, but it involves operations on the subject which inevitably destroy the normal condition of the sense.

According to Steinach we have in the negative variation in the electrical condition of nerves during stimulation a phenomenon which may be used in the determination of the threshold of stimulation as well as in the investigation of irritability. In a previous paper I have discussed the associational rôle of tactual impressions as well as the tactual reaction-time. All my observations lead me to believe that touch is a highly developed and important sense in the green frog.

Of the senses intermediate between touch and hearing that of equilibration has been most discussed. Certainly there is good reason to suppose that the sense-organs of the semicircular canals of the ear furnish the animal with impressions of position, movement, and possibly also of direction. Further study of the tactual-auditory senses of frogs may indicate the existence of conditions similar to those discovered by Parker in certain fishes, in which, as he remarks, "the skin, lateral line organs and ears represent, figuratively speaking, three generations of sense-organs. The oldest is the skin stimulated by varying pressures, such as are produced by irregular currents, and capable of initiating equilibrational responses. From the skin have been derived the lateral line organs stimulated by water vibrations of low rate, and also significant for equilibration. Finally, from the lateral line organs have come the ears stimulated by water vibrations of a high rate and important for equilibration. The ear, unlike the skin and lateral line organs, is differentiated for its two functions, the sacculus for hearing, the utriculus for equilibration."

The sense of hearing remains to be considered. My attention was first drawn to this subject by failure to obtain motor reactions to sounds in the investigation of the time-relations of the neural processes of the green frog. Although a large number of sounds of different qualities, pitches, and intensities were employed, no visible motor reactions were observed. This led me to seek the significance of what appeared to be either a surprising lack of sensitiveness to changes in the environment which would naturally be expected to stimulate the animal, or an interesting and important case of the inhibition of reaction to auditory stimuli. This suggested the question, Are frogs deaf, or do they under certain conditions completely inhibit their usual reactions to sound?

In the literature on the senses and reactions of frogs I have found nothing which contributes importantly to our knowledge of the sense of hearing. Most of the investigations which deal with the ear are concerned with the equilibrational and orientational functions of the labyrinth organs, and have nothing whatever to say about hearing. In the natural histories the existence of a well-developed sense of hearing is usually assumed, and numerous instances of what are supposed to be reactions to sound are cited. It is to be noted, however, that none of the observations in these popular works furnishes satisfactory proof of the exclusion of the influence of visual stimuli. Among the few references to frog audition of which I have knowledge, the only one which seems worthy of special notice is that of Gaupp in his Anatomie des Frosches. Since his few paragraphs sum up the state of our knowledge on the subject, while at the same time furnishing an illustration of the assumption of hearing on the basis of analogy, I present the substance of them in free and slightly abbreviated translation.

"The labyrinth organ has an acoustic and non-acoustic (static) function. For these two functions, according to the leading if not generally accepted view, entirely different portions of the organ are in question, and since the non-acoustic is attributed to the three Cristae acusticae ampullarum and the three Maculae (M. recessus utriculi, M. sacculi, M. lagenae), there remain for the acoustic function only the Papilla basilaris and the Macula neglecta. It is not certain, however, that the non-acoustic organs do not participate in the acoustic function.

"With regard to the acoustic sense of the frog nothing exact is known. That it exists, and that in good development, is certain. The existence of the drum and columella, and the fact that frogs have a voice are unmistakeable proofs of hearing. The participation of the Papilla basilaris in acoustic functions is rendered certain by comparative anatomical studies: the Papilla basilaris is the nerve end-organ from which, in the mammalia, the undoubtedly acoustic organ of Corti arises. From analogy of structure we may also infer an acoustic function in the Macula neglecta: on this, as on the Papilla basilaris, there is a simple tectorial membrane, and further the Pars neglecta, like the Pars basilaris, has a strong thick wall which only in a limited region, namely, where it approaches a part of the perilymphatic space, is markedly thinner." (For fish Breuer (1891) has already stated that if they really hear--which is proved--the Macula neglecta alone can come into consideration in connection with the function, for there is no Papilla basilaris in fishes, and the six other nerve end-organs apparently serve the non-acoustic function.)

As the green frog does not respond visibly to sounds under experimental conditions, I found it necessary to employ indirect methods in the study of audition. By observing the influence of sounds on respiration and on the reactions to certain electrical, tactual, and visual stimuli, I obtained results which, since they have already been described in detail elsewhere, may be summarized here as follows:

1. Observation of frogs in their natural habitat shows that they are stimulated by sounds, but the sense of hearing apparently serves rather as a warning sense which modifies reactions to other simultaneous or succeeding stimuli than as a control for definite auditory motor reactions.

2. Experimental tests prove that sounds modify the frog's reactions to visual and tactual stimuli. When the sound accompanies the visual or tactual stimulus it serves to reënforce the reaction to the other stimulus, but when given alone it never causes a motor reaction.

3. The green frog responds to sounds made in the air, whether the tympana be in the air or in water. There is some evidence that the influence of auditory stimuli is most marked when the drum is half-submerged in water. The influence of sounds upon tactual reactions is evident when the frog is submerged in water to a depth of 4 cm.

4. Sounds varying in pitch from those of 50 to 10,000 vibrations per second affect the frog. The most striking results were obtained by the use of an electric bell with a metal gong. With this sound in connection with a weak tactual stimulus a maximum reaction may often be obtained even when either stimulus alone causes no perceivable reaction.

5. Sounds modify the reactions of the frog after tympana and columellæ are removed. Cutting of the eighth cranial nerves causes disappearance of the influence of sound. It is clear, then, that the reactions to sounds are really auditory reactions and that the sense of hearing in the frog is fairly well developed, although there is little evidence of such a sense in the motor reactions of the animal.

6. Experiments during the spring months show marked influence of sounds for both males and females, whereas experiments made during the winter indicate a much diminished sensitiveness to auditory stimuli in both sexes, but especially in the male.

III. THE MUTUAL RELATIONS OF STIMULI

In studying the various influences of complication of stimuli in the frog, I have used two methods: the measurement of reaction-time and of the amount of reaction. The reaction-time results will be presented first.

Reaction-time to electric stimulation of the skin was studied with special attention to the influence of other stimuli which were given in definite temporal relation to the electric stimulus. A Hipp chronoscope, controlled by a Cattell's falling screen, served as a time-measuring apparatus. The other essentials of the apparatus were a reaction-box, and devices for giving the stimuli and indicating the reaction. On the bottom of the reaction-box a series of wires were so placed that an electric stimulus could be given to the frog resting upon them by the closing of a key in the hands of the experimenter. In preparation for each experiment the frog was placed upon these open circuit wires in such a position that the weight of its body pressed upon a delicate spring in the floor of the box, thus causing the chronoscope circuit to be completed. The forward jump of the frog in response to stimulation caused the breaking of this circuit by the release of the spring upon which the animal rested. When all was in readiness for an experiment the chronoscope was started, and a key closed which simultaneously gave an electric stimulus to the frog and completed a circuit which caused the chronoscope record to begin. The stimulus consisted of a current from one or more "Mesco" dry cells. The motor reaction of the frog broke the chronoscope circuit, thus causing the chronoscope record to stop. It was then possible for the experimenter to read from the dials of the chronoscope the time, in thousandths of seconds, intervening between stimulus and reaction (reaction-time). In case of additional stimuli in connection with the electric, various simple devices were introduced to meet the demands of the experiments. These will be described in connection with the statement of results in each case.

Electric and photic stimuli. A photic stimulus was given from one to two seconds before the electric stimulus by the turning on of a sixteen-candle-power incandescent light, which was placed thirty cm. in front of the frog in the case of one series of experiments and fifteen cm. above it in another. The light uniformly inhibited reaction to the electric stimulus, as is shown by the results of Table 1.

TABLE 1

Title of investigation,--Electric-Visual (Red Light). Experimented on,--Green Frog No. 4. Harvard Psychological Laboratory,--9.40 A.M., Feb. 28, 1902. Chronoscope control average, 189σ,--Electric stimulus, 1 Cell.

NO LIGHT.

Number of Experiment. Reaction-time. 1 152σ 2 145 3 221 4 327 5 263 6 271 7 329 8 215 9 225 10 216

LIGHT BEFORE ELECTRIC STIM.

11 No reaction. 12 No reaction. 13 No reaction. 14 No reaction. 15 No reaction.

NO LIGHT.

16 216

The inhibitory influence of light depends upon the intensity of the electric stimulus. Even a very strong light will not cause much retardation of reaction to a three or four cell current. As the strength of the electric stimulus decreases the delay of reaction increases, until finally there is complete inhibition. At this point, an electric stimulus, to which the frog would react almost invariably when there is no disturbing condition, will fail to cause reaction in the presence of a sudden increase in light intensity.

Merzbacher states that the leg reflex of a frog, so placed that its legs hang free in the air, is greater in response to a given cutaneous stimulus in darkness than in daylight.

Electric and visual stimuli (moving object). For the purpose of determining the effect upon reaction-time to an electric stimulus of stimulation of the eye by a rapidly moving object, experiments were made in which, as in the case of electric and photic stimuli, reactions to electric stimulus alone and to the visual and electric were observed alternately. Thus in the case of each pair of reactions it was possible to note whether the visual stimulus shortened or lengthened the reaction-time. The visual stimulus was given by quickly moving a finger before a window in the reaction-box.

Two series of twenty pairs of reactions each were taken with each of two frogs. In the first series the finger was suddenly moved across the window and the electric stimulus was given either simultaneously or a small fraction of a second later. It was impossible to arrange for accurate measurement of the temporal relations of the two stimuli in the case of these tests. In the second series the finger was moved back and forth before the opening in the reaction-box for an interval of at least a second before the electric stimulus was given.

These experiments, which were in the nature of preliminary tests, yielded the following results. When the stimuli were given almost simultaneously the visual reënforced the electric as was indicated by a shortening of the reaction-time. As appears in the upper part of Table 2, the average time of forty reactions, twenty for each frog, to the electric stimulus was 148^{σ}, and to the same stimulus when it followed the visual 128^{σ}. Furthermore, examination of the several pairs of reactions shows, as is indicated in the table, that there were twenty-seven cases in which the visual stimulus caused shortening of the reaction-time (reënforcement of the electric stimulus) to thirteen in which it caused lengthening (inhibition). When the visual stimulus preceded the electric by at least a second, the reaction-time to the electric stimulus was greatly lengthened. The averages are 150^{σ} for the electric stimulus alone, 178^{σ} when it was preceded by the visual. In this series there are twenty-five cases of inhibition to fourteen of reënforcement.

Electric and visual stimuli (moving red disc). The indications of the importance of the temporal relations of stimuli, so far as reaction-time results are concerned, furnished by these crude preliminary observations led to a more accurate study of the subject. A revolving disc, which moved at the rate of one revolution per minute, was so arranged that at a certain point it closed an electric circuit in which a magnet had been placed. This magnet attracted a steel arm at the end of which a disc of red cardboard 12 mm. in diameter was suspended. With the making of the circuit the steel arm was drawn downward suddenly and the red disc, by reason of the vibrations of the arm moved rapidly back and forth in front of a window in the reaction-box. In this way the moving object was exposed to view about ten cm. to the right and three cm. in front of the right eye of the frog. The revolving disc, a fraction of a second later, completed the electric stimulus circuit. Thus both stimuli were given automatically, at such an interval apart as the experimenter desired. In the two series of results now to be described the intervals were 0.1 and 0.5 second respectively.

TABLE 2

Reaction-time to Electric Stimulation Alone, and to the Same when preceded for 0.1, 0.5, or 1.0 Second by Visual Stimulus.

Key: 1 = Frog. 2 = Electric Alone. 3 = Visual 0.1" before elect. 4 = Number Inhibited. 5 = Number Reënforced. 6 = Number Equal. 7 = Visual 1.0" before elect.

1 2 3 4 5 6 2 7 4 5 6

Preliminary Series. Visual Stimulus Moving Finger. Averages for 20 reactions. No. 5. 179ˢ 158ˢ 6 14 0 163ˢ 206ˢ 14 6 0 No. 6. 116 98 7 13 0 136 150 11 8 1 Gen. Aver. 148 128 13 27 0 150 178 25 14 1

Visual Stimulus Moving Red Disc. Visual 0.1" before electric. Visual 0.5" before electric.

Series I. Averages for 25 reactions.

No. 5. 177 163 10 15 0 170 255 15 9 1 No. 6. 148 112 6 19 0 115 178 18 7 0

Series II. Averages for 25 reactions.

No. 5. 135 120 7 18 0 155 259 24 1 0 No. 6. 128 111 6 19 0 132 227 17 7 1 Gen. Aver. 147 126 29 71 0 143 230 74 24 2

These series consisted of twenty-five pairs of reactions each, with two animals. The results of the series are presented separately, in the lower half of Table 2, because the experiments which constitute them were separated by a period of three weeks. It is to be noted that these results agree fully with those of the preliminary series. The visual stimulus of a moving red disc, given 0.1 second before a 2 cell electric stimulus, reënforces the electric reaction, i. e., it shortens the time of reaction. The same visual stimulus given 0.5 second before tends to inhibit the electric reaction, i. e., it lengthens the time of reaction.

Tactual and auditory stimuli. Since in the frog auditory stimuli under experimental conditions seldom if ever cause visible motor reactions, the study of the influence of this mode of stimulation upon the reactions to other simultaneous or succeeding stimuli is of special interest. In the investigation of the relations of auditory stimulation to other forms of reaction amount of reaction instead of reaction-time was taken as a measure of the influence of the stimulus. By a method the details of which may be most easily understood by reference to the plan of the apparatus in Figure 1, the influence of auditory stimuli on the leg-movement induced by tactual stimulation was observed.

In these experiments the frog sat astride a wooden support, held in position by linen bands over the back and a wire screen cap over the head. The hind legs hung free, and any movement of one of them in response to a stimulus could be read in millimetres by reference to a scale on the wooden support. This method of measuring the value of a stimulus in terms of leg-reflex has been used by several investigators--most recently by Merzbacher. I have found it desirable, as did Merzbacher, to observe the movements of a shadow of the leg on the scale and thus read the amount of movement, rather than to watch the leg itself and attempt to project it upon the scale.

As is indicated in Fig. 1, the auditory and tactual stimuli were given automatically by means of a swinging pendulum, P, which was held in position by the magnet a until released by the experimenter. Early in its swing the pendulum turned the key, m, thus completing a circuit which caused the auditory stimulus to be given; later in the swing the key, n, was turned, and the tactual stimulus thus given through the magnetic release of the lever, l. The interval between the auditory and the tactual stimuli could be varied from 0 to 2" by changing the position of the key, n. For intervals over 1" it was necessary to arrange this key so that the tactual stimulus was given at some time during the return swing of the pendulum.

The auditory stimulus used was either the sound of a quick hammer blow (momentary stimulus of Series I), or the ringing of an electric bell for a certain length of time (prolonged stimulus of Series II). In Fig. 1 the bell is shown. It was placed eighty cm. from the frog, and in order that the influence of vibration of the experiment table might be avoided it was suspended from the pendulum frame. When the hammer was used it was placed sixty cm. from the frog, on the pendulum table. The holder for the frog and the tactual apparatus occupied a separate table which was not disturbed by the jars of the pendulum table.

The tactual stimulus was given by a rubber cone, T, two mm. in diameter at its apex. This rubber point, after the electric release of the lever to which it was attached, struck the frog at the middle point of a line drawn between the posterior margins of the tympana. The intensity of the stimulus could be varied by weighting the lever, l, at w.

All experiments were made with the green frog, Rana clamata Daudin. The reactions were taken regularly at half-minute intervals in pairs: first, a tactual stimulus reaction, then an auditory-tactual reaction. Ten, fifty, or one hundred pairs constituted a series. So far as the condition of the frog is concerned there seems to be nothing undesirable in long series, for fatigue does not appear, and so long as the animal is kept moist and in an unconstrained position, it continues to react normally, and without frequent struggles to escape. The advantage for the purposes of this investigation of taking the reactions in pairs, rather than taking separate series of reactions for each stimulus or combination of stimuli, is obvious. It enables us to compare directly the reactions of each pair, in other words those reactions which took place under most nearly identical conditions, and to note at once whether the auditory stimulus reënforced or inhibited the tactual reaction.

During a series the intensity of the tactual stimulus was changed as conditions demanded, but for any one pair of reactions it was always the same. It not infrequently happened that an intensity which at first caused merely a slight movement of the leg, later in the series uniformly brought about a maximal contraction, or the reverse might be true, and inasmuch as a maximal reaction to the tactual stimulus alone left no opportunity for judging of the influence of the auditory stimulus, when it was given in addition to the tactual, it was always necessary in such cases so to alter the intensity of the tactual stimulus that a medium reaction resulted.

The frogs, after being placed in the saddle-like holder and held firmly for a few seconds, seldom struggled very much, but if bound tightly they became irresponsive to the stimuli. It was, therefore, necessary after they had quieted down to loosen the bands which held them in position. For the purpose of excluding the influence of visual stimuli a wire screen cap covered with black cloth was put over the head; this served to keep the animal in position as well as to exclude visual stimulation.

a. Momentary auditory stimulation. Four frogs were used for a study of the influence of the momentary sound produced by a hammer blow, and for each of these animals fifty pairs of reactions were recorded in series each day. The temporal relation of the stimuli was changed daily during a week of experimentation: the results therefore consist of fifty pairs of reactions with each frog for each of the following seven intervals: (1) Auditory and tactual stimuli simultaneous, (2) auditory .25" before tactual, (3) auditory .45" before, (4) auditory .15" before, (5) auditory .65" before, (6) auditory .35" before, (7) auditory .90" before. The intervals were used in the experiments in the above order to avoid the formation of the definite habits of reaction which regular increase in the interval would have favored.

Typical of the results with all the animals are the following (Table 3) which were obtained with No. 1, a male. The figures in each case indicate the average of fifty reactions. Reënforcement and inhibition are expressed in terms of the tactual reaction, i. e., the auditory-tactual reaction is so many per cent greater (reënforcement) or less (inhibition) than the tactual. In the tables reënforcement is indicated by the + sign; inhibition by the - sign. In the last column of the table is given the number of reactions that were reënforced or inhibited. This was determined by comparing directly the reactions of each pair. Cases in which the two reactions were the same were distributed equally between the two classes: tactual reactions reënforced by auditory stimulus, and tactual reactions inhibited by auditory stimulus. Assuming that the auditory stimulus was without effect upon the tactual reaction, the number of reactions in these two classes would be approximately the same, hence all auditory-tactual reactions over half in a series, i. e., over twenty-five, which are greater than the corresponding tactual reactions, are reënforced reactions, and can be taken as a measure of the reënforcing influence of the auditory stimulus. In the same manner all reactions over half which show inhibition can be taken as a measure of the inhibitory value of the auditory stimulus.

As preliminary tests described in an earlier paper furnished evidence of sex-differences, it is worth while to compare the results given by the males and females in these experiments with momentary auditory stimulation. For purposes of comparison I have presented in Table 4 the reënforcement-inhibition values given by the males and females for each interval. Column one contains the value of the auditory-tactual reaction in terms of the tactual reaction; column two, the number of reactions in excess of half which were reënforced or inhibited.

TABLE 3. FROG NO. 1. MOMENTARY AUDITORY STIMULUS, HAMMER BLOW. WEIGHT USUALLY 5 OR 10 GRAMS

Reaction Reaction Amount of Number of to to Auditory Reënforc'm't reactions Tactual and Tactual or Reënforced Interval. Stim. Stim. Inhibition. or Inhibited.

0" 6.84mm. 11.08mm. +62.0% +17.0 .15 22.22 28.96 +30.3 +17.0 .25 16.30 21.72 +33.3 +13.0 .35 24.90 25.32 + 1.7 + 0.5 .45 17.56 13.64 -22.3 -10.0 .65 17.46 15.72 -10.0 - 6.0 .90 31.26 31.48 + 0.7 + 0.5

TABLE 4. MOMENTARY AUDITORY STIMULUS, HAMMER BLOW

Males Females Nos. 1 and 3. Nos. 2 and 4. Per centum No. of Per centum No. of Interval Diff. Reacts. Diff. Reacts. 0" +82.5% (Reënf't) +17.5 +58.0% +12.7 .15 +58.1 +17.0 +25.4 + 8.5 .25 +32.3 +12.7 +39.8 +12.7 .35 + 4.0 + 1.2 - 9.7 - 3.2 .45 -13.5 (Inhibition) - 7.2 -13.9 - 7.2 .65 -12.5 - 6.2 -11.8 - 7.2 .90 - 0.7 - 1.5 - 2.6 - 0.5

In these results two striking differences between the males and females appear: first, the reënforcement is not so great for the females as for the males; second, inhibition appears earlier and continues longer with the females than with the males. The average reënforcement with simultaneous stimuli is 82.5% for the males against 58.0% for the females. Inhibition begins to appear in case of the females when the interval between the stimuli is .25" to .35"; in case of the males it appears between .35" and .45". Finally at .90" interval inhibition is slightly greater for the females.

Although the exact significance of these facts is unknown, it is not improbable that they are indicative of fundamentally important sex-differences in reaction to sound. The males among frogs are usually the vocalists, although in some species the females also croak. Moreover, in case of the green frog the tympanum of the male is much larger than that of the female. The results presented would seem to indicate that certain sounds stimulate the males to activity, whereas they inhibit activity in the females.

Graphically represented, the results of the momentary auditory stimulus experiments with frogs Nos. 1, 2, 3, and 4 are as follows:

The curves are all plotted by the method which will now be described in connection with Fig. 2. This figure presents the reënforcement-inhibition curves for the males No. 1 (solid line in the figure) and No. 3 (broken line). If in this figure we let the zero-point on the ordinates represent the value of the reaction to the tactual stimulus when given alone, then the value of the reaction to the auditory-tactual stimuli would be represented at some point above the zero-point if this reaction was greater than the tactual reaction (reënforcement), and below the zero-point if the reaction was less than the tactual (inhibition). Since one of our chosen measures of reënforcement and inhibition is the amount, in per cent of tactual reaction, by which the auditory-tactual reaction exceeds or falls short of the tactual reaction, such a curve of reënforcement-inhibition as that of Fig. 2 (solid line) can be constructed at once from the data given in column four of Table 3. Here the auditory stimulus, when simultaneous with the tactual, caused 62% reënforcement, as is indicated in the figure. The figures in the left-hand margin of the curves indicate amount of reënforcement or inhibition in per cent of tactual reaction; those at the bottom of the curves mark the intervals. On the curves dots indicate the intervals used in the experiments. Each of the curves is plotted on the basis of 700 reactions.

In every way comparable with the curves for the males No. 1 and No. 3 in Fig. 2 are those for the females No. 2 and No. 4 of Fig. 3. The similarity of the two curves in each figure is noteworthy. Inasmuch as the conditions of experimentation were the same for all the animals this would seem to indicate sex-differences which are worthy of further investigation. The curves show clearly the greater reënforcement in the males, and the greater inhibition in the females.

Figures 4 and 5 are the reënforcement-inhibition curves for the same series of experiments plotted on the basis of the number of reactions in excess of half that were reënforced or inhibited. As there were fifty pairs of reactions with each frog for each interval, uniform reënforcement would be represented by twenty-five reactions above the base-line; uniform inhibition by twenty-five reactions below the base-line. The number of reactions is indicated by the figures in the left margin; the intervals, by those below the base-line. As an illustration of the application of the method of plotting, the curve for male No. 1 (solid line) of Fig. 4 is constructed from the data of column five of Table 3. With simultaneous stimuli 17 reactions in excess of half, i. e., 17 + 25, or 42, were reënforced; at .35" interval .5 of a reaction was the average amount of reënforcement; at .45" interval 10 reactions in excess of half, i. e., 35, were inhibited, therefore the curve falls to 10 below the base-line.

Just as Figures 2 and 3 permit of direct comparison of the results of the measurement of the amount of reënforcement and inhibition for males and females, so Figures 4 and 5 make possible comparison in similar fashion of the number of reënforced and inhibited reactions for the sexes. It is to be noted that the two sets of curves, plotted on the bases of amount and number of reaction, agree in all important respects.

Figure 6 is the composite curve of amount of reënforcement-inhibition for the four animals; Figure 7 is the composite curve of the number of reactions reënforced and inhibited.

Summarily stated, the results of the experiments thus far described are: (1) The auditory stimulus of a quick hammer blow produces the maximum amount of reënforcement of tactual reaction when it is given simultaneously with the tactual stimulus; (2) as the interval between the auditory and the tactual stimulus approaches .35″ the amount of reënforcement gradually decreases; (3) when given .35″ before the tactual stimulus the auditory is practically without effect upon the tactual reaction; (4) as the interval increases above .35″ inhibition begins to appear; (5) the inhibitory influence of the auditory stimulus is greatest when the interval is about .45″; (6) when the interval is as long as .90″ the auditory stimulus is again ineffective. It thus appears that the reënforcement-inhibition curve of this particular stimulus under the conditions described is representative of a neural process which completes itself, in passing through two phases, a positive phase (reënforcement) and a negative phase (inhibition), in about one second.

b. Prolonged auditory stimulation. The experiments previously described have proved that a momentary auditory stimulus, which when given alone never produces a visible motor reaction, either reënforces or inhibits the reaction to a tactual stimulus which it accompanies or precedes. The experiments now to be described were made for the purpose of ascertaining whether reënforcement and inhibition occur in the same way if the auditory stimulus is prolonged, instead of momentary.

In a trial series of experiments with frog No. 1, one hundred pairs of reactions were recorded for each of six intervals of auditory stimulation. The auditory stimulus was given by the ringing of an electric bell. For all intervals the ringing of the bell continued until the tactual stimulus was given. When the two stimuli were given simultaneously the auditory stimulus was necessarily momentary, as in the foregoing experiments, but for all other relationships of the stimuli the bell rang for a certain length of time before the tactual stimulus was given. The six relations of the stimuli were: (1) simultaneous, (2) bell .2″ before and until tactual, (3) bell .6″ before, (4) bell 1.05″ before, (5) bell 1.5″ before, and (6) bell 2.0″ before. The other conditions of these experiments were the same as those previously described, except that the auditory stimulus was here given by the opening of the key which released the pendulum, instead of being given by the turning of a key in the course of the pendulum swing. This method of giving the auditory stimulus as the pendulum was released was found unsatisfactory because of the irregularity of the magnetic release; at one time the pendulum would start immediately, at another time there would be a delay of as much as .1″.

The reënforcement-inhibition curve plotted on the basis of the 1200 reactions in this series is presented in Fig. 8. Before stopping to consider the important features of this curve we should note the results of certain more accurate experiments with prolonged auditory stimulation.

With two animals, No. 2, a female, and No. 3, a male, fifty pairs of reactions were taken for nine different intervals (see Table 5) of auditory stimulation. Each of the curves of Figures 9 and 10 is therefore based upon 900 reactions. The conditions for these experiments were the same as those for the momentary stimulation series, save that the electric bell took the place of the electrically actuated hammer, as the mechanism for auditory stimulation.

The important facts exhibited by the results of these prolonged auditory stimulation experiments in contrast with those with momentary auditory stimulation are: (1) That whereas for the momentary auditory stimulus of a hammer blow the reënforcement is greatest for simultaneous stimuli, in case of the prolonged stimulation with the electric bell, reënforcement increases during an interval of .25″ of auditory stimulation. Hence, the two conditions of stimulation give us different types of reënforcement-inhibition curve. For the momentary stimulus the maximum reënforcement appears at simultaneity, and for the prolonged stimulus at .25″; (2) that the transition from reënforcement to inhibition occurs at 1.2″ in the prolonged stimulation curves, while in the momentary stimulation curves it occurs at .35″; (3) that the maximum inhibition which appears in the curves under discussion at about 1.5″ is less in comparison with the amount of reënforcement than that of the momentary stimulation curves; (4) that the auditory stimulus becomes ineffective when the interval during which it continues before tactual stimulation is 2.0″. The curves of Figures 8, 9, and 10 are then representations of a neural process which passes through a positive and a negative phase in about 2″. The effect of prolongation of the auditory stimulation interval is to lengthen the period of reënforcement; the period of inhibition shows little modification.

For the purpose of showing in greater detail the nature of the results of this work the data from which the curves of Figures 9 and 10 were constructed are presented in the accompanying Table 5.

Having now presented the results of my own investigation I wish to call attention to certain of their relationships to the work of other investigators, and to discuss briefly their significance.

TABLE 5. PROLONGED AUDITORY STIMULATION (ELECTRIC BELL)

Frog No. 2. Female. Weight usually 25 grams.

Auditory Number of and Amount of Reactions, Tactual Tactual Reënforcement Reënforced Interval. Stimulation. Stimulation. or Inhibition. or Inhibited.

0″ 9.20 mm. 12.12 mm. + 31.7% +10.0 .25 4.56 11.88 +160.5 +22.5 .45 8.94 16.94 + 89.5 +18.5 .65 17.18 22.50 + 31.0 +15.5 .90 9.42 13.32 + 41.4 +14.5 1.20 10.54 9.64 - 8.5 - 2.5 1.40 24.00 20.64 - 14.0 - 6.0 1.58 19.16 17.80 - 7.1 - 7.0 1.95 14.50 15.40 + 6.2 + 5.5

Frog No. 3. Male. Weight usually 5 or 10 grams.

0″ 14.92 mm. 25.20 mm. + 68.9% +11.0 .25 15.88 38.54 +142.7 +24.0 .45 13.48 26.02 + 92.9 +13.5 .65 18.30 27.94 + 52.6 +13.0 .90 20.94 29.06 + 38.8 + 9.5 1.20 21.90 30.58 + 39.6 + 1.0 1.40 19.18 18.34 - 4.4 - 5.5 1.58 32.24 26.30 - 18.1 - 3.0 1.95 13.86 14.14 + 2.0 + 2.0

VI. DISCUSSION OF LITERATURE AND RESULTS

The literature on reënforcement and inhibition is large, and even that portion of it which deals especially with the importance of the temporal relations of stimuli in connection with reënforcement and inhibition is so extensive that it does not seem worth while to attempt to give a systematic résumé of it for the purposes of this paper. I shall therefore call attention merely to those investigations which have contributed directly to the solution of the problems with which we are now concerned.

Bowditch and Warren discovered that knee-jerk in the human subject is reënforced when an auditory, a visual, or a tactual stimulus precedes the tendon blow by .1″ to .5″, whereas the same stimuli have an inhibitory influence when they are given from .5″ to 1.0″ before the tendon blow.

At the suggestion of Bowditch, Cleghorn undertook to investigate the influence of complication of stimuli upon voluntary movements. In this research graphic records taken in connection with an ergograph indicated (1) that "a sensory stimulus" applied just as the muscle was beginning to contract (voluntarily) caused an increase in the height of the contraction, and (2) that the relaxation following a contraction with intercalated sensory stimulus is quicker and more complete than when no stimulus is given (p. 344). Cleghorn did not give special attention to the significance of the temporal relations of the stimuli which he employed, and his work was limited to the phenomenon of reënforcement of voluntary action by reason of the appearance, during the progress of his research, of an excellent paper on the interference of stimuli by Hofbauer.

Hofbauer covered thoroughly the ground which Cleghorn had planned to work over. The ergographic method was employed also by Hofbauer in his very careful study of the interference of impulses in the central nervous system of man. It was noticed that while the subject was rhythmically contracting a certain group of muscles in response to some prearranged signal (e.g., the sound of a metronome) the report of a pistol caused the contraction which immediately followed it to be much greater than the average of the rhythmic series, while the next contraction was correspondingly less than the average. It thus appeared that the sudden sound caused, first, reënforcement of the voluntary movement, then, inhibition. The reënforcement is greatest, according to Hofbauer, when the voluntary movement occurs immediately after the pistol report. When the report precedes the metronome signal by .2″ reënforcement is still marked, but thereafter it decreases rapidly in amount, until finally at .5″ inhibition appears. When the interval between the two stimuli is 1.0″ the first stimulus has practically no effect upon the voluntary movement in response to the second. (Hofbauer, p. 558.)

What Bowditch and Warren, not to mention other students of the subject, have described for reflex action in man, Hofbauer, Cleghorn, and others have shown to hold true also of voluntary movements. Unfortunately my own investigation was completed up to the point of the writing of this paper before I read Hofbauer's work, so I have not followed methods of dealing with my data which would make our results directly and easily comparable. But, whatever may be the relations of our results in detail, there can be no doubt that what he has demonstrated for man is true in its important aspect of the reënforcement-inhibition phenomena for the frog.

Important in their bearings upon the phenomena of reënforcement and inhibition which we are now considering, are the various studies of refractory period and rhythm of nerve cell and fibre. The existence of a refractory period in neural substance, similar to that demonstrated for certain kinds of muscle by Marey, Englemann, Kaiser, Cushny and Matthews, Woodworth, and many others, has been proved by Broca and Richet.

Broca and Richet found that in the normal dog the refractory period of the nerve substance is too short to be easily detectable, they therefore experimented with animals which were lightly chloralized and kept at a temperature of 30 to 34° (the mean normal temperature of the dog is about 39.5°). Under these conditions a dog, when two identical stimuli (quality and intensity the same) were applied to the cerebral cortex successively, exhibited the following reactions: (1) When the stimuli were separated by .01″ they reënforced one another (addition); (2) when the interval was .1″ they inhibited the reaction partially (subtraction).

Concerning this phenomenon Richet writes in his dictionary of physiology (p.5): "Marey showed, in 1890, that the heart of the frog, at certain moments of systole, was inexcitable. Now our experiments prove that the cerebral apparatus, a certain time after the excitation, also ceases to be excitable: it then has a refractory phase, and this refractory phase is much more prolonged than that of the cardiac muscle." In a later publication Richet makes the somewhat startling statement that a refractory period is not exhibited by the nerves of cold-blooded animals. In the tortoise, according to his results, reënforcement occurs so long as the interval between the two stimuli is not greater than 2″, while for longer intervals each stimulus to all appearances works independently. Richet seems to have generalized from a study of the tortoise. That his generalization is unwarranted seems to me highly probable in the light of the results of this paper, for there are many reasons for supposing that the reënforcement-inhibition phenomena with which we have been dealing in case of the frog are manifestations of the existence of the same process in the nervous system which under somewhat different conditions of experimentation exhibits itself in the so-called refractory period.

The researches of Richet and his students indicate that the time of the process which conditions the phenomena of reënforcement-inhibition is about .1″. Stimuli given at .1″ intervals do not interfere with one another. That the process underlying the refractory period and the reënforcement-inhibition phenomena of our experiments is a rhythmic double-phase process is made still more probable by the following results. Horsley and Schäfer found that the rate of response of the monkey to cortical stimulation was 12 per second, and Schäfer discovered that the maximum rate of volitional impulses in man is 10 to 12 per second.

It was shown by Exner that certain movements of the foot of a rabbit could be produced by stimulating either the cortex or the skin of the foot. Simultaneous stimulation of both regions gives reënforcement. Stimulation of the cortex, if given not more than 3″ before subliminal stimulation of the skin, renders the latter effective. When both stimuli are subliminal each makes the subsequent one effective if the interval between them is not over 1/8″ (Schäfer). Similarly for the dog Exner proved that cortical and cutaneous stimuli reënforced one another, when both were subliminal, if the interval between them was not greater than .6″. Cortical and auditory stimuli, and auditory and cutaneous (of the skin of foot) gave similar results.

Physiologists have long been familiar with several aspects of the phenomena of reënforcement and inhibition in the frog, but I know of no detailed study of the significance of the temporal relations of stimuli in this connection. Goltz called attention to the inhibition of the croaking reflex by peripheral stimulation, as well as to several similar phenomena. Nothnagel, Lewisson, and Wydensky further contributed to our knowledge of the interference effects of stimuli in the frog. Wydensky proved that the application of an induced current to a nerve-muscle preparation may result in either contraction or relaxation of the muscle, according to the frequency of stimulation.

More recently Merzbacher has dealt with the influences of complication of stimuli in the frog with the purpose of ascertaining the relations of the sense-organs to the reflex movements of the animal. His first paper is concerned especially with the functional importance of the eye in connection with reflexes. Unfortunately for the demands of this research, he did not attend particularly to the temporal relations of his stimuli. That a visual and a cutaneous stimulus were given either "at the same time or within a short interval of one another" (p. 250) is not the sort of information our problems demand.

According to Merzbacher's very interesting results a visual stimulus reënforces the reaction to a cutaneous stimulus. As the results of this paper show, this is only half a truth, for the two stimuli may either reënforce or inhibit one another's reactions. As Merzbacher observed no evidences of reaction to auditory stimulation he presumably did not attempt to study the influences of the ear in connection with reflexes.

There can be no doubt that the words reënforcement and inhibition as at present used in connection with the functions of the nervous system cover a multitude of widely differing phenomena. We can at once distinguish at least two important kinds of reënforcement or inhibition: first, that which is due to the functioning of special augmentary or inhibitory portions of the nervous system; second, that which is the result of the complication of stimuli. Any and every process in the nervous system may have either a reënforcing or an inhibiting influence upon simultaneous or succeeding processes; doubtless most processes or impulses at various times have both effects. The nervous system is constantly being modified by impulses from many sources, which suppress or strengthen one another according to their relative intensity, their temporal relations, and the motor relations of the portions of the organism which they affect.

The existence of the so-called refractory period in brain cortex and nerve indicates that every stimulus causes certain fundamentally important changes in the condition of the neural substance. These changes we may for convenience of illustration describe as modification of excitability, or of the functional capacity of central or peripheral tissues. Every stimulus causes a portion of the neural substance to pass from its normal state through a condition of increased excitability, which we may designate the positive phase, to a condition of diminished excitability, the negative phase. There is first an increase in the functional capacity of the tissues, then a decrease. If during the course of the change produced by a given stimulus a second stimulus becomes effective its result in reaction is determined by the particular phase of the tissues upon which it intrudes. If the nervous system is in the condition of increased excitability, and the two stimuli act upon sensory regions whose motor connections are not antagonistic, the reaction will be reënforced, as we say, by the previous stimulus; if, however, the second stimulus falls upon the negative phase of the nerve substance, the reaction will be partially or totally inhibited.

The facts which are most prominent as the result of this investigation are, first, that the temporal relation of stimuli is an important condition of certain forms of reënforcement and inhibition; second, that the interference effects of two stimuli cannot be studied to advantage without attention to the relations of the forms of reaction which are appropriate to each stimulus.

V. SUMMARY

1. Motor reactions of the green frog to electric stimuli are inhibited either partially or wholly by photic stimuli. The visual stimulus of a moving object has a like effect. It has been found, furthermore, that the same visual stimulus may either inhibit or reënforce the motor reaction in response to electric stimulation. When the two stimuli are given simultaneously reënforcement occurs, when the visual stimulus precedes the electric by half a second or more inhibition appears.

2. An auditory stimulus, which does not produce any visible reaction when given alone, modifies respiration and the reactions to other stimuli when given in connection with them.

3. The momentary auditory stimulus of a quick hammer blow when simultaneous with tactual stimulation reënforces the reaction to the latter stimulus. This reënforcement, or increase in the amount of reaction, ranges from 50 to 100% of the average reaction to the tactual stimulus alone. When the auditory stimulus is given before the tactual reënforcement occurs in gradually decreasing amount until the interval between the two stimuli reaches .35″; at this point the auditory stimulus has no apparent effect upon the tactual reaction. As the interval is still further increased inhibition appears and continues for intervals between .35″ and .9″. Reënforcement is greatest when the two stimuli are simultaneous; inhibition is greatest when the momentary auditory stimulus precedes the tactual by .4″ to .6″. When the interval reaches .9″ the first stimulus does not affect the reaction to the second.

4. Reënforcement is greater for the males than for the females; inhibition appears sooner and lasts longer in case of the females. This apparently indicates that the males are stimulated to activity by certain auditory stimuli, whereas the females are rendered passive by similar sounds.

5. Prolonged auditory stimulation by means of an electric bell causes reënforcement and inhibition, according to the temporal relations of the stimuli, as does momentary auditory stimulation, with the following differences: The maximum reënforcement occurs when the tactual stimulus is given about .25″ after auditory stimulation has begun; reënforcement continues for a period of 1.2″, i. e., when the electric bell continues to ring until the tactual stimulus is given, it reënforces the tactual reaction from simultaneity to 1.2″. Inhibition then appears, and continues until 1.8″. Both momentary and prolonged auditory stimulation cause first reënforcement, then inhibition of the appropriate reaction to a tactual stimulus.

6. The reënforcement-inhibition curves for the frog are very similar to those for man.

7. In case of the several pairs of stimuli whose interference effects have been studied reënforcement-inhibition appears. The first stimulus reënforces reaction to the second so long as the interval between them is not more than about .4″, while it inhibits the reaction when the interval is longer. Whether this reënforcement-inhibition curve as given in the experiments described may similarly be obtained for any and every pair of stimuli, no matter what their relation to reactions, remains to be determined.

8. In connection with the study of the mutual relations of stimuli of which this paper gives an account certain facts concerning the sense of hearing have been discovered. A summary statement of the results on hearing may be found on page 551.

FOOTNOTES:

THE TEMPORAL RELATIONS OF NEURAL PROCESSES

BY ROBERT M. YERKES

Muscle contraction-time, according to the determinations of several investigators, varies about .0035". Sanderson states that the time for direct stimulation of the muscle is approximately .0035" and for indirect stimulation, by means of the nerve, .007". The rate of nerve-transmission in the frog ranges from 25 to 35 metres per second.

Reflex reaction-time, as might be expected, varies widely with the nature of the reaction elicited by a stimulus, the condition of the animal, and the quality and strength of the stimulus. For many of the simple motor reactions of the frog it ranges between 20 and 60^{σ}. Whether reflex reaction-time is to be sharply contrasted with instinctive and voluntary reaction-times, or whether they indistinguishably merge into one another is a question of considerable interest and importance for the student of the evolution of activity.

Voluntary reaction-time may be as short as 150^{σ} or as long as life, in an animal capable of profiting by experience as does the frog. It is preëminently the delayed type of reaction-time.

So much concerning the temporal relations of neural processes in the frog being well established, the purpose of the present paper is to call attention to some experimental results which indicate the existence of clearly defined types of reaction, and suggest possible values of reaction-time as a sign of mind.

The specific problems to be considered are: (1) Do reaction-times, in any given animal, range with equal frequency of occurrence from short to long, or are there certain modes (most frequented classes) which indicate definite types of reaction, such, for example, as the reflex, instinctive, etc.? (2) If there is distribution of the reaction-times about one or more modes, what are the types of reaction indicated thereby? (3) Finally, is reaction-time of service as a sign or measure of consciousness?

I wish especially to call attention to the fact that this paper deals with the reactions of the frog, not with animal reactions in general.

REACTIONS TO ELECTRICAL STIMULATION AND TYPES OF REACTION

Two years ago in connection with a discussion of the reaction-time of the green frog to electrical and tactual stimuli, I presented a curve showing the distribution of 277 reaction-times to an electrical stimulus. The curve exhibited two clearly defined modes: one at between 60 and 70^{σ} and the other at about 160^{σ}. There was further a group of delayed reactions ranging about 500^{σ}. This form of distribution was interpreted, at the time, as indicative of three types of reaction, called, respectively, the reflex, the instinctive, and the delayed.

I have since obtained and examined with reference to form of distribution the further data which are presented in this paper. The reactions are all those of the green frog to electrical stimulation. The stimulus was applied by means of wires on the reaction-board on which the frog rested during the experiments. When reaction occurred in response to the electrical stimulus a circuit through the time-measuring apparatus was broken by the release of a delicate spring which had been held in place up to the instant of reaction by the weight of the frog. A Hipp chronoscope, controlled by a Cattell falling screen, served as a time-measuring mechanism. Three intensities of stimulus were used: (1) A current from one Mesco dry cell, (2) from two cells, and (3) from four cells.

Of the reactions whose time was measured there are three series. Series I is constituted by the recorded reaction-times in response to a one-cell stimulus, Series II, those in response to a two-cell stimulus, and Series III, those in response to a four-cell stimulus. The number of reactions, range and mode of each series are as follows:

Number of reactions Range Mode Series I 193 161-798^{σ} 235^{σ} Series II 288 41-647 235 Series III 256 61-178 105

The distribution of the 481 reaction-times of Series I and II is shown by Figure 1; that of the 256 reaction-times of Series III, by Figure 2. For both of these distribution polygons the reaction-times were arranged in ten^{σ} classes, beginning with the class 41-50^{σ} in the case of the combined Series I and II and with the class 61-70^{σ} in the case of Series III.

Series I exhibits a primary mode at 235^{σ}. There are no reflex reactions in this series, unless it be maintained that the reflex reactions of the frog may have a reaction-time of over 160^{σ}, but there are a number of delayed reactions, some of which have reaction-times as long as 798^{σ}. This intensity of stimulation (one cell) may be said to call forth prompt reactions, which we may provisionally call instinctive, and delayed reactions, which have all the appearances of voluntary acts. There are no reactions which come within the range commonly considered as the reflex range of the frog (20-60^{σ}), and there are relatively few delayed reactions: almost all centre about the mode 235^{σ}.

Series II, in contrast with Series I, exhibits a secondary mode at 65^{σ} in addition to the primary mode at 235^{σ}. The stimulus-intensity of this series (roughly twice as great as that for Series I) induces a variety of short reaction, which did not appear in the case of the one-cell stimulus, and at the same time fewer delayed reactions. The range of the reaction-times for the two series is about the same, but the lower limits are markedly different.

Observation of the subjects during the experiments revealed two methods of reaction to the two-cell stimulus: a locomotor reaction (jump) which at once removed the animal from the source of stimulation, and a twitch of the hind legs which was instantly followed by the above-mentioned locomotor reaction. The leg reactions constitute the reflex group of Fig. 1, the usual prompt locomotor reactions, the instinctive group, and the slow locomotor reactions, the delayed or voluntary group.

It is to be noted that the instinctive reaction-time mode is the same for the two intensities of stimulation. This apparently indicates that change in intensity of stimulation causes a change in the type of reaction, not merely a gradual change in the position of the mode. For example, the modal reaction-time of 235^{σ} given by a one-cell stimulus did not shift to 200^{σ} or lower, as might have been expected, but instead there appeared a new type of reaction. The average reaction-times for the two series indicate a decrease in time with increase in intensity of stimulation, but they give no indication of the really important difference in the two series of reactions. The great importance of the distribution of the data, in addition to the common statistical quantities, is manifest.

Series III, whose reactions occurred in response to a very strong stimulus, differs in several important respects from the other series. Its range is much narrower, only 117^{σ}. Delayed reactions are lacking, and so also, curiously enough, are the reflex reactions of Series II. Instead of either or both of the modes of Series II, there appears in Series III an intermediate mode at 105^{σ}.

Our interpretation of these facts is facilitated by results of observation of the reacting subject. The leg reflex which frequently occurred in response to the two-cell stimulus never appeared in response to the four-cell stimulus. This in part explains the lack of the short reaction-time mode of Series II; it does not, however, account for the lack of delayed reactions. The latter fact may be referred to the intensity of the stimulus. Another difficulty in interpretation appears in connection with the intermediate mode, 105^{σ}. Is this to be considered an instinctive mode, as were those at 235^{σ}, or a reflex mode? Where is the line between reflex and instinctive action to be drawn? These results very clearly indicate that no line can be drawn, except quite arbitrarily. Reflex reaction-time, in the case of the frog, is continuous with instinctive, yet for any given situation the reflex, instinctive, and delayed (voluntary?) modes are likely to appear, as, for example, in the case of the data of this paper. Our conclusion must be, therefore, that although types of reaction are indicated by reaction-time results, the mode for a given type varies too much in position with different conditions to make it possible to say that a particular reaction-time is that of a certain type.

We may safely say, then, that for any given subject, the muscle contraction-time, nerve transmission-time, and simple sensory reaction-time to the constant stimulus in question being known, we should be able safely to interpret reaction-time records in terms of reaction types. For reflex, instinctive, and voluntary are terms which designate modes of reaction, albeit not isolated classes, for they intergrade.

Whether there are more types of reaction than are indicated by the data of this report does not concern us at present, for the practical as well as the theoretical bearings of our conclusions depend upon the existence of types, and not upon their number.

REACTION-TIME AS AN INDICATION OF CONSCIOUSNESS

Hesitation in reaction is commonly accepted as an important sign of volitional consciousness in man; consequently delayed reactions in lower animals are supposed to be indicative of psychic processes. Granting this much, reaction-time may be used as a sign of consciousness. It cannot be denied that the longer the reaction-time of a given animal the greater the probability that the reaction is conditioned by mental processes. Such a statement, it is true, has a basis neither better nor worse than that of most of our inferences concerning the nature of the actions of our fellow beings. As I have already attempted to show in a discussion of criteria of consciousness in animal psychology, there is no one criterion of consciousness which can be used alone satisfactorily, but instead there are numerous signs of mind each of which has value according to the number and variety of our observations concerning its occurrence in connection with states of consciousness. The more of such signs we discover and learn to evaluate properly in relation to consciousness in its different grades and to one another, the safer will be our inferences concerning the existence of mental processes in animals.

Reaction-time is presented in this paper as an additional sign of mind. Like all other signs it is of value only if used as one of a series of indications of mental life. For if we attempt to judge of consciousness by reference to reaction-time alone, we may be seriously misled, whereas if we use it in connection with docility, variability, neural specialization, and other recognizedly valuable signs, we may be greatly aided in our inference. As in juristic procedure judgment is not based upon one bit of evidence nor even upon the evidence of a single witness, but upon evidence accumulated from all available sources, so in our attempts to judge of the existence of consciousness, it matters not whether the being be human or infra-human, we should make use of all phenomena which are recognized as signs of mind. The chief task of comparative psychology at present is the discovery and evaluation of signs of mind.

Reaction-time data, however, furnish another sign, or, as I prefer to call it in this case, measure of the intensity of consciousness; for variability of the time of reaction as well as its duration is significant. Reflex reaction-time is relatively constant, instinctive varies considerably, and the variability of voluntary reaction-time is extremely large. Degree of variability of reaction-time may be used as an indication of consciousness in the same way that variability in the form of reaction is used. The higher the power of consciousness the greater the variety in form of reaction and the variability of the reaction-time.

Reaction-time studies, as well as introspection and the investigation of animal behavior, indicate the importance of three activity concepts: automatism, instinct, and will. The automatic act is quick and relatively constant in form as well as reaction-time, while all signs lead us to infer that consciousness, when it accompanies the act, is a sequent phenomenon and not a condition of the act. The instinctive act is both slower and more variable in form and time than the automatic: consciousness is indicated as an accompaniment, and apparently it is at times a condition of the act. The will-act is extremely variable, unique in form, and almost without limits of reaction-time, for the conscious organism may react to the present situation in a fifth of a second, a day, or a year. Will is experience in action: it is our name for individually acquired control, and voluntary action is above all consciously conditioned activity.

Reaction-time, with respect to its two aspects of duration and variability, may be used as a sign or criterion of consciousness, for in accordance with the nature of these two sets of facts we classify acts as reflex, instinctive, or voluntary.

FOOTNOTES:

THE MENTAL LIFE OF THE DOMESTIC PIGEON

AN EXPERIMENTAL STUDY OF CERTAIN EMOTIONAL AND ASSOCIATIVE PROCESSES

BY JOHN E. ROUSE

I. INTRODUCTION

Naturalists have observed the habits of pigeons, and physiologists since Flourens have subjected them to numerous experiments, but so far they seem to have received little psychological study. As a contribution to this interesting field the present paper reports an investigation of certain emotional and associative processes of the domestic pigeon. Since the literature of the subject is meagre, I shall state at the beginning a few related facts which I have gathered from various sources; then I shall discuss in detail the problems, methods, and results of my several experiments.

The brain of the pigeon is well developed, although the hemispheres are unconvoluted. When they are removed, the animal retains unaltered its reflex and vital activities, but ceases for a time at least to show evidence of mental life, for example, memory and will. In the normal animal sight and hearing are acute, and touch seems keen, although the claws are not used for grasping and eating, as in the case of more intelligent birds, especially, parrots. There is considerable sensitiveness to temperature changes. Taste, and probably smell, appear to be deficient. The "sense of support" is marked, even in the young.

Since the pigeon seems to dream and also to miss its absent mate, some observers believe that imagery is present. There is certainly local memory, and also capacity to observe. Various intelligent acts have been reported. The remarkable homing habits of the carrier pigeon have received no satisfactory explanation. While Cyon suggests the stimulation of the nasal organs by air currents, Thauzièr holds to the electrical theory; they agree, however, that certain higher psychical processes are probably involved.

Graber's tests indicate that pigeons have no color-preference. Beebe's statement concerning birds in general is peculiarly true of pigeons: "There are few species which do not show the emotions of love and sympathy, and ... one will sometimes pine and die of grief at the loss of its mate." After referring to their patient care of the young, he adds: "Indeed, sympathy is the keynote in the development of the higher mental faculties." These birds communicate, but their language consists of comparatively few sounds. As in many other birds, the play-instinct is highly developed.

II. PROBLEMS AND METHODS

My study of the pigeon's emotional life had for its object certain respiratory "expressions." These were investigated by means of a pneumographic tracing, secured while the animal was comfortably fastened in a shallow nest, partially open below. A small box was placed over the bird, and apparatus was so arranged that the time of giving various stimuli was recorded automatically on the smoked paper of the kymograph drum, below the breathing-curve. A third line indicated rate of drum movement. Although some interesting results were obtained, the chief significance of the research consists in its demonstration of the fact that this method of studying animal mind is valuable.

In the study of association I sought to determine the sense-data which the process involves, its method of formation (with due regard to social conditions), its rapidity, permanence, and modifiability, and also its probable degree of complexity. Material contributing to the subject was secured by observing the behavior of the animal when seeking to obtain food by overcoming such obstacles as labyrinths with wire passages, and latches, when the food was left in view, or by finding it when out of sight. In the latter case it was placed in a box occupying a customary place in a group of exactly similar boxes, or else in a box of color or form unlike the other members of the group and variously arranged, from time to time, with respect to them. When the animals were learning the labyrinth habits, various stimulations were given them; later the character of some of these was altered, and the resulting changes in behavior were noted. After the habits had been thoroughly learned, the birds were given a rest for some weeks, and then tested again under the old conditions. A few trials were arranged with special reference to the study of imitation; the animals here were tested as to their ability to execute simple but unfamiliar acts, after having only seen them performed by an animal previously trained. Throughout the associational tests the animals very seldom received food in their cages; but as they were tested daily and allowed to satisfy their hunger completely at the last test, they were never in a state of "utter hunger"--a condition which most experimenters think best to avoid. A series of tests, given at the conclusion of the investigation, indicated that the odor of the food had not assisted the animals in reaching it.

In the two series of experiments (emotion and association) thirty-five animals in all were used. They were confined in large cages in a fairly well lighted and ventilated room, and were fed wheat, cracked corn, and occasionally fruit, and kept well supplied with fresh water and sand. They generally remained in a healthy condition throughout the tests, especially during the winter. To exclude, as far as possible, the disturbing influence of fear, they were usually handled only after the room had been darkened. As the noise made by the curtains was objectionable, the birds were tested with the room illuminated by incandescent lamps; the light was turned off before the birds were placed in position for the trials, and again before they were removed from the apparatus to the cages. It is generally agreed that an experimenter should be out of sight when giving a test. I am convinced that it is important to avoid being seen by the animals at any time. This involves great inconvenience, especially when one employs the pneumographic method, but better results are thus obtained.

For practical suggestions as to apparatus and methods I am greatly indebted to Dr. Robert MacDougall, at the beginning of my investigation, and to Dr. Robert M. Yerkes, throughout. I also owe much to the researches of Zoneff and Meumann, Thorndike, Mills, Small, and Kinnaman. Porter's interesting study of sparrows was made almost simultaneously with the investigation here reported. Fewer animals were used by him, but in some instances more tests were given.

III. INVESTIGATION OF EMOTION

1. Respiration in general. The normal breathing-curve in pigeons is quite similar in contour to that of the human subject, although the rhythm is more rapid and the pauses are less pronounced. When acoustical, visual, olfactory, or tactual stimuli are given, various modifications appear, for example, quickening, deepening, and minor irregularities. It was noticed that meaningless stimuli (pistol-shots) quickly lose their disturbing influence, whereas the breathing remains sensitive to those of a significant character, such as the noises made by other birds. It was also found that a stimulus which no longer affects the breathing will sometimes occasion disturbance if accompanied by a second stimulus of another order, although of a weak intensity (summation).

2. Respiratory reactions to light. As the easy control of conditions makes vision an excellent field in which to work, light reactions were investigated in detail. Two distinct series of tests were given. One sought to determine the relation between quality of light and reaction; the other, between intensity of light and reaction. Four colors of one intensity and three intensities of one color, respectively, were used. In the first series four stimuli, one for each of the colors, red, yellow, green, and blue, were given daily; in the second series five daily stimuli were given, of the same intensity for any one day, and one minute apart; this made it possible to observe also the effect of repetition. Each stimulus was given at the beginning of a respiration and continued two seconds. When the tracings were studied, various modifications were noted, but special attention was paid to alterations in rate of breathing. In the case of both sets of trials an immediate quickening usually occurred after stimulation, and occasionally shallowing and minor irregularities of contour.

In the first set of tests ten animals were used for twenty-five days. Average results indicated that red and yellow are less stimulating than green and blue. To secure data that would assist in the interpretation of these results, an investigation was made of the animals' color-preference. This was done by recording, at thirty-minute intervals, the position of the birds when confined, singly, in a box one half of which was illuminated (from the side) by light of one color, and one half by light of different color but of the same intensity. A water screen excluded the heat rays. After nine records had been taken the colored glasses were interchanged, and the animal's position relatively to the two colors was observed as before. This was repeated with the other colors until each of the four had been used with each of the other three. There were far more choices of green and blue than of red and yellow, though none of the colors was avoided. It seemed a question of degree of liking, rather than of liking or disliking. As stated, Graber's experiments indicated that pigeons have no color-preference, but his results are probably untrustworthy, since he tested several animals at once and apparently was not careful to change the colored glasses regularly. Putting together our two sets of data (the latter stated first) we have the following comparison:

R Y G B Color choices of } 5 animals } 72 129 167 172

Breathing-rise of} 10 animals } 9.94% 10.39% 10.41% 12.11%

Although the proportions do not hold, there is a direct correspondence between the two series of responses; hence it would seem that increased respiratory activity is an expression of agreeable feeling in pigeons, and this especially since the breathing, when varying at all in amplitude, usually became shallower, and also showed certain minor irregularities of contour, as often occurs in human respiration during moderate stimulation of a pleasant character.

In the second series of respiratory tests four animals were used for fifteen days. Average results showed nothing as to the relation between intensity of stimulus and amount of quickening, since the three intensities used, 1, 2, and 4, produced reactions, respectively, as follows: 6.6%, 4.3%, and 6.4%. This may have been because the three intensities were employed each on different days. When the reactions are averaged according to daily succession, without regard to the intensities of the stimuli, we get the following results: first reaction, 8.0% rise in rate; second, 3.7%; third, 4.1%; fourth, 5.7%; and fifth, 6.9%. We should have expected the second daily response to be less vigorous than the first, since the animals were perhaps better prepared for the second stimulation. That the reactions increased thereafter was probably due, partially to summation, and partially to the fact that the short illuminations occasioned mental action (perception of interior of box, increased desire to escape, etc.) which involved heightened, rather than depressed, breathing activity, and thus worked directly against the dulling tendency of repetition.

IV. INVESTIGATION OF ASSOCIATION

1. Labyrinth experiments. Four labyrinths in all were used (L, M, H, O). Each was constructed by attaching moveable wire partitions in a wooden box, covered with chicken wire. The pigeon was admitted through a small entrance compartment which was fastened at one end of the box, and which communicated with it by means of a lifting door, operated by pulling a cord from behind the observation curtain. Food was placed within the maze, and usually at the opposite end. Before beginning the tests the bird was allowed to become thoroughly familiar with the box without the partitions. After a few trials it learned to go to the food immediately upon entering the box. The partitions were then put into position, and the bird was tested as to the time required (except in the case of labyrinth O) and as to the method employed in reaching the food. The time was measured by means of a stop-watch, and the bird's horizontal movements were recorded on a small plot of the labyrinth; other general observations were added.

A. Habits in Labyrinth L

In this labyrinth (Fig. 1) six animals were tested once daily for thirty days, and five of these again after two and six weeks, respectively. On entering the labyrinth with the partitions in place the first time, a bird started on its usual direct course toward the food-box; running against the first partition it made vigorous efforts to push through, flying at the wire and often clinging to it for a short time; some of these random movements eventually brought it to the left of the compartment, and thence, through the opening, into the second compartment, and so on through the others, until finally it reached the food by a series of fortunate accidents. The same general reaction was shown in case of the next few tests, except that fewer and fewer useless movements were made, and that the right ones were carried out with greater and greater precision. Later the animal had no difficulty in reaching the food; it did not run against the partitions, enter the blind alley, nor display such general signs of uncertainty as pausing and looking about. The process of learning in this case was obviously one of "trial and error," or the selection of useful movements. From the mass of random movements constituting the reaction to the unfamiliar environment, only those which enabled the bird to reach the food were retained and improved; the others gradually disappeared until finally the path taken became the shortest one possible, and was entered upon and pursued without hesitation by each animal as soon as it was allowed to enter the labyrinth. The time required for the tests is given in Table I.

It will be seen at a glance that the absolute time required for reaching the food varied for the individuals (see especially the results given by different birds in the case of test 1), but that the several periods for any one bird were relatively similar to those for another; and also that the time was long at first, but rapidly shortened from test to test, thus showing a steady advance in the learning process. Various lapses occurred (for example, A, 8; C, 13; E, 10) after the habit had been fairly well fixed.

In tests 18-22 the time-shortening was due principally to quickening of movements which had already become well defined. The great importance of visual data is brought out by the abrupt lengthening of the periods in the case of tests 23-25, and 26-30, where the light intensities were decreased. The lengthening was roughly proportional to the change of illumination. In the relative darkness the birds had to re-acquire the habits. The same mistakes were made as at first (running against partitions, and into the blind alley), yet here, as before, there was a ready adjustment. That the food was out of sight, or at least very much less visible, probably made no difference, since it was found that the birds would readily go to the old place after both food and food-box had been removed. In order to exclude the light entirely without making their movements invisible to me, I blindfolded the birds by means of a thin black hood, comfortably adjusted over their eyes and top of head; as a result, none was able to make the course in twenty minutes. The first turn, however, was usually made naturally, perhaps because associated with certain non-visual sense-data (sound of the lifting door, and perhaps tactual impressions of the close entrance compartment, etc.). Rats seem far less dependent upon visual data than do pigeons. The great permanence of the pigeons' habits is shown by comparing the periods for tests 31-3, given after two and six weeks of rest, respectively, with those for tests 18-22.

TABLE I. TIME REQUIRED TO REACH FOOD IN LABYRINTH L

Animals

Trials, A B C D E F Average 1 daily. ' " ' " ' " ' " ' " ' " ' " { 1) 28:50 :59 42:20 49:04 22:13 4:04 24:35 { 2) 7:22 :22 25:47 10:17 :48 2:02 7:46 { 3) 1:18 :12 8:29 12:35 :12 1:41 4:05 { 4) :32 :21 10:51 1:26 :19 :52 2:24 { 5) :24 :28 2:36 2:18 :12 1:33 1:15 { 6) :25 :26 1:10 :55 :12 1:50 :50 { 7) :15 :24 :28 :32 :15 2:09 :41 { 8) 1:05 :23 :33 1:19 :17 1:46 :54 1 { 9) :16 :24 :57 :58 :10 :26 :32 {10) :24 :32 1:15 :51 2:12 :31 :58 {11) :12 :21 1:40 :30 :17 :54 :39 {12) :16 :32 :49 1:34 :22 1:18 :49 {13) :13 :18 2:30 :18 :10 :42 :42 {14) :29 :32 :27 :31 :25 :36 :30 {15) 1:00 :24 :30 :31 :12 :35 :32 {16) :19 :52 1:10 :22 :17 :24 :34 {17) :14 :14 :31 :39 :13 :57 :28

{18) :13 :09 :29 :13 :16 :29 :18 {19) :10 :10 :36 :26 :07 :14 :17 2 { 20) :11 :15 :34 :17 :07 :10 :16 {21) :13 :14 :34 :16 :09 :21 :18 {22) :09 :16 :26 :14 :08 :11 :14

{23) :12 :42 1:29 :39 5:53 :13 1:31 {24) :20 :17 1:31 :33 :15 :13 :31 3 { 25) :15 :24 :40 :28 :19 :20 :24

{26) 1:21 16:29 1:22 13:54 3:51 1:36 6:26 {27) 3:36 4:45 :44 1:03 1:09 2:59 2:23 4 { 28) :51 1:24 :46 1:04 :56 1:09 1:02 {29) :51 :41 1:10 :40 2:17 :11 :58 {30) 2:04 :18 :41 :14 :07 :19 :37

5 { 31) :08 :33 :25 :07 :07 :16 6 { 32) :09 :15 :20 :08 :31 :17

1: With 18-candle-power illumination of the room.

2: Same illumination; tests given after the animals had heard four other pigeons pecking in the labyrinth.

3: With 2-candle-power illumination, other conditions the same.

4: With a slight illumination through single curtain, other conditions the same.

5: After two weeks' rest, conditions as in 2.

6: After six weeks' rest, conditions as before.

Let us now notice the gradual progress of learning in three important parts of the maze, as shown in Fig. 2. It will be seen that in the beginning the animals started upon their usual course and pressed against the first partition (stage 1), but that later they touched it less and less (stages 2 and 3), and that finally they avoided it entirely (stage 4). The adjustment here was fairly simple: the sound made by the opening of the entrance door, and the glimpse thus given of the labyrinth, gradually came to be conditions of the movements of turning to the left, on emerging from the entrance, and passing along the compartment toward the opening, where impressions, mostly visual, in the same manner determined the movements of turning to the right and entering compartment 2.

The blind alley was naturally a decided obstacle. The pigeons learned to avoid this compartment by going around it only after many unsuccessful attempts to go through it. During the first test the animals entered it many times (stage 1); on emerging they returned to the second or the first compartment, only to encounter the pen again when they re-advanced toward the food; finally, on reëmerging from the annoying enclosure, perhaps for the eighth or tenth time, they might happen to turn to the right instead of going forward as usual toward the entrance of the box, and thus make their way along the new passage and reach the food. For the next few tests they usually entered the blind alley, but less frequently, and they remained for shorter periods (stages 2-4). Later they merely entered (stage 5); and still later they passed very near the opening without entering, or only paused a moment before it (stage 6); and finally they passed it without the slightest hesitation, walking briskly, but with well-directed movements, midway between the partitions (stages 7-8). The act of turning seemed to be an especially important factor in this habit. We notice that it was a turn to the right (most probably accidental) which first enabled the animals to get beyond the opening of the blind alley; that this same act was repeated in each successive trial until, by the gradual shortening of the loop forming the path taken by the animals in passing into, and from, the labyrinth it was finally reduced to a mere pause (stage 6); and that this later disappeared entirely, leaving only the left turn, which instead was now conditioned by the visual data at that part of the labyrinth and carried the animal past the entrance of the blind alley.

The animals did not come in contact with the second partition until they had almost learned to pass the opening of the blind alley (see stage 6). This was probably because the turn to the left which was made on approaching x was associated with visual data derived from points farther along the course (y), and when the animals reached z, compartment 2, these same data were received and were sufficient to occasion the turn to the left there also, thus bringing the birds against the partition. The adjustment was made principally on the basis of new sense-data arising from running against the wire, looking at it more closely, etc. For a few trials the birds made the turn at x too quickly, and thus failed to reach the third compartment. One of my most intelligent subjects made this mistake in the third test, and again in the sixth and seventh, and retained the act almost unchanged through the tenth, eleventh, twelfth, thirteenth, fourteenth, sixteenth, seventeenth, twentieth, and twenty-first tests, so strong was the tendency to continue a movement once begun, though it was really disadvantageous.

After reaching the food and satisfying their hunger, the animals often returned to the maze passages, seeking an exit; but they never "explored" passages or showed other evidence of "free curiosity" and "desire to know all their new surroundings," as Small reports concerning rats.

B. Habits in Labyrinth M

Five of the animals previously used were next tested twice daily, forenoon and afternoon, for five days, in a larger, more complicated maze. It had two blind alleys, and the food-box was near the centre (see Fig. 3). The animals' general behavior was similar to that before observed. The periods are given in Table II.

TABLE II. TIME REQUIRED TO REACH FOOD IN LABYRINTH M

Animals Trials, A B C E F Average 2 daily. ' " ' " ' " ' " ' " ' " (1) 16:25 2:55 6:33 3:26 4:11 6:42 (2) :55 4:10 2:24 3:36 :23 2:18 (3) 1:12 :55 8:27 9:06 2:07 4:21 (4) :48 :44 2:31 :34 1:04 1:08 (5) :27 :16 :14 :16 :14 :17 (6) :18 :32 :25 :15 :27 :23 (7) :14 :11 :31 :44 :30 :26 (8) :12 :16 :57 :16 :48 :30 (9) :12 :18 :23 :16 :16 :17 (10) :10 :19 :16 :15 :28 :18

Although this maze was much more difficult, it will be seen that the animals learned the route to the food far more readily than before. The first period in this series was only about one fourth as long as the first period in the other, and the course was mastered sooner (by the fifth trial instead of by the ninth). There was less pressing against the wire than before, and unsuccessful movements were sooner discontinued. This improvement was probably due entirely to experience gained in dealing with the first maze. Thorndike speaks of the gradually increasing ability of animals to deal with successive contrivances. The average results given in Tables I and II are plotted in Fig. 4, next page.

C. Habits in Labyrinth H

Since hearing is an important sense in pigeons, we should expect them to be capable of useful acoustical associations. Several things occurred in the course of the two preceding experiments which seemed to indicate that this is true; for example, although I could move about, rather noisily, in the darkened room, without apparently disturbing any of the birds, some few showed signs of fright (moving about restlessly) on hearing the low, grating noises made by lifting the hanging door of the cage, sounds which had always preceded the handling of the subjects before experimentation, and which had probably become signs to them of being taken.

To investigate this kind of association I constructed a labyrinth (see Fig. 5) in connection with which sounds could be utilized as one form of sense-data. The passages were so arranged that along the route leading to the food there were three blind alleys, which the animals would surely enter before mastering the course. In another part of the room were placed, very close to each other, two electric gongs of the same size, but of different material. One was of metal and gave a clear ringing sound; the other was of wood and gave a low rattling noise.

When an animal was learning the route (as in the other two mazes) I sounded the gongs, the metallic, as the bird approached and entered the blind alleys, by openings M, O, and R, and the wooden, as it emerged from them and proceeded along the proper course, and occasionally after it had reached the food. The ringing sound was also given after the animal passed P and was approaching Q. When the new route was fairly well learned, I changed the order of the sound stimuli, ringing one gong at the places where the other had previously been sounded, and compared the records thus obtained with those obtained when the sounds were given in the original order. As an animal is liable to become confused by the sounds, or else quickly accustomed to them, I thought it best to give only a few trials, one trial with the usual order of sound stimulations, the next immediately following with the reversed order, and so on till four pairs of records had been secured, the series of trials being completed in a single day. Four animals were thus tested. The periods of the various trials are shown in Table III.

TABLE III. SOUND ASSOCIATION, LABYRINTH H

Time required to reach food under different sound conditions

I Order of gongs the same as when course was being learned.

Animals Trials A B E G " " " " (1) 13 19 24 17 (2) 16 12 13 16 (3) 10 16 20 14 (4) 12 19 10 12

Total, 51 66 67 59 " 243

II Order of gongs reversed.

Animals Trials A B E G " " " " (1) 14 16 18 27 (2) 37 17 16 27 (3) 14 19 26 11 (4) 27 22 21 14

Total, 92 74 81 79 " 326

In the case of thirteen of the sixteen tests given with reversed acoustical conditions (see column II) the periods were longer than the corresponding ones given alternately with them for comparison, and there was an average time-lengthening of 5.2 seconds per trial, or 34.2%. The following is a short description of the animals' reactions to the changed conditions. It corresponds to the time-values expressed in column II of the table.

Bird A: test 1, animal undisturbed; test 2, drew back from S, turned to the left and went toward R, but later returned and passed S without pausing; test 3, paused at O for a short interval, but did not enter the blind alley; test 4, paused at O again, later drew back from S, turned to the left and entered blind alley 3; it soon escaped, and this time passed S without being disturbed, although it paused at T and U.

Bird B: tests 1 and 2, animal apparently undisturbed; test 3, a few slight pauses at openings; test 4, drew back from S, entered blind alley 3, but soon escaped and passed S without hesitation.

Bird E: test 1, undisturbed; tests 2 and 4, paused at openings; test 3, turned back from S, entered blind alley 3, and paused at several places later when passing toward F.

Bird G: test 1, many pauses; test 2, turned from S and entered blind alley 3; test 3, undisturbed; test 4, drew back from S, went toward R, but did not enter the blind alley.

As the animals gave little attention to the wooden gong, but were always sensitive to the metallic one, their observed movements probably must be accounted for chiefly on the basis of certain visual and organic sense-data now with, and now without, the ringing sound. The data governing the start (as already noticed) were probably sufficient for the avoidance of the first blind alley when the gongs were reversed. In case of the other two, however, the birds had come to depend upon acoustical data, and when these were lacking as they approached the openings O and Q, the left turn could not readily be initiated, hence certain hesitations and misdirections of movement frequently occurred. Experience with the blind alley in the first experiment assisted the animals in dealing with the second blind alley here, but mistakes were made. Visual data usually were sufficient to produce the proper turn at Q, but when the ringing sound was given just afterwards, it sometimes occasioned the left turn, thus bringing the animals toward the opening of the blind alley. While the tests given were not such as would indicate how far pigeons can discriminate sounds, they certainly show that these birds are capable of useful sound associations, although visual ones are evidently of greater importance to them.

D. Habits in Labyrinth O

I next made tests in which tactual and electrical sense-data could also be utilized. In one of the passages of a simple labyrinth was placed a board 8 in. square and 3/4 in. thick, over which were stretched copper wires which formed a series of interrupted electrical circuits. By closing a key a bird could be stimulated whenever it stepped upon the wire surface. A second key was connected with a metallic gong. When an animal on its way through the maze first stepped upon the wire surface, electrical and acoustical stimuli were given; later it was allowed to walk across the board without being thus stimulated; afterward acoustical stimuli were given it at various parts of the maze.

Eight animals were used. All were found quite sensitive to the electrical shocks, and when next tested they avoided the board, especially if the gong sounded as they approached. Some would show signs of uneasiness anywhere in the maze on hearing the gong. When the board was so placed that they had to pass over it in reaching the food, when once on it they moved very leisurely, often lingering; and if they stepped upon the wire surface in the darkened maze, they showed no evidence of being frightened. Evidently no association had been formed between the peculiar tactual stimulus of touching the wires and the electrical shocks which had at first been given. But the tactual stimulus may have been below the threshold. Yerkes saw evidence of association of this kind in the frog; this animal, however, is probably much more sensitive to tactual stimulation received from surfaces over which it passes than is the pigeon.

The results of these four experiments indicate that the pigeon easily acquires complicated labyrinth habits; that these remain fixed for some weeks at least; that acoustical, visual, and certain organic data are the most important sensory factors; and that the process of learning is one of "trial and error," in which the animal comes to form such a close connection between the sense-data of the interior of the box and those other sense-data arising from movements involved in reaching the food, that when the box impressions are again encountered the other sense-data are revived and readily condition the proper movements. How much memory of eating was involved in these tests cannot be told; but it was certainly not an essential part of the mental act. Proper guidance throughout the course was the main thing, and this was determined by definite sense-data. That recognition, discrimination, and perhaps choice were to some extent present seems likely from the animal's hesitating movements at certain critical points. Thus it is highly probable that when the bird approached a blind alley which it had always entered before (see Fig. 2, stage 6), two alternatives were recognized, to enter, as before, or not to do so, as was usual thereafter, and that the pause had for its mental correlate a state closely bordering upon what in us would be deliberation.

2. Release experiments. Under this heading are included certain cage experiments in which some act, such as touching a lever, pecking, or stepping upon a platform, resulted in the opening of the door, and thus enabled the animal to escape and secure the food lying in view without. The animal was admitted to the cage through an entrance compartment as in the case of the maze trials. Before being tested it was allowed to become familiar with the cage and to reach the food directly by passing out through the open door. When first in the cage the animals did not seem to notice the release apparatus, and hence they probably did not begin learning the method of escape until later when they entered the cage and found the door closed, and the ordinary exit thus obstructed.

A. Latch Tests

The cage here employed was an 18-inch cubical box. The top was of chicken wire and the bottom and three sides of heavy boards; the fourth side was formed by narrow vertical bars and a wire door which opened inwardly and was held by a latch working on the outer surface of the bars. At first a long wooden latch was used, which the animals raised when reaching out for food. As this seemed an unnatural act, downward pressure was substituted by attaching to the latch, now made smaller and of brass, a string which ran over a pulley above the door and down into the cage. As nooses did not seem adapted to the birds, the end of the string was attached to a wooden lever which worked on the inner surface of the bars, about three inches from the floor. Eight animals were tested four times daily (twice in the forenoon and twice in the afternoon) for ten days. The time required to escape and the animals' behavior were recorded as in the case of the labyrinth tests.

When they first entered the box (singly as in the other experiments) and found the usual exit closed they made various attempts to push through between the bars, springing and often flying about with great force and persistency. In course of their random movements they touched the bar and opened the door and thus escaped. Later the unnecessary movements were mostly dropped and the necessary ones became highly specialized. The first association was established between the box impressions received on entering and the movements involved in approaching the front of the box and depressing the lever; later a connection was formed between the sensations of touching the lever, of hearing the sound of the opening door, of feeling the jar, etc., and the movements of turning away from the lever and passing out. The sight of the opening door seemed to be of less service to the birds than the sound and jar. Each animal soon came to touch the bar at the point of least resistance, and usually with considerable precision. The time required by the several birds is shown in Table IV, next page. The daily average results are plotted in Fig. 6, above.

TABLE IV. TIME REQUIRED TO ESCAPE FROM CAGE BY USING LATCH

Animals Trials, Daily 4 daily. A B C E F G H I Av. Average ' " ' " ' " ' " ' " ' " ' " ' " ' " ' " (1) :03 :06 3:20 :08 :03 1:10 4:30 :23 1:13 (2) :05 1:10 :59 1:50 2:00 :10 2:46 1:00 1:00 (3) 5:28 :29 1:04 :21 :29 :52 2:33 4:05 1:55 (4) :31 1:45 3:52 :14 :06 :23 2:40 2:30 1:30 (1:25) (5) :15 :10 2:25 :03 :10 :07 :31 :50 :34 (6) :17 2:00 5:03 :03 :05 :15 :39 :10 1:04 (7) :08 :20 :36 :31 :03 :11 :30 :46 :23 (8) :03 :54 :47 :28 :03 :55 :06 :47 :30 (:38) (9) 1:12 :34 :39 :17 :03 :04 :21 :20 :26 (10) :02 :19 :28 :07 :01 :03 :20 :47 :16 (11) :02 :51 :09 :14 :03 :02 1:46 :47 :29 (12) :02 :15 :24 :12 :02 :02 1:17 :27 :20 (:23) (13) :03 :19 :22 :07 :06 :04 :48 :11 :15 (14) :02 :15 :22 :09 :02 :02 :20 :19 :11 (15) :08 :06 :12 :05 :02 :02 :03 :06 :06 (16) :03 :18 :44 :10 :02 :03 :02 :21 :13 (:11) (17) :02 :17 :41 :03 :07 :07 :11 :06 :12 (18) :02 :42 :26 :05 :01 :03 :31 :15 :16 (19) :04 :13 :48 :04 :01 :02 :15 :14 :13 (20) :04 :32 :35 :04 :01 :02 :08 :22 :15 (:14) (21) :03 :13 :10 :13 :09 :01 1:10 :16 :17 (22) :03 :05 :10 :11 :03 :02 :39 :26 :12 (23) :02 :07 :17 :05 :03 :15 :20 :26 :12 (24) :02 :14 :04 :02 :01 :02 :09 1:03 :12 (:12) (25) :01 :03 :05 :04 :01 :02 :31 :24 :09 (26) :02 :04 :08 :01 :01 :05 1:34 :33 :19 (27) :02 :03 :03 :03 :10 :05 :15 :39 :10 (28) :02 :04 :03 :01 :06 :05 :23 :51 :12 (:12) (29) :02 :04 :10 :03 :02 :06 :23 :21 :09 (30) :01 :03 :03 :02 :11 :02 :09 1:08 :12 (31) :21 :03 :11 :03 :01 :03 :28 :11 :10 (32) :03 :02 :04 :02 :01 :03 :11 :42 :09 (:10) (33) :02 :02 :03 :10 :01 :03 :18 :17 :07 (34) :02 :02 :03 :02 :01 :03 :11 :11 :04 (35) :01 :02 :03 :02 :01 :02 :18 :09 :05 (36) :01 :03 :03 :03 :03 :02 :04 :10 :04 (:05) (37) :01 :02 :07 :02 :04 :01 :11 :09 :05 (38) :01 :02 :02 :02 :02 :01 :08 :03 :03 (39) :02 :04 :11 :01 :02 :02 :14 :08 :06 (40) :01 :02 :03 :01 :01 :03 :10 :03 :03 (:04)

The periods here were similar to those given in the maze tests--the time was long at first, then it shortened very rapidly for a few trials, then more slowly but still constantly, until the act became thoroughly familiar. The process was one of "trial and error" throughout. As before, various lapses occurred, even although the animals were as persistent as usual in their efforts to escape. When the lever was moved to the side or back of the box, none of the animals could escape. In general, pigeons show less ingenuity in dealing with latches than do sparrows, according to Porter's observations, although in some other tests they are equally apt.

B. Pecking Tests

The preceding series of trials proved the animals' ability to utilize certain touching or clawing movements, at first accidental, in making an escape, and showed that these could become highly specialized. Desiring to carry out similar tests in the case of pecking movements, which are quite as natural to pigeons, I arranged a contrivance by means of which the act of pecking at corn-grains fastened to a small piece of cardboard (placed just outside the cage, but within easy reach through an opening in the wire) would open the cage-door by making a delicate electrical contact. Four animals were tested.

On entering the cage they endeavored to escape as before; failing in this they began pecking about until they found the corn-grains and made the contact which opened the door and allowed them to escape to the food without. Three of the animals made their escape in this way several times; but the habit seemed to be one that could not be readily learned, as the successive periods showed little shortening.

Thinking that the pecking of things at a definite place perhaps complicated the matter, I removed the electrical apparatus and arranged to open the door myself by pulling a string whenever the pigeon pecked anywhere upon entering the box. Preliminary experiments with Bird J indicated some ability to profit by this kind of experience. As the act of pecking could be used to advantage in a series of imitation trials, this animal only was allowed to learn to escape by actually pecking; the others were reserved for the imitation tests next to be reported.

C. Imitation Tests

In the experiments already reported the animals were used individually and usually out of sight of the others, although in the same room and within hearing of them. When efforts were made in some of the experiments to test the animals in a separate room, signs of fear and discontent were often noticed, and it was necessary to return to the first room to continue the tests. Some instances were noticed in which a pigeon would do what it saw another doing. For example, one of my subjects would not eat one day, being ill apparently; but when I put two others into the compartment with it, and they began eating the food lying about, it also began pecking. Its act could not have been due to its only then happening to see the corn, for it had before looked toward the food when this was thrown to it.

Desiring to test, under definite conditions, the imitative ability of these animals, I arranged trials in which birds were allowed to see a useful, but simple act performed by another bird, and then were given an opportunity to execute the act themselves. Using the animal which I had trained for the purpose, I allowed its series of acts (entrance to box, pecking, release, and food-eating without) to be observed by another animal, confined in a small wire compartment (similar to the entrance compartment before used) attached to the side of the large cage. Care was taken to see that the confined bird was observing, or at least was looking toward the acting one; in case of doubt, the trial was repeated. Later the trained animal was replaced by the observing one, and the latter's reaction was noted. Five animals, in all, were tested, and each was given two opportunities to escape after having seen the trained animal perform the act ten successive times. None of them, however, showed any signs of trying to escape by repeating the movements so often performed by the bird familiar with the act, but each rushed against the sides of the cage and tried to push through at various places, just as the trained bird had done when first learning the habit.

As the act, or series of acts, was rather too complex to be easily observed and utilized by the other pigeons, I arranged two much simpler tests. In one case the leading bird was taught to open the cage-door by stepping upon a platform (the lowering of which made an electrical contact); in the other, to avoid a blind alley, enter a short passage, and ascend a wooden plane (inclined at an angle of thirty degrees) which led into another box containing food. In these tests it was more difficult for the series of acts to be viewed, but the animals, singly as before, were placed at a point of vantage and apparently saw the movements of the other animals.

Of the five birds tested in the platform experiment, four utterly failed to escape in the two trials given. The fifth, in its second test, went to the platform promptly and thus made its escape, but the success may have been accidental, or due to the animal's experience of seeing and approaching the platform in its first test. In the labyrinth experiment only one bird (second test) avoided the blind alley and went directly up the inclined plane to the food; this success was probably due to experience in the first test. There was certainly no evidence that the animals had grasped the nature of the problem; nothing to indicate that the performance of the trained animal had supplied data for the guidance of future conduct, that is, for the conditioning of the necessary movements, in this case those of pecking, stepping upon a platform, or avoiding a blind alley and ascending an inclined plane.

These results are similar to those which other experimenters have secured in the case of chicks, cats, dogs, monkeys, and also rats. Although the method I employed is doubtless open to the criticism of being artificial, some value at least should be attached to the results; if so it would seem probable that imitation in pigeons is not above the "instinctive" stage, and that learning depends entirely upon first hand experience, upon really doing the thing, and not upon merely seeing it done.

3. Position, Color and Form Tests. The apparatus used in these experiments consisted of small boxes, six inches in height, and open at the top. Sometimes they were exactly similar, and sometimes they differed in color or in form. They were moveably attached, six inches apart, to a board which was placed in a large wire-covered box, having an entrance compartment as in the case of the mazes and cages. Food was placed in one of the small boxes, and the pigeons were allowed to find it twice; later each bird was tested as to its ability to return to the food by depending upon the position of the box in the group, or upon its color or its form. Tests were given in series of six, and the box which was first approached was recorded as the animal's choice for that test. If it made a wrong selection, it was allowed to look about until it found the box containing the food, but in no case was it permitted to satisfy its hunger until the last test of the series. The animal apparently did not see the food until it approached the box, and subsequent tests demonstrated that it was not guided by odor.

A. Position Tests

In this series of trials I used at first six, later nine food-boxes, four inches square and covered with dark gray paper. The board to which the several receptacles were fastened was shifted at irregular intervals to various oblique angles; this was done to prevent the animal from being assisted by the position of the food-box in the larger box rather than in the group of similar boxes. After a bird had been tested sufficiently for one position it was then used, for a week or so, in some other experiment, and thus given an opportunity to forget, to some extent at least, the old experience before being taught to find the food in a box placed elsewhere in the group. For the positions 2, 3, and 4, in the group of six similar boxes, eight animals were each given thirty tests in series of six, as stated above. For the positions 5, 6, and 7, in the group of nine similar boxes, the experiments were shortened. Six animals were given twenty-four trials each, two animals for each of the three positions.

The animals quickly learned the position of the food-box and passed to it promptly when released from the entrance compartment. Changing the position of the board to which the food-boxes were attached did not affect the animals' ability to reach the food readily. They usually selected the proper box as before, although frequently they went around the end of the board and approached the food from the opposite side. The general distribution of choices in the case of positions 2, 3, and 4 is given in Table V; the rate of learning, in Table VI.

With a single exception (Bird B, box 3) the box containing the food was far more often chosen than any one of the empty boxes, and usually more often than all of them combined, the average right choices being: position 2, 62%; position 3, 57.7%; position 4, 53.3%. It will be seen that the animals were more successful in finding the food in the second position than in either the third or the fourth, that is, positions nearer the end were more easily located. But what is of greater interest to us is the rate of learning to go to the right box. This is indicated by the increasing number of right choices from series to series. (Table VI.) That the increase was small was due to the fact that the animals learned so quickly in the first series of six tests that little improvement could be made thereafter; what they could learn they acquired early in the experiment. There is some evidence of improvement after the first series in the case of box 4, a more difficult position, and the average for all three boxes shows a slight improvement from series 1 to series 5, although a falling-off is seen in the last series: 26, 27, 27, 31, 28. The same general features appear in the case of the incomplete tests (positions 5, 6, and 7). For each box there were, on the average, 26 right choices in the possible 48. There were more right choices in the case of box 7 than in the case of either box 6 or box 5, box 7 being nearer one end (box 9). There was little evidence of learning after the first series of trials.

TABLE V. ASSOCIATION OF POSITION: GENERAL DISTRIBUTION OF CHOICES

Choices of boxes 1 to 6 when food was placed in boxes 2, 3, and 4

Food in box 2 Food in box 3 Food in box 4 Boxes Boxes Boxes Animals 1, 2, 3, 4, 5, 6. 1, 2, 3, 4, 5, 6. 1, 2, 3, 4, 5, 6. (B) 1 18 4 6 1 0 1 1 8 7 8 5 0 0 2 21 4 3 (C) 2 20 6 1 1 0 0 9 19 2 0 0 0 7 2 15 6 0 (E) 4 18 2 3 2 1 0 6 18 3 3 0 2 6 2 15 3 2 (F) 5 20 1 2 2 0 1 6 18 3 2 0 0 2 5 14 8 1 (G) 0 18 5 2 2 0 0 3 13 5 7 2 0 0 3 17 9 1 (H) 0 22 4 3 1 0 1 0 18 4 6 1 0 4 8 15 3 0 (I) 4 18 3 1 3 1 0 1 23 6 0 0 0 6 7 15 2 0 (J) 1 17 7 5 0 0 3 0 21 3 2 1 1 4 3 16 6 0

Total, 17 151 32 26 12 2 6 26 138 33 28 9 3 29 32 128 41 7

TABLE VI. ASSOCIATION OF POSITION: DISTRIBUTION OF RIGHT CHOICES

Choices from series 1 to series 5 in the case of boxes 2, 3, and 4

Box 2 Box 3 Box 4 Series Series Series Animals 1, 2, 3, 4, 5. 1, 2, 3, 4, 5. 1, 2, 3, 4, 5. (B) 4 4 5 2 3 2 2 1 1 2 4 4 4 5 4 (C) 5 4 2 5 4 4 4 4 4 3 3 3 2 2 5 (E) 3 4 4 3 4 4 4 2 5 3 3 2 2 4 4 (F) 5 4 3 5 3 5 3 3 4 3 0 2 4 4 4 (G) 3 4 4 3 4 2 2 4 2 3 3 3 2 4 5 (H) 4 4 4 5 4 4 3 4 5 2 3 3 4 3 2 (I) 1 4 3 5 5 4 4 5 6 4 2 1 3 4 5 (J) 4 5 4 3 1 3 4 6 4 4 3 3 3 4 3

Total, 29 33 29 31 29 28 26 29 31 24 21 21 24 30 32 Average for the three boxes, 26, 27, 27, 31, 28.

The method of learning in these position tests was the same as that noticed in previous experiments, namely, building upon chance successes. When first admitted to the large box containing the row of small ones at the farther end, the animal accidentally found the receptacle containing the food, and later associated the movements involved in reaching that position with various sense-impressions of the box, especially those experienced upon entering--certain tactual impressions of the small entrance compartment, sound of the lifting door and sight given of the interior of the large box.

While the results clearly indicate that pigeons readily learn the position of objects, nothing is proved as to "counting." Some experimenters speak of similar trials as "number-tests," just as they do of "form-tests," but this is probably going too far. To investigate counting in animals, experiments should be arranged which minimize spatial responses. These tests certainly show that pigeons can discriminate positions readily, especially toward the ends of the group, but little more is certainly indicated. Porter says: "If we do not find in birds the power to count, we have in their nice sense for the location of a member of a series ... something of that preliminary number-sense which Ribot describes as belonging to children and savages."

B. Color Tests

To investigate the animals' ability to utilize colors in finding their food, I employed the same apparatus as before, except that six boxes were used throughout and each was covered with paper of a different color: red, yellow, green, blue (Bradley's standards, except red, RO being substituted), gray, and black. The boxes covered with black and gray paper were employed merely to complete the group of six. The same method as before was employed, except that the board to which the boxes were attached was left stationary at the end of the large box, and also that the position of all six boxes was changed irregularly for each test.

The general behavior of the animals at the beginning of these tests was quite similar to that shown in the preceding experiment; but it was soon evident that colors occasioned them far more difficulty than positions. The general distribution of choices is given in Table VII. It will be seen that the proper box was usually chosen more often than any one of the empty ones, but never oftener than the other five combined, as occurred in the position tests; also that in the case of each color there were instances in which another color was as often, or more often selected. Yet it is clear that colors may serve as valuable sense-data for these animals. In the first series of six tests (see Table VIII) there were few right choices or none, but in each succeeding series the number increased. The learning process was evidently of the same type as before observed (selection, in this case gradual, of chance but useful movements), and involved visual data largely.

TABLE VII. COLOR ASSOCIATION. GENERAL DISTRIBUTION OF CHOICES

Choices of all 6 boxes when food was placed in red, yellow, green, or blue boxes

Food in Red Box Food in Yellow Box Animals R, Y, G, B, B'k. G'y. R, Y, G, B, B'k, G'y. (B) 12 6 7 3 2 0 2 12 6 7 0 3 (C) 15 8 1 4 2 0 3 12 1 6 3 5 (E) 11 8 5 3 3 0 3 10 5 3 5 4 (F) 7 5 9 6 2 1 6 6 5 6 3 4 (G) 9 1 7 5 3 5 5 11 6 3 2 3 (H) 13 3 5 3 3 3 5 9 6 3 3 4 (I) 11 5 2 4 6 2 4 10 5 4 3 4 (J) 10 0 6 8 4 2 5 10 3 3 5 4

Total, 88 36 42 36 25 13 33 80 37 35 24 31

Food in Green Box Food in Blue Box Animals R, Y, G, B, B'k, G'y. R, Y, G, B, B'k, G'y. (B) 4 5 12 4 5 0 1 3 5 10 6 5 (C) 3 4 14 5 1 3 3 1 2 13 2 9 (E) 1 9 7 11 0 2 3 5 5 10 3 4 (F) 0 8 9 6 0 7 0 3 6 11 5 5 (G) 2 5 16 2 4 1 5 5 4 5 4 7 (H) 1 3 15 5 6 0 7 5 5 6 4 3 (I) 0 4 15 6 2 3 5 5 3 9 5 3 (J) 4 1 12 9 1 3 6 4 4 7 5 4

Total, 15 39 100 48 19 19 30 31 34 71 34 40

TABLE VIII. COLOR ASSOCIATION. DISTRIBUTION OF RIGHT CHOICES

Choices from series 1 to series 5 in the case of red, yellow, green, and blue boxes

Red Box Yellow Box Green Box Animals 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, (B) 0 1 2 4 5 1 2 3 2 4 1 2 2 4 3 (C) 2 2 4 3 4 1 2 3 3 3 3 2 3 4 2 (E) 2 1 2 3 3 1 0 3 2 4 0 1 1 1 4 (F) 0 1 1 2 3 0 2 2 1 1 0 1 2 3 3 (G) 0 2 3 2 2 2 2 1 3 3 2 4 3 3 4 (H) 1 1 3 4 4 0 2 2 2 3 2 3 4 3 3 (I) 1 2 3 3 2 1 2 2 3 2 1 3 3 3 5 (J) 1 2 2 2 3 1 2 1 2 4 1 1 3 3 4

Total, 7 12 20 23 26 7 14 17 18 24 10 17 21 24 28

Blue Box 1, 2, 3, 4, 5, 1 3 1 2 3 3 2 2 3 3 0 1 3 2 4 1 1 2 4 3 1 1 0 0 3 0 1 2 2 1 0 2 2 3 2 1 2 1 1 2

7 13 13 17 21

There is no evidence that the color-preference of the animals assisted them in choosing correctly, in fact, they were rather less successful in dealing with those colors for which they had previously shown decided preference, since the whole number of right choices was less in the case of the green and blue boxes (85) than in the case of the red and yellow ones (92), and since there was a relative diminution in the rate of learning toward the last in case of the former boxes.

To test the animals' ability to discriminate shades of colors in finding their food, two birds were used, with four boxes, each covered with a different shade of red paper, and two with the boxes covered with green paper. The brightness of the different shades was not measured, but to the eye it seemed to be equal in each of the cases. The food was placed in the box having the most nearly saturated color, and twenty-four trials in series of six, as before, were given each bird. The results were quite similar to those secured with different colors. With the red shades there were twenty-two choices of the best saturated shade to eight, ten, and eight, respectively, of the other three; and with green, twenty-one to nine, ten, and eight. The 43 correct choices were distributed from series 1 to series 4 as follows: 7, 11, 12, and 13, which shows learning as before. The relatively large number of right choices was probably due, partially to the fact that fewer alternative choices were possible since only four boxes were used, instead of six, and partially to the fact that the box containing the food may have been slightly brighter than the others.

Throughout these trials the position-element was a decidedly disturbing factor. When the animals were first learning to choose a box of a definite color, some would show a marked tendency to approach a receptacle occupying a certain position, and would persist in this from series to series. Others at first showed no special preference for certain positions, but, after happening to make a correct choice, they would return to that same place the next time, and thus miss the right box which had been changed for the new test.

C. Form Tests.

In this experiment the six food-boxes were each of different form: triangular, square, oblong, hexagonal, circular, and elliptical. They were of the same capacity, and were covered with light-brown paper. As in the preceding experiment, the birds were tested for only four of the boxes, and were given thirty trials each. Six animals were used, and as they were not the same as those previously employed, the square box (which had always been used before) had no advantage over the others in attracting the birds at the beginning of the trials. The tests were given as in the preceding experiment, except that it did not seem necessary to change the position of each of the six forms before giving each test; it was thought sufficient to move the food-box, and, if a wrong choice had been made in the preceding test, also the box wrongly chosen. The results are shown in Tables IX and X.

TABLE IX. FORM ASSOCIATION. GENERAL DISTRIBUTION OF CHOICES

Choices of all 6 boxes when food was placed in Tri., Sq., Hex., or Cyl. boxes

Food in Tri. Food in Sq. Animals Tri. Sq. Ob. Hx. Cyl. El. Tri. Sq. Ob. Hx. Cyl. El. (U) 8 6 5 5 2 4 5 9 6 1 6 3 (V) 9 2 3 6 7 3 5 8 5 2 5 5 (W) 8 7 3 4 5 3 5 8 5 3 3 6 (X) 11 5 2 5 3 4 5 8 3 6 3 5 (Y) 11 4 5 4 2 4 4 9 5 6 3 3 (Z) 8 3 7 4 4 4 6 11 7 3 1 2

Total, 55 27 25 28 23 22 30 53 31 21 21 24

Food in Hex. Food in Cyl. Animals Tri. Sq. Ob. Hx. Cyl. El. Tri. Sq. Ob. Hx. Cyl. El. (U) 5 5 3 11 3 3 6 5 3 5 8 3 (V) 4 4 5 8 5 4 6 6 3 4 7 4 (W) 3 5 3 10 7 2 4 3 4 3 10 6 (X) 4 5 4 7 6 4 2 5 4 4 9 6 (Y) 3 4 2 8 6 7 3 2 5 6 9 5 (Z) 5 4 2 11 3 5 3 5 4 4 9 5

Total, 24 27 19 55 30 25 24 26 23 26 52 29

TABLE X. FORM ASSOCIATION. DISTRIBUTION OF RIGHT CHOICES

Choices from series 1 to 5 in the case of Tri., Sq., Hex., and Cyl. boxes

Food in Tri. Food in Sq. Food in Hex. Food in Cyl. Animals 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, (U) 2 0 2 1 3 1 2 2 2 2 2 1 2 3 3 1 1 2 1 3 (V) 1 1 2 3 2 2 1 1 2 2 0 1 2 3 2 1 2 2 1 1 (W) 1 2 1 2 2 2 1 2 1 2 1 2 2 2 3 1 2 2 3 2 (X) 1 2 3 3 2 0 2 2 3 1 1 2 2 1 1 1 1 2 3 2 (Y) 1 2 3 2 3 2 2 2 3 2 1 1 2 1 3 0 2 2 2 3 (Z) 1 2 2 2 1 0 2 3 3 2 1 2 2 4 2 1 2 3 1 2

Total, 7 9 13 13 13 7 10 12 13 11 6 9 12 14 14 5 10 13 11 13

It will be seen that each animal chose the right box oftener than any other one box, but not oftener than all of them; also that there was a small increase in the number of right choices from series to series. No one of the four forms seemed better discriminated than the others if we may judge from the practical equality of right choices made in each case (55, 53, 55, 52) or from the similar increase in number of right choices from series to series; the hexagonal and cylindrical boxes received fewer choices in the first series than did the triangular and square, but this was exactly counterbalanced in the last series. The triangular box was more often confused with hexagonal and square, and the square with triangular and oblong, than with the others. For the hexagonal box the cylindrical was more frequently mistaken than were the other forms, especially the oblong; and with the cylindrical the elliptical was more frequently confused than were the others, especially the oblong. In this series of tests nothing new as regards general behavior or method of learning was observed.

TABLE XI. POSITION, COLOR AND FORM ASSOCIATION

Total Right Choices Right choices from series 1 to series 5 1 2 3 4 5 Position 57.9% 54.2% 55.6% 56.9% 63.2% 59.0% Color 35.3% 16.2% 29.7% 37.0% 42.7% 51.6% Form 29.8% 17.4% 26.4% 34.7% 35.3% 35.3%

If we compare the results obtained in these three experiments (see Table XI and Fig. 7), we shall see that the pigeons were governed much more by the position of the food-box than by either its color or its form, and that color was better associated than form. Position was a most important factor throughout, as was observed also by Porter in the case of the English sparrow. Porter also found that his sparrows could associate color better than form. In the position-tests the pigeons showed very little improvement from series to series (see table); almost all that the animals could learn was acquired at the beginning. The more difficult color- and form-trials, however, showed almost constant improvement, although we should have expected this to be greater in the latter case than it was. When judged entirely by the actual number of right choices in a given kind of tests, some of the birds made a very poor showing; but from the standpoint of increasing number of right choices they appeared in a wholly different light.

Thus, for example, bird F (Table VII) made only 33 right choices in a possible 120, yet their arrangement is significant, being, from series 1 to series 5, respectively, 1, 5, 7, 10, 10. It is probable that there would have been still greater improvement had the tests been continued; perhaps the animal would have become as proficient in finding its food by depending upon the color of the receptacle usually containing it, as by relying upon the position of the box in the group.

V. SUMMARY

1. Respiration in pigeons is sensitive to various stimuli, and since its alterations of rate, amplitude, etc., can be easily recorded pneumo-graphically without frightening the animals, it may well serve as a process through which to study their mental life.

2. By repetition meaningless stimuli, for example, pistol-shots, quickly lose their disturbing influence; whereas the breathing remains sensitive to those of a significant character, such as the noises made by other birds.

3. Reaction to light of moderate intensity consists principally in an immediate quickening, the amount varying with the color; since a direct correspondence was found between color-preference and breathing-rate, it would seem that here agreeable feeling involves increased breathing activity.

4. Visual, acoustical, probably tactual, and certainly organic data, are the principal sensory factors of the associations of pigeons.

5. The animals readily form useful associations by a method of "trial and error," or the selection of successful movements which were at first accidental.

6. Apparently a pigeon does not learn by merely seeing a new act performed by another pigeon; yet there are instances of simple ("instinctive") imitation, and "trial and error" learning is not wholly independent of social conditions, since it proceeds much more satisfactorily if the animal is trained at least within hearing distance of other pigeons.

7. When a habit is being formed, the "period" required for the first test is usually very long, but learning proceeds quite rapidly during the next few trials; later it is more gradual, but it continues till the act becomes thoroughly familiar.

8. Associations are fairly permanent, and some remain practically unaltered for at least six weeks. Modification is easily accomplished, however, on the basis of new experience.

9. Pigeons differ widely both as to the ease with which they acquire associations and also as to their permanence. Difference in activity seems the chief reason for this.

10. While these birds seem mentally inferior to English sparrows and to various mammals which have been tested in a similar manner, they are capable of numerous ready adjustments. They discover circuitous labyrinth passages, they learn to manipulate latch apparatus when adapted to their natural habits and conveniently placed, and they easily reach their food by depending upon the position, color, or form of the box containing it. But the process is apparently simple association throughout. There is no evidence of higher mental activity--no looking the situation over and acting accordingly, no "reasoning" in the proper sense of the word, but only blind movements, some of which are retained and become highly specialized, merely because successful.

FOOTNOTES:

REACTIONS OF THE CRAYFISH

BY J. CARLETON BELL

The crayfish has long been the typical Crustacean for anatomical and physiological investigations, but it is only recently that its reactions to sensory stimuli have been made the object of experimental study. The purpose of this paper is to describe the reactions of the animal to certain sensory stimuli under experimental conditions, and to estimate the relative importance of these stimuli in the life of the organism.

I. REACTIONS TO VISUAL STIMULI

Huxley states that crayfish avoid direct sunlight, hiding under stones during the day, and becoming active in the evening. On the other hand, they are attracted like moths to fires lighted on the bank at night, and may be scooped out by hand. Abbott, giving an account of the burrowing crayfish, Cambarus diogenes, states that it is very difficult to observe the animals at work, since all their digging is done at night. It would seem from the account of Miss Hoppin, quoted by Garman, that the blind crayfish, Cambarus pellucidus, is not altogether insensitive to light, for, reporting on the fauna of the caves of Missouri, she says that the crayfish are all found near the entrance to the cave, where there is considerable light. In the dark recesses there are only little white fishes. Blind fish and crayfish are also taken from the wells in the neighborhood, where the crayfish are found only in wells that are rather shallow and light; the fish, on the other hand, are only obtained from deep, dark wells.

According to the above accounts it would appear that the crayfish is negatively phototactic to direct sunlight or diffuse daylight, but positively phototactic to a light at night, and moreover, that light may influence the behavior of the animal even when the eyes have ceased to function.

The directive influence of light upon the movements of the crayfish has never been experimentally studied to my knowledge. Dearborn thinks that light has no effect upon the animals. Yerkes and Towle have shown that Daphnia move toward the light. Bethe finds that Carcinus is negatively phototactic, and also shows a tendency to hunt out corners. When the eyes are varnished with lampblack, the phototaxis disappears, but the tendency to seek out corners still remains. Bethe says that he has observed the same phenomenon in the crayfish. Keeble and Gamble discovered that Hippolyte varians responds positively to light under all conditions, and Palæmon is just as markedly negative. Macromysis, however, reacted now positively now negatively, depending on the background. A black (absorbing) background called forth a positive response, while a white (scattering) background produced a negative reaction. Spaulding, in studying the habits of the Hermit Crab (Eupagurus), found that it is strikingly positively phototactic. When animals are placed in an aquarium, one half of which is shaded, none of them are ever noticed inside of the dark line. Herrick notes that lobsters are nocturnal, and avoid the light when placed in a tank, and Bateson says that prawns and shrimps lie hidden during the day, and are active only at night. Parker, in a study of Copepods, finds that the females have a strong positive phototaxis for light of a low intensity, while males show a weak negative phototaxis. To light of over 100-candle power at a distance of 10 cm. or to direct sunlight the female Copepods are negative, while the reaction of the males does not seem to be altered.

In his work on Carcinus, Bethe obtained retraction of the eye-stalks by suddenly throwing a strong light on the eye by means of a mirror. "Usually the eyes were quickly drawn in and protruded again, sometimes several times in rapid succession, like a man blinking under a sudden, strong light." When a dark object, the size of the hand, was moved just over the water, the eyes were seldom retracted, but the antennules were usually drawn in. Lemoine observed that in Astacus retraction was due to touch alone, and that no light, however strong, was able to bring about such a reaction. Gulland takes just the opposite view with reference to Astacus, stating that there are no setæ of any sort on the eye-stalk, and therefore it is insensitive to touch, but is withdrawn only because the animal sees the object by which the stimulus is given. If a curved needle is used, and the stimulus is applied from behind, no retraction follows. Dearborn, however, working with Cambarus, agrees with Lemoine in saying, "Withdrawal of the ophthalmites into their sockets occurs only on contact with some hard object,--not from any light-stimulus of an ordinary sort." I may say in passing that in none of the following experiments on Cambarus was there ever a sign of retraction due to stimulation by light, the retraction always taking place in response to a touch-stimulus.

Lyon, in his study of compensatory movements of the eye-stalks, found that when the eyes were painted with lampblack, the crayfish showed a reduction of about 10% in the compensatory movements when rotated in vertical planes, but the compensation remained the same for rotation about the dorsi-ventral axis. On rotation in the dark the compensatory movement of the eyes was found to be from 5° to 8° less than in the light.

EXPERIMENTAL

In the investigations to be described, 58 crayfish of the species Cambarus affinis were made use of, and for identification the animals were marked on the back with white enamel paint, the males receiving the even numbers from 2 to 64, the females the odd numbers from 1 to 51.

1. Reactions to White Light

The questions proposed for investigation were, (a) How does the crayfish react to diffuse daylight; (b) to reflected sunlight; (c) to direct sunlight; (d) to artificial light of different intensities? (e) What is the influence of previous conditions of exposure to light upon the reactions of the animal? (f) Do changes of temperature affect the reactions?

A wooden box, 80 cm. long, 25 cm. wide, and 20 cm. high, painted black on the inside, and constructed so as to hold water, was covered with a heavy black cloth to exclude the light from above. The front end of the box was of glass, thus admitting the light from the end. In all the experiments except those with direct sunlight, this glass end was covered with black cardboard in which a hole 10 cm. long and 5 cm. high had been so cut that the light was admitted at the middle of the bottom of the glass. The direct sunlight was admitted through the whole of the glass end. At the rear of the box a piece of black cardboard was so arranged that an aperture was afforded for observing the animals without admitting any appreciable amount of light, and this aperture could be readily closed by a slide when not in use.

The method of experimentation was to place the animal in the box about 20 cm. from the glass end, and observe whether it went toward or away from the source of light. The animals were experimented on in two groups of five each, and one hundred observations were made on the individuals of each group with each intensity of light, that is, twenty observations on each animal. In order to check the influence of the orientation of the animal at the time of exposure to the stimulus, the following four positions for placing the animal were chosen: (1) Head toward the light; (2) Head away from the light; (3) At right angles to the light with right side toward it; (4) At right angles with the left side toward the light. Thus five observations were made on each animal of each group in each position, exposed to each of the different intensities of light.

Seven different intensities of light were employed, and the results have been arranged in eight sets, as follows: I. Diffuse daylight in dry box, i. e., the animals were taken out of their ordinary medium, water, and were exposed to the stimulus of diffuse daylight in the air. The reactions under these conditions, however, were so slow and so unsatisfactory that the test was abandoned after the first group, and thus the second group has nothing to show for itself under this head. The remaining seven sets of observations were made on animals placed in 10 cm. of water at 15° C. II. Diffuse daylight. III. Reflected sunlight. The box was placed near a window on a clear day, and the sunlight was thrown in horizontally by means of a mirror. IV. Direct sunlight. On a clear day the box was placed in such a position that the sun shone in directly and illuminated the front half of it. V. 9-candle-power incandescent electric light. This lamp was marked 16 c., but it had been used a great deal, and on being tested with a Lummer-Brodhun photometer showed only 9 c. VI. An incandescent electric light of about 50 c. This lamp was marked 100 c., but had been used considerably and was slightly smoked. Unfortunately it was broken before there was any opportunity to test it. Judging from the fact that another 100 c. lamp of the same manufacture, in slightly better condition, measured 64 c., the estimate of 50 c. seemed a safe one. VII. The incandescent electric light alluded to above, which measured 64 c. VIII. An arc light which varied in intensity from 150 c. to 250 c.

In intensities V and VI the lamp was placed 5 cm. from the glass end of the box to allow the interposition of a heat-screen consisting of an alum solution in a flat glass jar 5 cm. thick. Reckoned in candle-metres, therefore, the intensity of the illumination at the surface of the animal in V was 144 c. m., and that in VI was about 800 c. m. In VII two heat-screens were used, and between these was placed a lens of considerable but not accurately determined curvature, so that it is impossible to express the intensity in candle-metres. In VIII the light was so variable that such an expression would mean nothing.

Unfortunately it was impossible to keep the two groups constant throughout the whole series, owing to the death of two individuals in each group during the experimentation. Group 1 was composed of nos. 1, 3, 4, 8, and 9, of which 1 and 8 were replaced by nos. 13 and 42 respectively. Group 2 was begun with nos. 23, 27, 32, 34, and 38, and the vacancies caused by the death of 23 and 32 were filled by nos. 21 and 36. The following table exhibits the reactions to the different intensities of light, + indicating an orientation toward the source of light, - an orientation away from the light, and ± an indifferent orientation, which usually means no movement at all.

TABLE I. SUMMARY OF REACTIONS TO WHITE LIGHT

Group 1 Group 2 Totals + - ± + - ± + - ± I. 23 48 29 23 48 29 II. 49 47 4 19 81 68 128 4 III. 40 60 30 70 70 130 IV. 47 52 1 36 64 83 116 1 V. 51 49 30 70 81 119 VI. 39 61 39 61 78 122 VII. 28 72 28 71 1 56 143 1 VIII. 35 63 2 37 60 3 72 123 5

312 452 36 219 477 4 = 531 929 40 = 1500

The following table gives the average time required for orientation for each group. The time of each animal in seconds was noted with a stop-watch from the instant the animal was placed in the box until a definite orientation was assumed with reference to the light. If no orientation followed within three minutes, the result was called indifferent.

TABLE II. AVERAGE TIME OF ORIENTATION

Intensities I II III IV V VI VII VIII Ave. Group 1 (144) 52 11 4-1/2 9 8-1/2 11 19 16-1/2 Group 2 9 3 4 4 5 4 5 5

Inspection of these tables shows that when the animals were taken from the water and placed in diffuse daylight in the air (I), their movements were so sluggish that in twenty-nine cases out of one hundred there was no orientation within the three-minute limit. Moreover, in the seventy-one cases where there was definite orientation the average time was over two minutes (144 seconds). While, therefore, the conditions were quite different from the normal environment of the animal, it is interesting to note that of the cases where orientation did take place the negative reactions were more than twice as many as the positive. In II, where the animals were under the same conditions of diffuse daylight but in the water, a wide variation in the reactions of the two groups is noted. In group 1 they are about equally divided between positive and negative, while in group 2 there is the largest proportion of negative reactions in the whole series. It will be observed that the time for group 1 is extremely long compared with the other averages, and this doubtless indicates a general sluggishness and lack of sensitiveness to stimuli in the animals, which might to some extent account for the difference in reaction. If we consider the totals of both groups for each intensity, we are led to conclude that there is no appreciable difference in the reactions of crayfish to diffuse daylight, to sunlight, or to artificial light within the limits here employed. A slight exception to this is found in VII, where the 64 c. lamp with the lens caused a somewhat more uniform negative reaction. The action of direct sunlight in IV is rather remarkable in that with the lowest proportion of negative reactions in the whole series we also observe the shortest time-average, indicating that the animals are the liveliest and most sensitive. This would seem to indicate that while the animals are in general somewhat negatively phototactic to all light-stimuli of moderate intensity the action of direct sunlight tends to reduce the negative phototaxis to a minimum. If we consider the totals of the two groups separately we observe that group 1 has only 57% of negative reactions while group 2 shows 68%. This is rather in accordance with what we would expect from the general time-average, which is over three times as much for group 1 as for group 2. But although we may in a general way connect rapidity of orientation with a large percentage of negative reactions, it will not do to carry it to individual cases, for it was observed that no. 27 showed 83% of its reactions negative, yet its total time-average was 10 sec., the highest in its group.

In general, then, we conclude that the crayfish is negatively phototactic in the proportion of about two to one. This apparently contradicts the statement made by Huxley that crayfish "are attracted like moths to fires lighted on the bank at night." For surely if this were the case some such tendency would have been observed in these experiments. On the other hand, there is no such marked and definite response to light as in the case of Daphnia or Hippolyte or Palæmon or the Hermit Crab. The action of the stimulus is by no means mechanical and constant, but there is wide variation in individuals.

As was mentioned in the description of the method of experimentation, four different positions for placing the animal were chosen with the idea that the initial position of the animal with respect to the light might have some influence on the direction of its movement. To determine what this influence might be, a careful record was kept of the orientation with reference to each one of these positions, and the following table gives a summary of these observations. In the table position I is where the animal is placed with its head toward the light; position II, with head away from the light; position III, at right angles to the light with the right side toward it; position IV, at right angles with left side to light.

TABLE III. INFLUENCE OF POSITION ON LIGHT REACTIONS

Position I Position II Position III Position IV Totals + - ± + - ± + - ± + - ± + - ±

Group 1 85 108 7 61 135 4 74 114 12 92 95 13 312 452 36 Group 2 39 133 3 48 126 1 57 118 75 100 219 477 4

124 241 10 109 261 5 131 232 12 167 195 13 531 929 40

Since the animals have been shown to be somewhat negatively phototactic, we should expect that position II, with the head away from the light, would show the largest number of negative reactions, and this is what we find if we take the sum of both groups. But by the same course of reasoning we should expect position I, with head toward the light, to yield the smallest number of negative reactions, a condition which prevails neither in the sum nor in either of the groups. On the whole we can only say that difference of position seems to have remarkably little influence on the orientation of the animals.

In his work on the eye of the crayfish, Parker called attention to the migration of the pigment in the retinular cells under the influence of light. The question now arose, what influence, if any, does this pigment migration exert upon the reactions of the crayfish to light? The time required for pigment migration in the eye of the crayfish has never been determined to my knowledge, but from the work of Parker on Palæmonetes it was thought that confinement in the dark for about an hour would be sufficient to bring about a retraction of the pigment. Accordingly, group 1 was kept in the dark for one hour, group 2 for one hour and a half, before experimentation. A further test was made to observe the effects of pigment expansion, both groups having been exposed for one hour and a half to the rays of a 32 c. incandescent electric light at a distance of 40 cm. The apparatus used in the reaction-tests was the box described above, with the 64 c. light as a stimulus. As to the method of observation, each group was placed in the centre of the box at right angles to the horizontal rays of light, and the position of each animal was accurately noted at intervals of one minute for one hour. In reporting the results, all the observations of animals in the half of the box nearest the light are denominated positive, those in the half farthest from the light negative. The results are given in Table IV, where line I indicates the reactions after confinement in the dark, line II those after exposure to the light.

TABLE IV. INFLUENCE OF PREVIOUS CONDITIONS UPON REACTIONS TO LIGHT

Group 1 Nos. 13 3 4 9 42 Totals + - + - + - + - + - + - I 54 6 60 5 55 48 12 60 167 133 II 59 1 60 60 9 51 55 5 183 117

Group 2 Nos. 21 27 31 36 38 Totals + - + - + - + - + - + - I 5 55 5 55 7 53 10 50 4 56 31 269 II 38 22 60 13 47 45 15 16 44 116 184

Let it be said at once that these results do not offer an altogether satisfactory basis for an answer to the question proposed above. Some of the animals would take up a position during the first ten minutes and remain in it for the rest of the hour. Whether the position taken was due solely to the light, or was owing to thigmotactic influences, or whether it depended on the way in which the animal was released, are questions which cannot be answered, and for this reason the conclusions to be drawn from the table are tentative. If we examine the table we find that the totals of both groups agree in manifesting a decrease in negative results for line II, after exposure to the light, as compared with line I, after confinement in the dark. This is what we would expect from negatively phototactic animals. When taken from the dark the pigment is retracted, and the sensitive retinal substance is exposed to the direct action of a rather strong light. The negative tendency of the animal we should expect to find accentuated. The decrease in negative reactions is especially marked with group 2, which was shown above to be much the livelier of the two, and all the individuals except no. 27 share in the change. In group 1 the decrease is not so striking, and is observed to be due to two individuals solely. Inexplicable is the preponderance of positive over negative reactions in the results for group 1.

All of the experiments thus far described were carried out in water at approximately 15° C. The question naturally arises, what will be the result of raising or lowering the temperature upon the reactions of the animals to light? Unfortunately the experiments anent this question are fragmentary and incomplete, but the results will be given for what they are worth. The same apparatus and the same intensity of light (64 c.) were used as in the preceding paragraph. The results, presented in Table V, are arranged in three sets, as follows: The line marked I represents the results obtained from group 1 at a temperature of 5° C. The animals were placed in the box one at a time, as in Table I, and their orientation noted. They were set at right angles to the rays of light, five times with the right and five with the left side toward the source of the stimulus. No observations were made upon group 2 at 5° C., and those on group 1 are so few as to have a questionable value. Line II gives the reactions of both groups of animals in water at 25° C., and in this set the animals were placed in all four of the positions indicated for Table III. Line III presents the reactions of the animals in water at 25° C. by the method outlined for Table IV, i. e., each group of animals was placed in the centre of the box, and observed at intervals of one minute. To obviate the objection of the animals remaining in one spot, they were reset every ten minutes in the middle of the box, at right angles to the entering rays.

TABLE V. REACTIONS TO LIGHT AT DIFFERENT TEMPERATURES

Group 1 Nos. 13 33 4 42 9 Totals + - + - + - + - + - + - I 7 3 4 6 2 8 2 8 8 2 23 27 II 16 4 14 6 11 9 9 11 13 4(±3) 63 34(±3) III 19 41 13 47 39 21 17 43 27 33 115 185

Group 2 Nos. 21 27 31 36 38 Totals + - + - + - + - + - + - II 12 8 5 15 12 8 15 5 13 7 57 43 III 12 48 12 48 31 29 42 18 44 16 141 159

Judging from lines II and III we may say that there is a tendency toward a decrease of the negative phototaxis with an increase in temperature. It is true that group 1 in line III maintains the average, 62% negative reactions, but the others are much lower than this, line II even going over to positive phototaxis in both groups. In line III the animals of both groups were extremely active during the first ten minutes, rushing about from one end of the box to the other, pushing each other back and forth, and in general exhibiting great restlessness. Some of the animals when first put into the water reacted with a sort of cramp reflex, which was followed in a few seconds by intense activity. After the first ten minutes the animals began to grow more quiet, and in twenty or thirty minutes they had become quite sluggish, scarcely moving out of the position in which they were reset. During the period of restlessness the males showed marked sexual activity, rushing up to the females, pushing them about, seizing them, and trying to turn them over in spite of their vigorous resistance. One of the males, no. 36, did succeed in turning a female on her back twice, although she struggled violently to escape,--a thing which the female never does in the ordinary sexual act. The rise in temperature, therefore, seemed to stimulate the males to sexual activity, but not the females.

2. Reactions to Colored Light

No observations have ever been made, so far as I know, on the reactions of the crayfish to colored light. Lyon, in his work on compensatory eye-movements, found that rotation in blue light gave a compensatory movement only slightly less than that in white light, while in red light the compensation was only a little larger than in darkness. In some animals the interposition of an opaque object between the eye and the source of light caused an elevation of the eye 1° or 2° toward the vertical. Red glass acted like an opaque object, blue glass produced no effect, i. e., blue light had the same effect as white light. To observe whether the same thing applied to movement reactions was the object of the following experiments.

a. Reactions to Horizontal Colored Light. The same apparatus was used as in the previous experiments, viz., the dark box with light from the 64 c. lamp entering horizontally at the end. Across half of this end were placed pieces of colored glass of a saturated blue, green, yellow, and red. The colored light obtained by this means was not spectrally pure, but it was the nearest to it that could be obtained. A more serious objection is that the intensities were not the same, the red and the yellow being very appreciably brighter than the blue and the green. In addition to observations with these colors, a piece of black cardboard was introduced in the same position as the glass, thus cutting off the light from that half of the box. This, to preserve the uniformity of the series, is denominated black. The animals were placed in the centre of the box, on the line separating the white from the colored light, and were observed at intervals of one minute for forty minutes, the position of each animal being accurately noted. At the end of every ten minutes the animals were reset at the centre of the box. The following table gives a summary of the results for each individual. Here again it was impossible to keep the groups constant owing to the death of individuals during the progress of the experiment.

TABLE VI. REACTIONS TO HORIZONTAL COLORED LIGHT

Group 1 Group 2 Animals 13 37 41 42 44 33 4 9 43 46 Sum 21 27 31 36 38 52 Sum Sum Total Blue 12 9 21 30 18 90 23 2 24 13 37 99 189 White 28 31 19 10 22 110 17 38 16 27 3 101 211

Green 15 9 23 36 17 100 12 19 3 28 40 102 202 White 25 31 17 4 23 100 28 21 37 12 98 198

Yellow 20 10 31 28 19 108 17 11 28 37 35 128 236 White 20 30 9 12 21 92 23 29 12 3 5 72 164

Red 26 9 32 20 17 104 2 30 16 29 36 113 217 White 14 31 8 20 23 96 38 10 24 11 4 87 183

Black 24 27 12 27 26 116 15 18 26 25 5 89 205 White 16 13 28 13 14 84 25 22 14 15 35 111 195

In this table the colored lights are arranged in the order of the spectrum from blue to red. On the hypothesis that blue light has practically the same effect upon animal reactions as white light, while red is about the same as darkness, we might expect that the reactions would be about equally divided between the blue and the white, and that there would be a gradually increasing difference in number as we go down the table, reaching the maximum with the last pair, black-white. This, we see, however, is not quite the case. In both groups the white has a slightly larger number of reactions than the blue, while the pair green-white shows numbers more nearly equal. In the sum totals the yellow shows a greater preponderance over the white than any other color, and the black and white are very nearly equal. Group 1, it is true, shows a fairly regular ascending scale in reactions to the colored lights with the exception of the red, and the same might be said of group 2 if it were not for the very low number of reactions to the black and the exceptionally high showing of the yellow. On the whole, however, the differences are so small and the individual variations are so large that we can only conclude that for these conditions colored light has little or no effect on the reactions of the animals.

In the foregoing experiment the light came from a broad spiral coil inside the bulb of the lamp, and the distance from it to the edge of the glass was so small compared with the length of the box that there was no sharp dividing-line between the colored light and the white, but rather a wedge-shaped block of lessening saturation of the color, and this wedge, having the point toward the light, took up the whole of the box at the extreme farther end. Thus the imaginary central line dividing the white light from the colored departed farther and farther from the reality as the rear of the box was approached. To obviate this difficulty and to get a check on the previous work, the following series of experiments was undertaken.

b. Reactions to Vertical Colored Light. The same box was used as in the previous experiments, but the end was closed with a black cloth, and an electric light marked 32 c. but measuring only 22 c. was hung exactly over the middle of the box, 40 cm. from the bottom. By means of wires it was arranged that a plate of colored glass could be swung in such a manner that all of one half the box (the whole of one end) was illuminated with the desired color, while the other half was either left white or illuminated with another color. In this way there was a fairly sharp dividing-line between the two colors. The animals were observed at intervals of one minute for 40 minutes, and reset at the middle on the dividing-line every ten minutes as before. Table VII gives the results of the observations.

In this set of experiments it was possible to keep the groups intact except that in group 1 no. 13 had to be replaced by no. 46. If now we conceive the colors arranged in the order of the spectrum with black at one end and white at the other, and consider the black a lower stimulus than the white, we have the ascending series black, red, yellow, green, blue, white. Now since the animals have already been shown to be somewhat negatively phototactic, we should expect them to prefer a color of lower stimulus to one of higher. Turning to the sum totals in the table we find that the first color of each pair (which is always the lower stimulus) has the larger number of reactions in every case but one, the first pair of red-blue. As was stated above, it was impossible to secure colored light of the same intensity by means of the glass at our disposal, and in the present case the red was considerably brighter than the blue. Owing to the fact already mentioned that different intensities of white light seem to have no effect on the reactions it was thought that these differences in the intensities of the colored lights might be overlooked. Since the only thing that could be thought of to account for the anomalous behavior to the red-blue was this difference in intensity, another experiment was undertaken with the same animals under slightly different conditions. A glass aquarium about 40 cm. long by 20 cm. wide was covered with black cardboard and black cloth in such a manner that light could enter only through a space 5 cm. wide at the bottom of each end. Each of these ends was covered, the one with blue, the other with red glass, and 15 cm. from each end was placed an electric light marked 32 c. Later, however, it was found that one of these lamps measured 30 c. and the other 22 c. The red light was found to be much more intense to the eye than the blue, so the former was damped down with tissue paper until the two appeared to have the same intensity. The second pair of red-blue in Table VII gives the results of the observations under these conditions, and these are found to be in harmony with the rest of the table, i. e., the color giving the lower stimulus has the higher number of reactions.

TABLE VII. REACTIONS TO VERTICAL COLORED LIGHT

Group 1 Group 2 Group 3 Sum Animal no. 13 37 41 43 44 Sum 21 27 36 38 52 Sum 54 56 58 60 62 Sum Total

Blue 17 22 14 25 29 107 11 21 26 27 26 111 26 32 26 28 10 122 340 White 23 18 26 15 11 93 29 19 14 13 14 89 14 8 14 12 30 78 260

Green 17 26 11 21 29 104 21 27 30 7 26 111 29 31 26 10 9 105 320 White 23 14 29 19 11 96 19 13 10 33 14 89 11 9 14 30 31 95 280

Yellow 25 21 12 20 35 103 12 20 34 25 29 120 19 20 9 28 22 98 321 White 15 19 28 20 15 97 28 20 6 15 11 80 21 20 31 12 18 102 279

Red 37 33 23 22 28 143 21 30 30 33 22 136 32 37 26 33 28 156 435 White 3 7 17 18 12 57 19 10 10 7 18 64 8 3 14 7 12 44 165

Black 23 20 1 28 34 106 9 12 25 40 32 112 25 32 18 34 16 125 343 White 17 20 39 12 6 94 31 28 15 14 88 15 8 22 6 24 75 257

Black 20 15 13 27 37 112 16 11 16 27 24 94 36 40 18 29 29 152 358 Red 20 25 27 13 3 88 24 29 24 13 16 106 4 22 11 11 48 242

Black 20 25 26 20 32 123 22 15 39 40 32 148 33 22 34 34 11 134 405 Blue 20 15 14 20 8 77 18 25 1 8 52 7 18 6 6 29 66 195

Red 25 17 28 18 8 96 3 15 21 28 20 87 21 13 15 14 28 91 274 Blue 15 23 12 22 32 104 37 25 19 12 20 113 19 27 25 26 12 109 316

Red 21 27 22 10 36 116 13 40 24 21 28 126 22 25 18 21 16 102 344 Blue 19 13 18 30 4 84 27 16 19 12 74 18 15 22 19 24 98 256

The most striking feature of the table is the marked predominance of the red over the white. Here the red reaches 73% of the total number of reactions, and inspection shows that this predominance is uniform not only through the groups but even for the individuals. The constancy of this reaction and the fact that it is so much more frequent than the one to the black as compared with the white, would lead one to expect that the red would have the higher percentage in the combination black-red. Such, however, is not found to be the case, although it does happen with one group. If the arrangement of our color-scale in accordance with increasing intensity of stimulus were correct, we should expect a gradually increasing predominance in reactions to colored light over those to white in the first five pairs. Instead of this we find that green and yellow stand nearest to the white, blue and black come next and are almost equal, while red is very much higher than any. In the pairs black-red and black-blue the red holds its predominance over the blue at about the same rate as in the second pair of the direct comparison, red-blue. The wide individual variations, however, in all these reactions to colored light, except perhaps in the case of red-white, indicate that there is nothing very regular, stereotyped, or mechanical about them. The most that can be said is that in a general way the red end of the spectrum furnishes a less intense stimulus to negative reaction than the blue.

A tendency to habit formation was noticed during the course of these experiments, and it is possible that this may have influenced the results somewhat. Many individuals apparently formed a habit of going to a certain corner as soon as they were reset at the centre. The positions in which they were set were varied and they were headed in different directions, but within a minute after they were released in the middle of the box they would be found in their favorite corner. This was especially the case with no. 38 in Table VI, and I think accounts in some measure for the persistent avoidance of the white. In no case did this continue throughout the whole series, but would sometimes be noted for two or three days at a time in the case of an individual. What were the controlling factors in this habit formation, the means by which orientation and recognition were effected, I was unable to determine.

3. Reactions to Objects

In no case did an animal give any sign of perceiving stationary objects in its path or of avoiding them in any way that could be referred to a visual stimulus. When the animal approached an obstruction there was no hesitation in the movement until the object was touched. Usually even when the antenna had touched the object the animal did not stop, but continued until the contact of the chelæ or even of the rostrum made further movement in that direction impossible.

With moving objects the case was quite different. Here the condition and disposition of the individual animal seemed to be the deciding factors. Often when the animals were trying to climb out of a shallow pan in which they were kept in the experimenting-room, raising a finger or holding out a pencil would be sufficient to make them stop or even start back into the pan. Nor was this response occasioned by any change in the intensity of light, such as that caused by a shadow falling on the animal, for they would react to a movement made on the opposite side of them from the window. In fact, no. 56, the most active in response to moving objects, seemed to react more vigorously to a motion made on the opposite side than when it was made between him and the light. Whenever a person came near the aquarium he and one or two others would take an attitude of defence, and would "face about" to correspond to any movement the person made toward one side or the other. When in the pan mentioned above, any movement of a person within two or three yards of him usually called forth a reaction on his part, and if the pan were placed on the table and the person moved slowly round it, the animal turned with the person, making a complete circuit of the pan.

Reaction to a smaller moving object, however, was not so marked. A black object, 20×8×8 cm., was suspended above the middle of the pan so that if set swinging it would just pass over the top. When it was pulled to one side the animal responded slightly, but after the first swing he seemed to pay no more attention to it. When the operator stepped out from behind the screen, the animal was as keen in its response as before. The experiment was now tried of allowing the object to approach from one direction while the operator moved to a position at right angles to its line of movement. Without hesitation the animal moved so as to keep fronting the operator, without paying any attention to the movement of the smaller object, although this was much nearer.

These observations on the reactions of the crayfish to stationary and moving objects are in line with the conclusions of Plateau and Exner drawn from observations on other Arthropods. It is Exner's belief that the compound eye is a visual apparatus which is almost worthless for detecting the forms of objects, especially if these objects are stationary, but that it may furnish a very keen perception of moving objects.

II. EXPERIMENTS WITH SOUNDS

Hensen stated that Palæmon and Mysis reacted to sounds made by striking a thin, resonant board floating on the surface of the water, or by tapping the walls of the aquarium or of the room. Beer repeated Hensen's experiments, but denied that the Crustacea reacted to sounds, and claimed that their movements were due to visual and tactual stimuli. Prentiss confirmed Beer's results on Palæmonetes, and noted that the reactions were only slightly diminished by the removal of the otocysts, but that removal of the antennæ and antennules caused their almost complete cessation. More extended experiments were made on the fiddler crab, Gelasimus pugilator, which is on land a good deal of the time, and Prentiss's conclusions are: "(1) The reactions formerly attributed to sound-stimuli are nothing more than tactile reflexes. (2) The otocyst has little or no part in calling forth these reactions. (3) There is no direct evidence to prove that decapod Crustacea hear, and until such evidence has been obtained, we are not warranted in ascribing to the otocyst a true auditory function."

The experiments performed on the crayfish in this connection all resulted negatively and go to confirm Beer's and Prentiss's conclusions. Rapping upon a board floating in the water, and tapping the sides of the aquarium did not cause the slightest apparent reaction in the animals under observation, even though the vibration of the water could be plainly perceived by the sense of touch in the hand. When a rather large electric bell was sounded just over the surface of the water some reactions were observed which were evidently due to the movements of the hammer, but there was nothing which could be referred to the sound-stimuli. If the bell was held against the sides of the aquarium, or in the water near the animals, the vibration could be plainly felt by the fingers, yet no reactions on the part of the crayfish were observed. A metal snapper making a crack like a small pistol-shot was tried both in and out of the water but with no success in producing a reaction. A large hand tuning-fork, when held with its base pressed firmly against the glass walls of the aquarium, gave a deep rich tone of great volume, or when lightly touched to the glass produced a shrill, piercing, penetrating sound which was extremely sharp and disagreeable. Here again the vibrations of the water were quite perceptible to the hand at a distance of 10 cm., yet in neither case was there a sign of a reaction. Finally two electric tuning-forks, one of 256, the other of 512 vibrations per second, were tried on the animals taken one by one, and especial attention was given to the regular movement of the little thread-like appendages which keep up the current of water to the gills, with the idea that perhaps their rate of movement might be affected. In no case was there the slightest movement that could be referred to vibration, although here again the tactile stimulus was very perceptible to the finger. None of these experiments, then, give any indication that the crayfish reacts to vibratory stimuli which to the human ear produce sound.

III. ROTATION EXPERIMENTS

It has been found that the higher vertebrates, on being rotated on a turn-table, exhibit all the symptoms which accompany the sensation of dizziness in man. The question arises, to what degree and in what manner do invertebrates respond to rotation? Schaefer, the first to take up this question, denied on rather meagre observations that Crustacea respond in any way to rotation on the turn-table. Kreidl showed that this statement was altogether too sweeping, that Palæmon reacts very definitely to rotation by running in the opposite direction. Bunting tried the crayfish, but all the rotation experiments resulted negatively, so she was led to confirm Schaefer's statement so far as the crayfish is concerned. Bethe found that Carcinus behaved in a very definite manner on being rotated, that during the rotation the animals ran in the opposite direction to that in which they were turned, and as soon as the motion ceased they began running in the other direction. Finally Lyon, while agreeing with Bunting that adult crayfish do not react to rotation, discovered that young animals two or three centimetres long react very prettily to the movement, going in a direction opposite to the turn. To confirm and if possible extend these observations on the crayfish was the purpose of the following experiments.

It was soon found that a great deal depended on the method of experimentation. None of the experimenters mentioned above gives any detailed description of the manner in which the experiments were carried out. One is left uncertain whether the animals were placed on the periphery of the turn-table or over the centre, whether in the former case they were set with their heads toward the centre or away from it, or placed at right angles to a radius, or whether they were merely set down in any chance fashion and whirled about. The same indefiniteness exists in most of the accounts as to how fast they were turned, and whether the experiments were performed in the air or in the water. Finally it is not stated whether the rotation was always in the same direction, or whether its direction was alternated.

The turn-table used in the following experiments was one that had to be turned by hand, so that it was impossible to regulate the speed accurately. The crank, however, was not attached directly to the rotating board, but was connected with it by means of a gearing so that one turn of the crank produced about ten turns of the table. This gearing gave a steadying effect to the motion so that the speed could be kept tolerably constant. A circular pan, about 15 cm. in diameter at the bottom with the sides slightly sloping outward, was set so that its centre coincided with the axis of the rotating table. It was in this pan that all the experiments were tried. Through various preliminary experiments to determine the most favorable speed, it was found that a rotation rate of over one turn of the table per second produced such a strong centrifugal force that unless the animals were set exactly over the centre they were swept off against the side of the pan in such a manner that it was difficult to decide whether the rotation as such had any effect upon their movements. It was finally decided that the best results were obtained from a rate of approximately one rotation in two seconds.

It soon became evident that when the larger and more sluggish crayfish were merely dropped in the pan and rotated there was no particular reaction. This was true whether the animals rotated were in the air or in the water. The smaller and more active crayfish, however, showed a decided tendency to run either with or against the direction of the rotation, especially when the experiments were carried on in the water. In no case was there any tendency to go in the opposite direction when the rotation ceased, except in so far as the animals were carried along by the water. To get a quantitative expression for these tendencies a more delicate method of experimentation was resorted to. If there was a tendency on the part of the active animal to move either with or against the rotation, such a tendency might also be supposed to exist in the sluggish animal, only in the latter the inertia was sufficiently strong to prevent its appearance. If, however, the animals should be set radially to the periphery of the pan, the tendency to go with or against the rotation would be exhibited in the direction in which they turned out of the radial position. For it was found that no animal would remain in that position for any great length of time. Two groups of animals were used for these experiments, five animals in each group, and the first group was selected from the smallest and most active animals, the second from the largest and most sluggish. Each animal was set in two positions, position I, with the head toward the centre, position II, with the head away from the centre. Each animal was given ten trials in each position, and the number of times it turned in a direction with the rotation is set down in the + column, the number of times it turned against the rotation is indicated in the - column. In general from 5 to 15 turns were necessary for the orientation of the animal, though sometimes the number ran up to 30 or 40. Each trial was made in the opposite direction to the preceding one, in order to avoid the formation of any habit in turning. All these experiments were carried out in water, the depth of which in the pan was about 4 cm. In order that there should be no difference between the velocity of the water and that of the pan, the table was rotated a few times before the animal was put in. As a check a series of experiments of 5 in each position was performed in the air on the more active group.

The following table shows the results of these experiments in rotation:

TABLE VIII. REACTIONS TO ROTATION

In Water In Air Sum I II Sum I II Sum Total Group 1 + - + - + - + - + - + - + - 44 5 5 10 5 15 1 4 5 1 9 6 24 49 2 8 4 6 6 14 5 1 4 1 9 7 23 56 3 7 8 2 11 9 5 3 2 8 2 19 11 62 4 6 10 14 6 2 3 2 3 4 6 18 12 64 2 8 9 1 11 9 1 4 1 4 2 8 13 17 Sum 16 34 31 19 47 53 9 16 7 18 16 34 63 87 Group 2 21 5 5 3 7 8 12 27 2 8 5 5 7 13 36 7 3 8 2 15 5 37 2 8 3 7 5 15 54 5 5 6 4 11 9 Sum 21 29 25 25 46 54 46 54 Sum Total 37 63 56 44 93 107 9 16 7 18 16 34 109 141

Examination of the table reveals great individual variation. Some animals, as nos. 44, 49, and 37, turn rather constantly against the direction of the rotation, while others, as nos. 56 and 36, are almost as constant in their movement with the rotation. On the whole we observe that for each group, and for Group 1 in both water and air, there is a slightly greater tendency to go against the rotation than with it. This tendency, strange to say, comes out much more clearly in the air than in the water. It is evident, however, from the variation exhibited that there is nothing very stereotyped or mechanical about the reaction. Mention should be made of the fact that usually (though not always) the animals not only oriented themselves with reference to the rotation, but moved forward in that direction as long as the rotation continued.

IV. GEOTAXIS, BAROTAXIS, AND TURNING

(1) Geotaxis. So far as my knowledge extends, no experimental work has been done to determine the geotaxis of decapod Crustacea. Most of the vertebrates are positively geotactic, while a great many of the invertebrates, particularly unicellular organisms, larvæ of moths and butterflies, slugs, etc., are negatively geotactic. Parker found that in the case of the Copepod, Labidocera æstiva, the females exhibited strong, the males weak, negative geotaxis. In the investigation of the geotaxis of the crayfish, two sets of experiments were undertaken. In the first the method of procedure was as follows:

On a level table before a window a board was so arranged that it could be set at an inclination of 5°, 10°, 15°, 20°, and 25° either toward or away from the window. Starting, let us say, with the inclination toward the window, each one of a group of five animals was placed on the board with the right side to the window five times. The board was then inclined the same amount away from the window and the process was repeated. The same procedure was carried out with the animals set with the left side to the window. The following table gives the results of this set of experiments.

TABLE IX. GEOTAXIS IN FRONT OF WINDOW

10 12 14 16 18 Totals + - ± + - ± + - ± + - ± + - ± + - ± 5° 14 5 1 11 7 2 11 9 13 7 8 10 2 57 38 5 10° 13 7 12 7 1 13 6 1 16 4 14 6 68 30 2 15° 17 3 14 6 16 3 1 17 3 9 11 73 26 1 20° 12 8 17 3 18 1 1 19 1 14 6 80 19 1 25° 18 2 19 1 19 1 17 3 16 4 89 11

From this table it appears that the crayfish is positively geotactic, and that the positive geotaxis increases regularly with the increase in inclination. As a check on these results another set of experiments was undertaken with different animals under different conditions. The board was placed on a level table in the centre of a darkened room, and the operator stood behind a screen so as to be quite hidden from the animals. In order to observe the orientation a 2 c. incandescent electric light was suspended directly above the spot where the animals were set, at a distance of 60 cm. above the board. Each animal of a group of five was set five times in each of four positions, viz., head down the incline, head up the incline, and at right angles to it with first the right and then the left side down the slope. The results were as follows:

TABLE X. GEOTAXIS IN DARKENED ROOM

41 46 48 51 64 Totals + - ± + - ± + - ± + - ± + - ± + - ± 5° 12 4 4 13 4 3 8 10 2 14 5 1 11 7 2 58 30 12 10° 14 5 1 13 6 1 11 8 1 13 6 1 14 3 3 65 28 7 15° 16 4 15 5 13 7 11 6 3 14 2 4 69 24 7 20° 16 4 19 1 16 4 20 17 2 1 88 11 1

It will be observed that Tables IX and X agree quite well in the main, and we may conclude that the crayfish is positively geotactic and that the positive reactions vary from 58% at 5° to 89% at 25°.

(2) Barotaxis. Verworn uses the term barotaxis in an inclusive sense to cover all pressure phenomena that can be classed under the sub-heads of thigmotaxis, rheotaxis, and geotaxis. It seems preferable to me to employ the term in a more restricted sense of reaction to pressure other than the pull of gravity, the flow of a current, or the contact with bodies. The following experiment with the crayfish furnishes us, I think, with a case in point.

A glass aquarium, 54 cm. long and 28 cm. wide, was so inclined that the water was 20 cm. deep in one end and 8 cm. deep in the other. A board was so anchored that one end rested on the bottom at the shallow end of the aquarium while the other end projected slightly out of and above the deepest water. The board was about 45 cm. long, so that its slope was very gradual. Nine animals were placed in this aquarium and observed for three successive days. If we denote the bottom of the deep end of the aquarium by A, the shallow end under the board by B, the shallow end on top of the board by C, and the end of the board at the surface of the water by D, the results of the observations were as follows: On the first day 1 animal was found at D, 6 at C, and 2 at B. On the second day 5 were at C, 3 at B, and 1 at A. On the third day 1 was at D, 4 at C, and 4 at B. Totals, 2 at D, 15 at C, 9 at B, and 1 at A.

While these observations were too few to base very positive statements on, the striking fact that only one animal was found at A, the deep end of the aquarium, whereas 15 were noted on top of the board at C, indicates strongly that the animals avoid the deeper water. That the animals were found on top of the board, not under it, indicates that the observation is not to be referred to thigmotaxis, although the latter is doubtless very strong, as we shall see later. It should be observed that the negative barotaxis works against and overcomes the marked positive geotaxis which, as we have just seen, the animals exhibit in the air. Under the influence of the positive geotaxis, we should expect to find the greater number of the animals at A,--a condition which is speedily realized if we let the water run out of the aquarium. We conclude, therefore, that at certain pressures (specifically at the pressure exerted by water at a depth of 20 cm.) the crayfish is negatively barotactic.

(3) Turning. In the experiments with light it was observed that very seldom do the animals, when placed upon a surface, move off at once in a straight line, but usually they first turn through an angle of 90° or more and then start off straight. This came out strongly in the work on geotaxis, where oftentimes, when the animal was set with the head up the incline, the reactions would be preponderantly positive, whereas when set with the head down the incline the reactions were on the whole negative. In other words, when headed up the incline the animal would go down, and when headed down he would more often go up. Some experiments were tried under various conditions to determine how general this tendency is. The table presents the results in condensed form.

TABLE XI. EXPERIMENTS IN TURNING

Nos. 10 12 14 16 18 Sum 7 9 15 17 25 Sum I III 90°- 8 5 6 8 10 37 2 4 7 5 5 23 90° 4 6 7 2 2 21 3 1 4 90°+ 8 9 7 10 8 42 5 5 3 5 5 23 II IV 90°- 2 1 2 2 7 9 3 1 2 15 90° 1 1 2 4 3 5 3 3 1 16 90°+ 7 9 7 10 6 39 12 1 9 11 11 44

Nos. 41 43 46 48 64 Sum Sum V Totals 90°- 9 11 12 14 2 48 130 90° 2 1 2 1 3 9 54 90°+ 9 8 6 5 15 43 191

In this table the first line indicates the number of times each animal turned less than 90° when starting off from the position in which it was set, the second line the number of times the amount of turn was practically 90°, and the third more than 90°. The five parts of the table mark the different conditions; in Part I twenty observations were made on each animal placed on a level board before the window, and set now with the right now with the left side toward the window. In Parts II and III the animals were set with the head turned now toward now away from the window. In Parts IV and V the animals were placed on a level board in the middle of a darkened room with a 2 c. light about three feet above them. This was to exclude any possible directive influence of light. In all cases the operator was concealed by a screen. In Parts I and V twenty observations were made on each animal, in II and III ten, and in IV fifteen.

Rarely the animal would turn completely round and start off in the direction originally set, but usually the turn was between 90° and 180°. When once the animal began to move off, it would ordinarily keep to an approximately straight line. How seldom this was observed when the animals were first set down may be judged from the fact that out of a total of 375 observations in only 18 did the animals move straight ahead from their original position. From the table we observe that in over 65% of the cases (a proportion of almost two to one) the animals turned through 90° or more before starting off. At present the writer has no explanation to offer for this phenomenon.

V. THIGMOTAXIS AND TOUCH REACTIONS

(1) Thigmotaxis. Experiment has shown that there are some animals which tend to avoid contact with objects as much as possible, and on the other hand there are animals that seek to get as much of the surface of their bodies as possible in contact with objects. The former are spoken of as negatively, the latter as positively thigmotactic. Does the crayfish show any tendency in the one way or the other, and if so is it positively or negatively thigmotactic? In a large glass aquarium, 80 cm. long and 40 cm. wide, was a thin wooden box, 22 cm. long and 16 cm. wide, set in one corner 4 cm. from the glass walls. At the bottom of the box was an opening where the crayfish could enter. The following table shows the disposition of the animals for 27 different days, on which one examination was made each day. Line I indicates the number of times each animal was found against the walls of the aquarium, line II in the 4 cm. space between the box and the walls of the aquarium, line III inside the box against its sides, and line IV resting freely in the middle of the aquarium or in the middle of the box.

TABLE XII. THIGMOTAXIS REACTIONS

Animal 4 5 9 13 21 27 29 31 33 35 36 37 38 39 41 42 43 44 45 46 47 48

I 4 9 2 10 4 17 16 6 6 10 3 2 11 9 3 2 1 9 1 15 II 5 2 13 18 9 5 2 4 5 22 1 14 11 3 6 11 1 9 4 7 III 2 1 2 5 1 9 3 2 3 11 2 24 7 1 4 14 9 1 5 IV 1 11 2 1 2 4 4 1 2

Animal 49 51 52 54 56 58 60 62 64 Sum Totals I 6 11 3 1 2 4 7 10 12 195 II 8 3 7 5 8 1 3 2 1 190 III 1 16 9 5 10 3 2 2 154 IV 1 1 2 1 33 572

In order to appreciate the significance of the figures in this table it is necessary to consider the amount of lateral surface with which it was possible to come in contact in each case. In IV of course it was zero, in III it was 76 cm. with four corners in close proximity, in II it was only 38 cm. and one corner, but the space was so narrow that there was practically a contact-surface on both sides, and in I there was 212 cm. of lateral surface with three corners. I mention corners in this connection because they were almost invariably occupied. If we examine the table with these facts in mind, we find, (1) that the number of animals resting freely without contact with any lateral surface is very small, only about 6% of the whole; (2) that the number of animals found in the narrow space between the box and the walls of the aquarium is very large in proportion to the length of the space: indeed the animals were frequently found wedged into this space three or four deep; (3) that the number of animals found in the box was probably due largely to the fact that they found in it a greater lateral contact-surface, particularly in the corners, than was possible outside.

Two or three minor considerations are of interest. The animals were frequently observed "on edge" about half out of the water, that is, with the ventral surface of the body pressed against the vertical surface against which they were resting. This was also observed where the water was so deep that none of the members could touch the bottom. It was perhaps on account of the quality of the surface affording a rougher contact that so large a number of the animals were found in contact with the wooden box rather than the smooth, slippery surface of the glass. In the centre of the aquarium a wooden stopper 2 cm. in diameter projected about 15 cm. above the surface of the water. Very often a crayfish would be found almost at the top of this stopper, completely out of the water. This tendency to climb was frequently observed in the light-reaction experiments, where the animals would climb up on any piece of wood that chanced to be left in the box. It reminds one of the tree-climbing crabs of the West and East Indies. Along the creeks of Ohio I have frequently seen crayfish that had climbed up on logs or sticks that projected some feet out of the water.

In the table we see decided evidences of "habit" in the sense of an animal returning to the same place which it had occupied. No. 5 has almost half the observations in the open, nos. 21, 37, and 39 showed a decided preference for the space between the box and the aquarium wall, while nos. 38, 44, and 52 were more frequently found on the inside of the box. This recurrence to a particular position also came out in the light-reaction work, where an individual would return to the same spot in the box for days at a time as soon as released.

From the above considerations we conclude that the crayfish is strongly positively thigmotactic and that this thigmotaxis probably plays a most important part in the life of the animal.

(2) Touch Reactions. Lemoine investigated the reactions of crayfish to touch-stimuli and found that the plates of the telson, the sternal portions of the thorax, the abdominal pleopods, the chelæ, and particularly the antennæ toward their points are especially sensitive, but that nowhere, even on the back of the carapace, is a touch-stimulus altogether devoid of reaction. Gulland found that a needle could be inserted between the tufts of setæ on the chelæ without causing any reaction, but as soon as one of the hairs was touched, the chelæ closed with a snap. Considering the setæ as the organs of touch, he claimed to have found that the eyes, eye-stalks, and carapace (which he says have no setæ) are impervious to tactile impressions. This claim of Gulland's is strangely at variance with the facts. In no case have I been able to bring about retraction of the eye-stalk by visual stimulation, but a very light touch-stimulus on the eye itself or on the eye-stalk or a stronger stimulus on some portion of the head will cause the eye to be drawn in. It is true that after repeated stimulation the eye is retracted no longer, and with a heavy bristle one can make a perceptible indentation in the corneal surface without the eye being withdrawn.

The antennæ, from their anatomical structure, their position, and the manner in which they are carried, are generally considered the special organs of touch. Nevertheless, as far as the reactions of the animal are concerned, a stimulation of the antennæ by touch produces a less decided response than almost any other portion of the body. If the stimulus is very light no reaction at all is observed in most cases, and if stronger the antennæ are moved away, but that is all. A stimulation of the edge of the telson produces a more decided reaction. Either the animal folds it under the abdomen at once or faces about like a flash in an attitude of defence; frequently both reactions occur. While the response to stimulation of the chelæ was decided, that to touch on the first chelipedes was quicker and more accurate. The mouth-parts are also very sensitive to touch. I cannot agree with Gulland's assertions as to the insensitiveness of the carapace, for I have been able to find no place upon it where a light-stimulation would not produce a reaction. In this connection a curious phenomenon is characteristic of the animal. If the carapace or the front portion of the abdomen be lightly stroked with a solid object such as a pencil, the animal will slowly turn toward the stimulus on its antero-posterior axis. If, now, a like stimulus be applied on the other side, the animal will roll back through the normal position to a like inclination toward the stimulus on the other side. If the alternation be kept up and the change made quickly, a continuous and curious rolling movement is maintained, the animal growing more and more excited until it scampers off with a kind of cramp-like motion. With some animals this rolling reflex is more marked than with others, but in no case is it altogether lacking. Some animals have been known to roll so far over that they topple over on their backs. Dr. Yerkes informs me that he has observed the same phenomenon in a less degree with turtles when the edge of their shell is stimulated by scratching. The movement seems to be caused by the reflex stimulation of the extensor muscles on the opposite side of the body from the part stimulated. The thrust of the legs thereby brought about raises that side of the body and thus causes a rotation to some extent about the antero-posterior axis. But how was this connection between the stimulation of one side of the body and the contraction of the extensor muscles of the other side established? I have no doubt that it is intimately connected with the positive thigmotaxis described above. These animals live under loose stones for the most part, and thus the carapace gets a great deal of stimulation. If the animal is stimulated on one side, a contraction of the extensor muscles of the opposite side tends to roll the animal toward the source of the stimulus, and hence to increase the contact. In the race-history of the animal this has doubtless been advantageous in enabling it to escape the dangers of its habitat.

* * * * *

In a succeeding paper the writer hopes to discuss the reactions of the crayfish to chemical stimuli. In conclusion he desires to make acknowledgment to Dr. Robert M. Yerkes of the Harvard Psychological Laboratory for kindly suggestions and helpful criticism throughout the course of the investigation.

SUMMARY

(1) Crayfish are somewhat negatively phototactic, going away from rather than toward the source of light in the ratio of 62% to 38%. The different intensities employed in this investigation produced very little difference in the reactions. The average reaction-time was much less for the group of animals which showed the highest percentage of negative reactions, indicating a greater general sensitiveness. Variations of the position in which the animals were set affected the results very slightly.

(2) Previous confinement in the dark tended to increase slightly the number of negative reactions, and previous exposure to strong light tended to decrease the number, but the results were not constant. An increase in temperature tended to decrease the number of negative reactions to light, but here again the results were somewhat conflicting.

(3) Reactions to horizontal colored light showed a tendency to go to the colored light rather than to the white in the following order: Blue 47%, green 50.5%, black (or the absence of light) 51%, red 54%, yellow 59%. In the case of vertical colored light the comparison with the white resulted somewhat differently, as follows: Green 53%, yellow 53.5%, blue 57%, black 57%, red 72.5%. In the latter experiments the animals showed a marked and constant preference for the red.

(4) The animals showed no signs of reaction to static objects from visual stimulation, i. e., there is no evidence of visual perception of form in the case of stationary objects. Moving objects, especially large ones, are plainly perceived and definitely reacted to.

(5) There were no reactions whatever caused by those vibrations which to the human ear produce sound. So far as these experiments go, the animals cannot be said to hear.

(6) In rotation experiments individual animals were rather constant in moving either with or against the direction of the rotation, but no definite tendency for all animals was observed.

(7) The pull of gravity was followed with constantly increasing frequency from 58% at 5° to 89% at 25°. Therefore we conclude that the animals are positively geotactic. They are negatively barotactic, avoiding the pressure of water at the depth of 20 cm., and this is sufficient to overcome their positive geotaxis. When placed upon a level surface the animals show a peculiar tendency to turn through a greater or less angle before starting out in a straight line. In only 18 out of 375 observations, or 5%, did the animals start straight, in 30% they turned through an angle of less than 90°, and in 65% they turned through an angle of 90° or more.

(8) The crayfish is positively thigmotactic in a marked degree, as is indicated by the fact that in only 33 out of 572 observations, or less than 6%, were the animals found resting in the open, while in 190, or 33%, they were found in a narrow opening between two vertical surfaces.

(9) The animal is sensitive to touch over the whole surface of the body, but especially on the chelæ and chelipedes, the mouth-parts, the ventral surface of the abdomen, and the edge of the telson. If one side of the carapace or of the dorsal surface of the abdomen be stimulated, the extensors of the legs on the opposite side are contracted, and the animal turns on its antero-posterior axis toward the source of the stimulus. If opposite sides be stimulated alternately, a peculiar rolling motion is set up.

FOOTNOTES:

PRINTED AT THE RIVERSIDE PRESS H. O. HOUGHTON & CO. CAMBRIDGE, MASS., U. S. A.

* * * * * +----------------------------------------------------------------+ | | | Transcriber notes: | | | | Please note, underscore indicates italics, equals sign around| | the word indicates bold. | | Fixed various punctuation. | | P.47. 'to +5.93' may be 'to +:+5.93', however leaving it. | | P.142. The Standard for F calculation should be 110.40 total, | | not 122.40, with the average 3.94, not 4.35. Corrected table. | | P.47. 'for eyes at 0' added a degree like the others. | | P.272. 'abcab, cabe', changed 'cabe' to 'cabc 'as there is no | | 'e' and it looks to be 'c'. | | P.143. The Standards for Active M. Total calculation should | | be 65.40, not 59.40, with the average 4.67, not 4.24. | | The average for Total Passive M. should be 71.00 not 72.00. | | Corrected table. | | P.144. The Standards for F. Total calculation should be 120.11,| | not 117.04, with the average 4.45, not 4.33. The average for | | M. should be 4.19 not 4.20. Corrected table. | | P.145. Correct the net results on table IV. | | First table: | | Raised or lowered #25 2.1 to 2.7. #27 8.9 to 8.7. #28 3.5 to 4.| | Totals: 131 total is 132. 48.1 total is 47.9. 270 to 271.1. | | Net lowered 221.9 change to 223.2. | | Average lowering of each color judgment, 221.9/756 = .293 | | change to: 223.2/756 = .295 | | % of judgments of color not affected, 131/756 = 17+ is | | 132/756 = 17+. | | | | Second table: | | Raised or lowered, #8 36.3 is 36.4. #14 3.72 is 37.2. | | Total Raised 829.7 is 749.8, Net lowered 820.9 is 741. | | Average lowering of each tone judgment, 1.08+ is .98+ | | | | P.146. Table V. 2nd.--Lowered #18 8.2 is 8.6, change the | | total to 20.4. | | Net lowered change from 581.3 to 560.9 | | | | P.147. | |1st table: The Net + raised should be 640.5, not 649.5; with | | the 'Average raising of each tone judgment,' being 1.69, not | | 1.71. Changed in table. | |2nd table: -3.3 Net result lowered in #11 moved to | | #12 and is now +3.3. No figures affected. | | | | P.149. Table VII. 1st table. | | #15 Net result + 1 should be - 1. Moved. This does not affect | | negative figures. | | However, the + figure total is 197.5, not 186.5. | | Making the net raised 182.3. | | Averaging raising of each tone judgment, .45 is .48+ | | | | P.150. Table VIII. | | 1st table. #16. 1 Raised is 1 Lowered. Moved. #21 Raised 2.2 | | is 3.2, changed. | | Total Net Result needs to be changed from 11. to 12. | | Total Raised net result needs to be changed from 211.7 | | to 221.9, making Net raised = 209.9. | | Average raising of each tone judgment, .52 is .55+ | | | | P.151. Table IX. | | 2nd table. #13. raised 41.6 is 41.8. Changed. | | Total for Net Result Raised 130.0 is 150.2. . | | Net raised Grand total is 88.8. | | Average raising of each active touch judgment, .17+ is .22+ | | | | P.152. Table X. | | 1st table. #4 Net result raised, 4.3 is 4.6 = Total 57. | | Net lowered is 54.4. | | | | P.153. Table XI. First table. No. of colour 10-17 illegible. | | corrected. | | Points raised total & net raised 282.2 is 292.2. | | Average raising of each color judgment, .72- is .75-. | | | | 2nd. table. #2. Net Result - 13 changed to 11. | | Total Net Result - 56 is 54. | | Net raised, 104 is 106. | | Avererage .26+ is .27+. | | | | P.154. Table XII. | | First table. Total of column Raised, 226 change to 22 | | Total of column Lowered, 100 change to 101 | | | | P.313. 'obseved' changed to observed. | | Table I. Obs B row, 13.9 sec., taken out sec. | | | | P.400. (b) 'irrevelant' changed to 'irrelevant'. | | Footnote 128: 'Wahrscheinlichkeitseechnung' changed to | | 'Wahrscheinlichkeitsrechnung' | | P.463. moved 0 & 0 up one column in L. row of table from the | | percentage line | | P.499. Table. 23 Huggins. c. N. 98 first column '+55' changed | | to +'5.5'. | | P.550. Added closing quotation at: is markedly thinner." | | P.551. 'under expermental' changed to 'under experimental'. | | Footnote 141. 'Woods Holl' changed to 'Woods Hole'. | | P.583. 'thoughout' changed to 'throughout'. | | P.594. Fig. 4. 'The rise curve' changed to 'The rise on curve'.| | P.606. 'mimimize' changed to 'minimize'. | | P.638. Table XI. No. II, first row, 90deg: the minus was added.| | Note: Carat followed by { } indicates superscription. | | | +----------------------------------------------------------------+