Part 3

Where the difference between the maximum and minimum value is slight, the average of the separate determinations agrees closely with the number found by comparing the total substance used with the total product obtained. The latter method of calculation seems however to be preferable.

In the second series of eight determinations 28.6668 grams of cadmium bromide were equivalent to 22.7379 grams of silver.

Using the same values for silver and bromine, the atomic weight of cadmium = 112.245.

Maximum, 112.320. Minimum, 112.180.

The agreement of the separate determinations with each other is fairly close and the average of the two series of determinations is nearly the same. Huntington took great care in the purification of his material and in the carrying out of his method, which are strong arguments in favor of his work, yet his method is not as simple as could be desired where the nature of the work demands the greatest possible accuracy in all details and it also appears to be subject to some of the errors common to ordinary analytical operations.

Partridge. Amer. Journ. Science XL, 377. 1890. Methods: 1ˢᵗ. Decomposition of the oxalate to the oxide. 2ⁿᵈ. Reduction of the sulphate to the sulphide. 3ʳᵈ. Conversion of the oxalate into the sulphide. As an average of the determinations made by each method Partridge gives:

1ˢᵗ series, atomic weight of cadmium = 111.8027. 2ⁿᵈ ” ” ” ” ” = 111.7969. 3ʳᵈ ” ” ” ” ” = 111.8050.

An excellent agreement between results obtained by different methods.

That this very close agreement is only apparent has been shown by Clarke. He has found that the above calculations are based on the assumption that the atomic weight of carbon = 12, and that of sulphur = 32 when oxygen = 16. There seems to be little justification for this rather arbitrary selection by Partridge since the most refined work shows that whole numbers do not express the most probable atomic weights of carbon and sulphur in a system where oxygen = 16.

Amer. Chem. Journ. 13, 34. 1891.

The atomic weight of cadmium calculated from the total material used and the total product found in each of the three series is:

O = 16. C = 12. S = 32. At.Wt.Cd. 1ˢᵗ series, CdC₂O₄ : CdO = 12.66368g. : 8.10031g. 111.805. 2ⁿᵈ ” CdSO₄ : CdS = 15.93505g. : 11.02691g. 111.786. 3ʳᵈ ” CdC₂O₄ : CdS = 16.85228g. : 12.12906g. 111.806. difference, 0.020.

O = 16. C = 12.003 S = 32.059 At.Wt.Cd. 1ˢᵗ series, CdC₂O₄ : CdO = 12.66368g. : 8.10031g. 111.816. 2ⁿᵈ ” CdSO₄ : CdS = 15.93505g. : 11.02691g. 111.727. 3ʳᵈ ” CdC₂O₄ : CdS = 16.85228g. : 12.12906g. 111.610. difference, 0.206.

As Clarke has pointed out when those values are chosen for carbon and sulphur which are founded on the best experimental evidence the agreement between the different series of results as calculated by Partridge is somewhat modified.

I have repeated the work on which series I is based and would call attention to the following points in which it appears to have been experimentally defective.

1 The metal was only distilled twice in a vacuum. It has been found in this laboratory that perfectly pure cadmium or zinc can be prepared only by repeated distillations, each one being carried on slowly to allow the impurities to separate by means of their difference in volatility.

2 The supposed mixture of metal and oxide resulting from the decomposition of the oxalate was only moistened with a few drops of nitric acid in order to reoxidize any reduced metal. Unless the entire mass of metal and oxide was dissolved there would be danger of the presence of free undissolved metal which would possess an appreciable vapor-tension below the temperature of decomposition of cadmium nitrate. An appreciable loss in weight resulting from a distillation of the metal out of the crucible might easily result.

3 It seems very probable that the cadmium nitrate was not heated sufficiently to remove all traces of the oxides of nitrogen. I have found that this could only be accomplished by long continued heating. Constant weight was not sufficient to have decided this point since it was also found that this could be reached short of complete decomposition, if the temperature was too low to remove the last traces of these oxides. Some very delicate test for such oxides should have been applied at the end of each experiment.

The following table contains a summary of the results thus far obtained.

When two values are given for one series of determinations, the first is calculated from the total material used and the total product found, the second is an average of the results of the separate experiments. Oxygen is taken as 16 throughout.

Date. Investigators. At.Wt.Cd. 1818, Stromeyer, 111.483

1857, von Hauer, 111.935 } 111.940 }

1859, Dumas, 112.322 } 112.241 }

1860, Lenssen, 112.043 } 112.067 }

1882, Huntington, 1ˢᵗ series 112.239 ” 2ⁿᵈ ” 112.245

1890, Partridge, 1ˢᵗ series 111.805 ” 2ⁿᵈ ” 111.786 ” 3ʳᵈ ” 111.806

In the above calculation of Partridge’s results C = 12. S = 32. In the following carbon is taken as 12.003 and sulphur is 32.059.

1890, Partridge, 1ˢᵗ series 111.816 ” 2ⁿᵈ ” 111.727 ” 3ʳᵈ ” 111.610