CHAPTER X. Radioactive Minerals and Their Ages

RADIOACTIVE MINERALS AND THEIR AGES

The rate of helium production from uranium and thorium—Lead the final product of the uranium family—Its accumulation in geological time—Lead and helium ratios as a measure of geological time—Assumptions to be granted—Value of analyses in deciding quality of material for estimating ratios—Necessity for fresh, stable, primary rock minerals—Strutt’s work on the helium ratio—The lead ratio—Boltwood’s collection of analyses—Examples, with geological data—The author’s work on the Devonian minerals of Norway.

As we saw in Chapter VII the α-particles which are emitted at certain points in the line of descent of the radioactive elements have been identified with helium. Fortunately the evidence is conclusive, for upon this identification depends the latest and most elegant method yet devised of measuring geological time.

When all the members of a genetically related series of radio-elements are in equilibrium, the transformation proceeds in such a way that an equal number of atoms of each element disintegrates in the same time. When an atom of uranium disintegrates it does so with the production of two atoms of helium. At the same time, as indicated in the diagram, page 190, each of six other members of the family also emit a single atom of helium. Consequently, the total number of helium atoms which are liberated in the course of the complete transformation of a single atom of uranium is eight. From this result and the remarkable measurement made by Rutherford and Geiger—that 3·4 × 10¹⁰ α-particles or atoms of helium are expelled per second from a gram of radium—it is possible to calculate that the annual production of helium from a gram of uranium in equilibrium with all the other products of the series is 10·7 × 10⁻⁸ cubic centimetres. Measurements of the ionising power of thorium and its chain of dependent elements may be utilised to calculate the same rate in the case of the thorium family. It is found that one gram of thorium is equivalent in helium generation to 0·26 gram of uranium. The details of these calculations will be found in Appendix A.

While there is little uncertainty in these indirect results, it is gratifying to know that Prof. Strutt, during 1909-10, verified them both by a direct appeal to experiment. In certain minerals which contain radioactive constituents, the evolution and accumulation of helium must have been steadily proceeding for very long periods. Before a direct determination of the rate of helium production can be made, it is evidently necessary to use material entirely free from that element. Strutt worked with richly radioactive minerals, from large quantities of which he expelled completely the accumulated store of helium. This was done by preparing solutions with every precaution to avoid the presence of undissolved particles of mineral, and afterwards boiling them till the helium was removed. The solutions were then put aside until a fresh supply of helium had been generated in sufficient quantity to be detected and measured. To isolate the gas from the solutions and accurately to determine its volume was obviously a matter of great experimental difficulty, so minute were the volumes dealt with. However, as the result of the exquisite delicacy of his methods, Strutt brought his experiments to a successful issue.

The minerals used were pitchblende and thorianite, the former containing uranium alone, and the latter both uranium and thorium. The annual production of helium per gram of the parent element (in radioactive equilibrium with its respective family) was found to be:

(a) in the case of uranium: 10·6 × 10⁻⁸ ccs. or 1·88 × 10⁻¹¹ grams. 10·7 × 10⁻⁸ ccs. was the calculated estimate.

1 cc. would therefore be formed in 9,600,000 years.

(b) in the case of thorium: 2·4 × 10⁻⁸ ccs.

1 gram of thorium is therefore equivalent in its rate of helium generation to 0·23 grams of uranium, the calculated estimate being 0·26 grams.

The remarkable concordance of these results with the theoretical requirements is an eloquent tribute to the refined methods and experimental skill with which the measurements were carried out.

Experiments carried out by Boltwood and Rutherford in 1911 afford an equally striking confirmation of the conclusions on which the theoretical results were based. They measured the rate of production of helium from radium, the latter being in equilibrium with its early disintegration products, three of which also emit α-particles. Radio-lead and polonium were completely removed. The annual evolution of helium to be expected was 158 cubic millimetres. The first direct determinations were made by Sir James Dewar in 1908, and his best results corresponded to 169 cubic millimetres. Boltwood and Rutherford arrived at a much closer agreement, their figure being 156 cubic millimetres.

These results cannot fail to inspire the conviction that our atomic theory of matter is essentially correct. We are in possession of two experimental facts. The number of helium atoms expelled per second from a gram of radium has been directly counted, and the volume of helium accumulated in a year has been directly measured. The number of atoms in a given volume of helium (at N.P.T.) can be deduced at once, and the calculation is independent of any underlying theory. Calculation gives 2·69 × 10¹⁹ atoms per cubic centimetre; the atomic theory demands 2·72 × 10¹⁹.

Although there can now be no doubt that helium is one of the stable disintegration products, yet there is no direct evidence as to the identity of the ultimate products in the direct line of descent. In the uranium series indirect evidence points to lead with a considerable degree of certainty, but the end product of the thorium family is still unrecognised. In every uranium-bearing mineral the parent element slowly breaks down, while the final product of the transformation accumulates at its expense. Hence, if lead is the favoured element it ought to be found in association with uranium in all minerals which contain the latter. Moreover, in minerals which can be proved to be of the same antiquity, the amount of lead per gram of uranium should be constant; further, in minerals of various geological ages the proportion of lead should vary according to the latter. A mineral which began its accumulation of lead in pre-Cambrian times should certainly contain more at the present time than one in which lead has been collecting only since, say, the Tertiary outburst of igneous activity. The same statements apply equally well to the case of helium.

In so far as these principles may be used conversely to test the identity of lead with the ultimate product, they lend every support to that important conclusion. Dr. Hillebrand, the leading authority on the analysis of uranium-bearing minerals, has never in the course of a long experience found uranium unaccompanied by lead. It was this constant association which led Boltwood, in 1905, to suggest the probability of a genetic relationship existing between these two elements. In 1907, Boltwood went farther and showed that for minerals of the same age the amount of lead for each gram of uranium, or the ratio Pb/U, was, in general, nearly constant. He collected all the best analyses of primary uranium minerals, but unfortunately he omitted to give the geological details of their occurrence. As will be seen in the present chapter, when the relative ages of the minerals are compared with their lead ratios, a striking proportionality discloses itself.

The evidence of atomic weights is also favourable to lead. The complete disintegration of an original atom of uranium may be expressed as follows:

U ➜ 3He + Ra ➜ 8He + Pb 238·5 3·994 226·36 3·994 207·08 ———————————— ————————————— 238·5 238·34 239·03

Atomic weights are appended to each symbol, and the totals, which should be equal, are placed underneath. The agreement is close, but not as convincing as one could desire. Two alternative explanations of the discrepancies are suggested. Either lead is not the final product, or the atomic weights of both radium and uranium are too low by about 0·5. Neither alternative can readily be granted, but it may be pointed out that since uranium is the heaviest known element and radium follows not far behind, any impurities whatever, with the exception of thorium, would have the effect of lowering the observed atomic weights.

Accepting the above equation as substantially correct, the mass of lead generated in one year from a gram of uranium can now easily be calculated. For eight atoms of helium, one of lead is produced, or, mass for mass, six and a half times as much.

Consequently, as a gram of uranium involves the annual production of 1·88 × 10⁻¹¹ grams of helium, the associated lead which remains must amount to 1·22 = 10⁻¹⁰ grams. If this rate were constant we could find how long it would take for any mass of uranium to become completely converted into helium and lead. However, the rate is not constant, but is proportional at every moment to the quantity of uranium remaining unchanged. As the parent element becomes exhausted, it disintegrates more and more slowly.

Now in very considerable periods amounting to hundreds of millions of years only a very small fraction of the uranium originally in existence is decayed. For this reason, if only a small proportion of lead or helium has collected in a mineral since it began its life-history, then no serious error will be made in assuming their rates of evolution to have been constant. In minerals which have been in existence for 400 million years the slowing down is only about 5%. If an appreciable error should arise in ignoring this decline, then in place of the present day percentage of uranium the time-average must be substituted (see Ap. A, p. 179); that is to say, the amount of uranium which, if it did break up at a regular rate, would evolve the same quantities of the ultimate products.

Having calculated this uranium average, Uₘ, with the necessary approximation, the total quantity of lead or helium accumulated in a mineral would then provide a direct measure of its age.

Using lead as the age-index, and knowing its percentage, Pbₜ, in the mineral, then the time it has taken to collect, i.e. the age of the mineral, Pbₜ, is given by Pbₜ/Uₘ × 8200 million years.

Using helium as the age-index, both thorium and uranium must be estimated. The amount of thorium may be conveniently expressed in terms of uranium, since the latter is four times as active in its helium production as thorium. The total equivalent quantity of uranium, Uₑ, is thus known, and thorium then need play no further part in the calculations. It is found most convenient to measure the amount of helium, Heₜ, as the volume in cubic centimetres per gram of mineral. In this notation the age is given by Heₜ/Uₑ × 9·6 million years.

The validity of this procedure evidently demands the granting of certain obvious assumptions. Our choice of suitable minerals will not only be limited by these considerations, but the reason for a particular choice will be justified. The assumptions fall under the following headings:

(a) That no appreciable amount of lead or helium was present at the genesis of the mineral.

(b) That no lead, helium, or uranium has subsequently been added or removed by external agencies.

(c) That no lead or helium has originated by any other radioactive process than those already suggested.

The first and second suppositions bring up the whole problem of the origin of minerals. It is possible, by means of a physical examination, to decide whether a mineral is of the same age as the rock in which it occurs, or whether it is older or younger. In igneous rocks, such as a granite, the majority of the minerals are of the same antiquity as the rock, that is, they date from the period of consolidation of the rock magma. The component mineral particles of most sedimentary and detrital rocks existed long before the strata were laid down, whereas the cementing materials by which they are consolidated are partly furnished by percolating solutions and therefore may be subsequent to the period of deposition. In the same category come those ore deposits which occupy the fissures and crevices of pre-existing formations.

In whatever way a mineral may occur, its history can always be traced back either directly or by conjecture to an igneous rock, and it is rarely that it is possible to go beyond the magma from which such a rock must have consolidated. And even if this can be done in exceptional cases, there lies behind still another magma to which the material can be referred. Consequently, the minerals of igneous rocks are regarded as primary or original, and here we may briefly consider a few facts relative to their crystallisation from a molten condition.

A molten rock is regarded as a solution in which the numerous constituents are dissolved one in another. Now certain of the constituents can only remain in solution provided the latter is, with regard to them, very dilute. That is to say, they are only slightly soluble in a solvent composed of the rest of the rock material. Consequently, substances of which these constituents form an essential part will, as a general rule, be the first to crystallise; as examples, zircon, sphene, and apatite may be cited. For the same reason it happens that the magma does not remain homogeneous, but rejects certain of the rarer elements which collect together in a subsidiary magma of peculiar composition. To this concentrate of exceptional constituents, the gases and water vapour expelled during solidification are also added, and serve to maintain it in a state of æquo-igneous fusion, even when the bulk of the rock has already crystallised. The residual liquors yield the minerals of pegmatites and of drusy cavities. Certain minerals which are conspicuously rare in the body of the normal rock, are often developed on a large scale in pegmatite dykes, and it is from these that the most perfect and beautiful crystal forms are generally obtained.

Amongst the elements of limited solubility in a rock magma, uranium and thorium must be placed. The accessory minerals of ordinary igneous rocks, such as those already mentioned—zircon, sphene, and apatite—are rich in the radioactive elements when compared with commoner minerals like felspar and hornblende. In general, the richness of a mineral in uranium seems to depend on its position in the order of consolidation. The minerals first to be formed claim the greater part of the available store. The original surplus, unable to dissolve in the magma, is held over, and, should it be rich in radioactive ingredients, uranium- and thorium-bearing minerals may be formed during the later stage of pegmatitic intrusions. It is a striking fact that, as primary constituents, these minerals invariably occur in pegmatites associated with granite or syenite. As examples, pitchblende or uraninite, thorite, thorianite, and monazite may be mentioned.

We must now consider the part played by lead and helium during the genesis of minerals. Before the consolidation of the magma, both these elements must, of course, have been generated within it for an unknown period. As to the effect of physical conditions upon radioactive transformations it has already been shown (p. 102) that all the evidence points to the conclusion that these atomic changes are independent of the temperatures and pressures under which a molten magma exists. The helium already present at the time of crystallisation appears to behave physically in no way different from the other gases. There is no evidence that it tends to congregate in any particular mineral. A small proportion may be distributed through the resulting rock, but probably the larger share is expelled.

The lead which may be originally present follows a similar course. The metal is rejected, not only by the primary magma but, with rare exceptions, by the residual magma also. It finds no definite place in igneous rocks. Doubtless a certain amount of lead is retained in the molecular network of crystals, but that amount is not high. In the rocks of Leadville, Colorado, Hillebrand found an average of less than 0·002 per cent of lead. In the nepheline syenite of Southern Norway, using specimens free from minerals which one would expect to be comparatively rich in accumulated lead, the present writer was able to determine a percentage of only 0·0004. If, then, there should be initially a greater quantity of original lead, where are we to look for it? Probably the most of it goes to form lead-ores, such as galena. Separated from the pegmatites it appears in the later phases of ore deposition which follow on the heels of igneous activity. In company with hot gases, sulphide solutions, and a number of metallic companions, our lead is carried away and deposited in the fissures encountered by the mineralised waters. If the agency of magmatic gases appears to have been an important factor in the production of ore bodies, the origin of the latter is said to be pneumatolytic. Brögger has shown that in Southern Norway galena was one of the minerals to be formed in this way.

In the last phase of this complex series of operations, magmatic waters contribute their share to the filling of mineral veins, and it is amongst these hydatogenetic ores that galena is most usually found. The important point is that lead, for the most part, is drawn from the primary magma at, or perhaps before, the time of crystallisation, and it is not until the igneous activities have declined that it again appears in an active rôle.

Let us now consider the effect of the original distribution of lead and helium in a newly-formed rock. It will be clear that an analysis of the rock as a whole would give values of Pb/U and of He/U much higher than those corresponding to the period since consolidation. Of the total amounts of lead and helium, part would be originally segregated in the rock and part would be due to subsequent genesis. In most rocks, the former part is of sufficient magnitude altogether to invalidate the use of the ratios as age indices. This difficulty can be avoided by confining attention to particular minerals—indeed, to just those minerals which concentrate within themselves the radioactive parent elements. Within them lead and helium may accumulate to such a degree that the amount initially present becomes negligible. Zircon from the Devonian syenites of Southern Norway contains more than twenty times as much lead as the rock in which it occurs. Roughly, we may say that since the zircon came into being its content of lead has multiplied twenty times. Thorite may accumulate a hundred or a thousand times as much lead as it possessed at first. For the same reason, minerals like zircon and sphene often contain hundreds of times as much helium as the rock from which they are taken, and there is little possibility of error in assuming that they have themselves generated the whole of their supply.

Another difficulty, and a more serious one, must now be faced. Can we be sure that for periods of hundreds of thousands years a mineral has remained comparatively unaltered by external agencies? With regard to helium there is undoubtedly a tendency to escape. Strutt has demonstrated that when a radioactive mineral has been powdered, helium begins to leak, rapidly at first, then at a diminishing rate. Even crystals washed out of their original matrix showed a considerable leakage of helium. The observed rate of escape always exceeds the rate of generation, and it therefore follows that during the life-history of a mineral the conditions must be specially favourable to the retention of helium, for otherwise the latter could not have accumulated. Nevertheless, these experiments prove conclusively that the majority of minerals do not contain their full store of helium, and it is a matter for surprise that they contain so much. Consequently, ages deduced from the helium content of minerals can be regarded only as a fraction of the true age.

Dealing with lead and uranium, we must consider the tendency to alteration of the minerals in which they occur. From the surface down to the permanent level of the ground waters, rock material is subject to weathering. The more soluble constituents are leached out and complex silicate minerals are decomposed by the combined action of water, oxygen, and carbon-dioxide. It is in this belt of weathering that igneous rocks suffer most change. Many of the primary minerals are broken down and alteration products take their place. All the reactions involve considerable increase in volume, and not only are minerals altered in place, but material is carried away and deposited elsewhere as secondary minerals. Can we be sure that lead and uranium have remained untouched during this redistribution? In some cases we cannot, but fortunately for our purpose many of the most valuable minerals, like zircon, are dense and exceptionally stable. A mineral is only stable over a limited range of conditions. It adapts itself more or less readily to its physical environment. Certain minerals, however, are much more capable than others of withstanding great changes without undergoing metamorphism or alteration, and amongst these are many of the uranium-bearing minerals.

When there has been a migration of lead or uranium, an appeal to analysis will rarely fail to dispel the difficulty by disclosing the fact. It is inconsistent with the chemical properties of these elements that both should have been affected in the same proportion, and hence the ratio of lead to uranium obtained from different minerals of the same geological age affords an immediate test of the extent to which they have suffered from alteration in the course of their history. If the analyses give consistent results, it can be safely assumed that the effects of alteration have been inconsiderable; if there are marked discrepancies the results must be rejected as valueless from a chronological point of view. A microscopical examination of the minerals before analysis is a useful safeguard, for in this way altered material can often be detected. It is clear that reliable conclusions can only be drawn from minerals which are undoubtedly fresh.

Becker has criticised the method by directing attention to a suite of minerals from Llano Co., Texas. Their geological age is well defined. The Burnet granites with which they are associated are intrusive into a series of schists and quartzites, metamorphosed sediments of late Algonkian time. The Cambrian rocks lie upon this complex and the period of intrusion is therefore between two limits which are not very far apart. The lead ratios of these minerals are far from being constant, as the following examples show:

Yttrialite 1·15 Yttrialite 0·51 Mackintoshite 0·39 Uraninite 0·17 Fergusonite 1·04 Fergusonite 0·30

Boltwood found a satisfactory agreement in four cases, the ratio being 0·17, but, as he pointed out himself, most of the minerals from this locality are unsuitable, because of incipient or advanced alteration. The quartzose pegmatites in which the minerals occur are riddled with alteration products and secondary minerals, and the whole series is altogether unfavourable to accurate age determination. It is doubtful whether the apparent agreement of the ratios quoted by Boltwood ought to be accepted without further verification; for the present they cannot be regarded without suspicion.

This example shows how the actual results indicate the vicissitudes, varying from mineral to mineral, which the lead and uranium contents may have undergone. The method confirms or denies the validity of its application in every case. Judging from the relative solubilities of the constituents in question, uranium is likely to be abstracted from a mineral during the process of weathering more readily than lead, and consequently the age deduced from a weathered or altered specimen should in general be too high. A differential effect of this kind would account for the high ratios given by the Llano Co. minerals.

Strict attention must be paid to the question of origin, and secondary minerals avoided as carefully as altered primary minerals. Pitchblende is often secondary, e.g. when it occurs in veins with metalliferous sulphides. Other examples, of a rather different type, are autunite and carnotite. Secondary minerals are necessarily more recent than the rocks in which they occur, and many of them date back to no very remote period. Autunite is sometimes formed quite near the surface, within a few inches in fact. Its antiquity cannot therefore be more than a few thousand years, and in this time a detectable quantity of lead could not be generated. The traces actually found were probably in the original possession of the mineral. In an analysis made by the writer, only 0·06% of lead was found in specimens of autunite from Mozambique, where it occurs in bright green flakes attached to the large biotite crystals of pegmatitic dykes. In keeping with the age of the pegmatites and the high proportion of uranium—45%—it should have contained a hundred times as much had it been a primary mineral. The paucity of lead in autunite has even been put forward as an argument against the contention that lead is the ultimate product of disintegration of the uranium family. We now see how baseless is this argument when the origin of the mineral is remembered. It is not surprising that autunite should contain so little lead; on the contrary, it contains much more than the uranium can account for in the time at its disposal.

From these considerations it will be obvious that the only minerals to be chosen as material from which to determine the lead-ratio are fresh, stable primary rock-minerals. Having decided this, there remains a third possibility which might cast doubt upon the method. It can be objected that lead may originate as a product of some element other than uranium. Analytical results show clearly that thorium cannot give rise to lead, or a more proportionate relationship between these two elements would have announced the fact. There is also a possibility that certain of the longer-lived members of the uranium family may themselves be segregated in a mineral, independently of uranium. If so, they would gradually disintegrate, leaving no trace of themselves other than the residual helium and lead. In a magma containing 10% uranium, the radium would amount only to 0·0000034%. In actual magmas, even of pegmatites, the quantity present is always much less than this, and even if it be allowed that such tiny quantities may saturate the magmatic solution, the precipitation and concentration in any particular mineral would not be sufficient to leave an appreciable residue of lead.

The application of the accumulation of helium in minerals to the measurement of geological time, was first suggested by Rutherford in 1905, when he wrote: “I think that, when the constants required for these calculations are more definitely fixed, this method will probably give fairly trustworthy information as to the probable age of some of the radioactive minerals of the earth’s crust, and indirectly as to the age of the rocks in which they are found.”

During the years 1908-10, Strutt examined a great number of minerals, and determined the helium ratio whenever practicable. His first set of experiments dealt with phosphatic nodules and phosphatised bones. These may sometimes contain fifty times as much uranium as average rock material, and they have a further advantage in that they can be found in strata of nearly every age. As they frequently consist of fossils characteristic of the formations in which they occur, their age is well defined. However, the power of retaining helium is both poor and variable in the case of these phosphates, and the time relation is therefore obscured. Such materials never retain more than a small fraction of the helium which has been generated within them.

More suitable in their power of retention are certain iron ores, from which significant results were obtained. The helium ratio, and therefore the numerical age derived from it, showed a marked dependency upon the geological age of the mineral, as the following examples illustrate:

---------+---------------------+-------------- MINERAL. | GEOLOGICAL AGE. | MILLIONS OF | | YEARS. ---------+---------------------+-------------- Siderite | Upper Oligocene | 8·4 Hæmatite | Eocene | 30·8 Hæmatite | Upper Carboniferous | 141·9 Hæmatite | Devonian | 145·2 ---------+---------------------+--------------

Strutt next investigated the more compact minerals of igneous rocks, notably zircon and sphene. Zircon can be obtained from rocks belonging to several periods of igneous activity, and being a durable and stable mineral it is peculiarly fitted to retain the helium generated within it. Even allowing that the helium found does not represent the whole amount generated, it is unlikely that the fraction lost will vary as conspicuously as in the case of phosphates. In so far as that fraction depends on the structure of the mineral it is probably more uniform for zircon than for most other minerals. The helium ratio ought therefore to stand in a close relation to the geological age of the specimen. That it does so is clearly demonstrated by Strutt’s results, which are given below. The geological ages have been taken from the most recent literature, and are given in greater detail than those published in the original paper. Where two periods are bracketed together they are to be understood as referring to the limits between which the age of the igneous rock may fall.

-------------------------+---------------------+-------------- LOCALITY. | GEOLOGICAL AGE. | MILLIONS | | OF YEARS. -------------------------+---------------------+-------------- Mt. Somma, Vesuvius | { Recent | | { Pleistocene | 0·1 | | Mayen, Eifel | Pleistocene | 1·0 | | Campbell I., N.Z. | Pliocene | 2·5 | | Expailly, Auvergne | Miocene | 6·3 | | Brevig, Norway | Devonian | 54 | | Cheyenne Canon, | { Upper Cambrian | Colorado | { Archean | 141 | | Green River, | { Carboniferous | N. Carolina | { Archean | 147 | | Ural Mts. | Pre-Devonian | 209 | | Ceylon | Archean | 286 | | Blue Ground, Kimberley | Archean | 321 | | Sebastopol, Ontario | Archean | 622 -------------------------+---------------------+--------------

The results for sphene add but little to the above table, for most of the rocks from which workable quantities of sphene can be obtained are of pre-Cambrian age. The most notable helium ratio was from a specimen occurring in the Archean rocks of Ontario, and corresponded to an age of 715 million years. Summarising all the data afforded by Strutt’s work, we may graduate the geological column with a time-scale. It must be clearly understood, however, that the ages as expressed in years are, in the case of the helium ratio, minimum values only. How much greater the time represented by the geological periods actually is will appear from the ages as deduced from the lead-ratio. A few of these are placed in the table below for comparison.

-------------------------+----------------------------------- | TIME-SCALE IN MILLIONS OF YEARS. THE GEOLOGICAL SYSTEMS. +------------------+---------------- | HELIUM RATIO. | LEAD RATIO. -------------------------+------------------+---------------- Pleistocene | 1 | -- Pliocene | 2·5 | -- Miocene | 6·3 | -- Oligocene | 8·4 | -- Eocene | 30·8 | -- Cretaceous | -- | -- Jurassic | -- | -- Triassic | -- | -- Permian | -- | -- Carboniferous | 146 | 340 Devonian | 145 | 370 Silurian | } | } Ordivician | } | } 430 Cambrian | } 209 | -- Algonkian | } | 1000-1200 Archean | 710 | 1400-1600 -------------------------+------------------+----------------

The application of the lead-ratio to the measurement of the antiquity of minerals was first due to Boltwood. He tested its reliability by an appeal to the best analyses published up to 1907. Some of these, with geological details, will now be given.

In Glastonbury, and also in Portland, Connecticut, primary uraninite is found in the felspar quarries. The pegmatite in which the mineral occurs is associated with a granite which intrudes Lower Carboniferous strata. It is probably to be referred to the close of the Carboniferous period, and is certainly pre-Triassic. Five different specimens gave lead ratios in striking agreement, corresponding to an age of 340 million years.

It should be observed in all the following tables that in calculating the lead ratios the time-average of uranium has been used, and not the amount actually present. The statement of analyses is, of course, in percentage.

----------+-------+-------- URANIUM. | LEAD. | RATIO. ----------+-------+-------- 70 | 2·9 | 0·041 70 | 3·0 | 0·042 70 | 2·8 | 0·039 72 | 3·0 | 0·041 72 | 2·9 | 0·040 ----------+-------+--------

Similar crystals of uraninite have been furnished by the pegmatites of Branchville, Connecticut. The intruded strata are of either Silurian or Ordivician age, and the evidence suggests that the period of intrusion is possibly coincident with that of the earth movements which commenced at the close of the Ordivician. Here again the lead ratios closely agree, the age being 430 million years:

----------+-------+-------- URANIUM. | LEAD. | RATIO. ----------+-------+-------- 74 | 4·0 | 0·052 75 | 4·0 | 0·051 74 | 4·0 | 0·052 66 | 3·5 | 0·051 ----------+-------+--------

In North Carolina uraninite occurs in coarse pegmatites which are mined for their large flakes of mica. In this instance a good agreement is scarcely to be expected, as secondary products abound, and the three specimens from Spruce Pine which were examined by Hillebrand showed signs of incipient alteration. The fourth specimen was from South Carolina and this again lacked the freshness which is so essential. Zircon is also found in the North Carolina pegmatites, and the writer has examined two sets of specimens with the results given below. The material in this case appeared to be quite fresh.

The geological period is difficult to establish with certainty. The relations of the Appalachian rocks of Carolina are so obscure that the required age may be anywhere from pre-Cambrian to Carboniferous. Judging from the lead ratios, one is tempted to suggest that the age is not far from the Silurian, but until our time-scale is better determined one must be chary in this converse application of the method. It may confidently be hoped that in this way the geologist will be greatly helped in his attempt to unravel the history of igneous activity in the earth’s crust. But the time is not yet, for the accumulation of facts and data has only just commenced.

-----------+----------+--------+-------------- MINERAL. | URANIUM. | LEAD. | RATIO Pb/U. -----------+----------+--------+-------------- Uraninite | 77 | 3·9 | 0·049 Uraninite | 77 | 4·2 | 0·052 Uraninite | 67 | 3·3 | 0·047 Uraninite | 71 | 3·3 | 0·045 Zircon | 0·076 | 0·0036 | 0·046 Zircon | 0·130 | 0·0055 | 0·041 -----------+----------+--------+--------------

Other series of minerals could be given from the pre-Cambrian rocks of North America, Ceylon, and Mozambique, but it is unnecessary to multiply details further. In all cases the Archean seems to date from 1200-1600 million years.

We now turn to the pre-Cambrian rocks of Scandinavia and Finland, and for comparison with those of N. America the following classifications and correlations may be given. The chief unconformities are indicated by wavy lines. It should be pointed out that the correlation of the pre-Cambrian rocks over wide areas is, perhaps, the most difficult task the geologist has to attempt, and the scheme given opposite is based more on analogy than direct evidence.

Fennoscandia. North America.

Jotnian Keweenawan } } Jatulian Upper Huronian } } Upper Kalevian Mid. Huronian } Algonkian } Lower Kalevian } Lower Huronian } Bottnian }

Ladogian Laurentian } Katarchean Keewatin } Archean

It is not yet possible to support this correlation by a concordant system of chronology. Igneous rocks occur at most of the horizons, but it is extremely difficult to get samples of suitable mineral species for analysis. However, there is no doubt that in the future a definite time estimate will be attached to each of the above periods. When we are in full possession of this knowledge, and only then, will a reliable correlation of these rocks be possible.

The analyses collected by Boltwood include two groups which are of minerals taken from the pegmatites of Southern Norway, rocks famous for the occurrence of rare minerals. The first group, from the igneous complex of the Moss district, of which the average age is 1000 million years, is as follows:

------------+----------+-------+-------- MINERAL. | URANIUM. | LEAD. | RATIO. ------------+----------+-------+-------- Uraninite | 66 | 8·4 | 0·12 Uraninite | 68 | 7·8 | 0·11 Annerdödite | 15 | 2·2 | 0·135 Uraninite | 66 | 9·3 | 0·13 Uraninite | 57 | 8·0 | 0·13 Uraninite | 65 | 8·8 | 0·135 Uraninite | 68 | 8·8 | 0·12 Uraninite | 76 | 9·0 | 0·11 Thorite | 8·2 | 1·2 | 0·13 ------------+----------+-------+--------

The second group, from the complex of Arendal, demands an age of 1200 million years. The Scandinavian geologists believe that both groups of rocks are younger than the quartzites and other metamorphic sediments with which they are always associated, and that the latter rocks are of late Archean age. Sederholm, however, thinks they may be equivalent to his Kalevian division. The geological evidence, as far as it goes, does not point to any difference in the ages of these two sets of Norwegian minerals, but, on the other hand, there is no positive evidence that the ages are the same. An analogy made in ignorance cannot be held to constitute a proof. Field work in this case does not disqualify the testimony of the radioactive minerals; it rather invites their co-operation in the perplexing task of disentangling the intricate structural relations of the rocks.

----------+----------+-------+------------- MINERAL. | URANIUM. | LEAD. | RATIO Pb/U. ----------+----------+-------+------------- Uraninite | 56 | 9·8 | 0·14 Uraninite | 61 | 10·2 | 0·15 Uraninite | 56 | 9·4 | 0·15 Thorite | 9 | 1·5 | 0·16 Orangite | 7·5 | 1·2 | 0·15 Xenotime | 2·9 | 0·62 | 0·19 ----------+----------+-------+-------------

Amongst these rocks correlation is exceedingly difficult, and even their relative ages are hidden in obscurity. Högbom holds that the massifs of Moss and Arendal are contemporaneous with the Ser-archean granites of Sweden. Similar hyperites and quartzites are found in nearly every locality and the granites always appear to be younger than these. After their intrusion, an enormous thickness of rock was denuded away before the Jatulian sediments were laid down. The physical break here indicated is one of the greatest in the history of the earth, and undoubtedly represents an immense lapse of time. The granites and pegmatites must therefore be considerably older than the Jatulian rocks.

With a view to testing the constancy of the lead-ratio in a series of minerals from a single igneous complex, the author, in 1911, made a number of experiments on carefully chosen material. There occurs in the Christiania district of Norway, a geologically depressed area of nearly 4000 square miles, which is separated by faults from the surrounding pre-Cambrian rocks on every side. Within this area there is a nearly complete sequence of early Palæozoic rocks, surmounted by a few beds of red sandstone of Lower Devonian age. Over these beds and intercalated with them are lava flows; and finally, penetrating the whole mass, and representing a later phase of the same period of igneous activity, are great intrusions of plutonic rocks. Amongst the earliest of the intrusions is a series of thorite-bearing nepheline-syenites. Brögger believes them to be of Middle or Lower Devonian age, most probably the latter. The minerals occurring in them are, in many instances, notably radioactive, and thus they afford an admirable series in which to investigate the consanguinity of lead and uranium. A suite of minerals was obtained from Brevig, and estimations of these elements made, with the following results:

--------------+--------------+--------------+------- | URANIUM. | LEAD. | MINERAL. | GRS. PER 100 | GRS. PER 100 | Pb/U | GRS. MINERAL | GRS. MINERAL | --------------+--------------+--------------+------- Thorite | 10·1040 | 0·4279 | 0·042 Orangite | 1·2437 | 0·0570 | 0·046 Orangite | 1·1825 | 0·0542 | 0·046 Thorite | 0·4072 | 0·0196 | 0·048 Homelite | 0·2442 | 0·0121 | 0·049 Zircon | 0·1941 | 0·0085 | 0·044 Pyrochlore | 0·1923 | 0·0120 | 0·062 Pyrochlore | 0·1855 | 0·0093 | 0·050 Biotite | 0·1602 | 0·0069 | 0·043 Tritomite | 0·0631 | 0·0026 | 0·041 Freyalite | 0·0526 | 0·0028 | 0·053 Mosandrite | 0·0432 | 0·0024 | 0·056 Aegerine | 0·0253 | 0·0015 | 0·060 Astrophyllite | 0·0140 | 0·0007 | 0·050 Catapleite | 0·0132 | 0·0009 | 0·068 Nepheline | 0·0010 | 0·0004 | 0·400 Felspar | 0·0006 | 0·0003 | 0·500 --------------+--------------+--------------+-------

It will be noticed, that with a few exceptions, the value of the ratio increases as the percentage of uranium diminishes. This is probably due to the relative importance of lead originally entangled in the minerals at the period of their crystallisation. Thus it would seem in the case of nepheline and felspar that almost the whole of the lead found was originally present, while that which has since been generated is very small in comparison. Minerals with so little uranium contain too much occluded lead to be reliable, and are, of course, valueless in age-estimations. When sufficient uranium is held by a mineral, the lead generated becomes increasingly important, until the original amount is of negligible consequence. There is always the possibility that in some of the richer minerals larger quantities of lead were occluded than in felspar, but the agreement among the ratios renders this improbable. Rejecting all the results after that of biotite because of the low percentage of uranium, and omitting that of the first specimen of pyrochlore because the estimation could not be verified owing to lack of material, the mean ratio is 0·046. Replacing the uranium percentage by its time-average value, the ratio becomes 0·045 and the corresponding age 370 million years. It may be thought somewhat arbitrary to select certain results preferentially, but in view of the interpretation placed upon them the choice is not unfair, nor without justification.

Most of the available evidence drawn from radioactive minerals has now been passed in review. As yet it is a meagre record, but, nevertheless, a record brimful of promise. Radioactive minerals, for the geologist, are clocks wound up at the time of their origin. After a few years’ preliminary work, we are now confident that the means of reading these time-keepers is in our possession. Not only can we read them, but if they have been tampered with and are recording time incorrectly, we can, in most cases, detect the error and so safeguard ourselves against false conclusions.