Procedure
Enough cadmium oxalate for a determination was placed in a weighing tube which had been tared against a similar vessel and dried at 150°C. until the weight remained constant. It was then poured into a weighed porcelain crucible. The tube and its tare were now dried again at the same temperature to constant weight in order to avoid any error resulting from moisture being absorbed by the cadmium oxalate which adhered to the weighing-glass. The crucibles used in these determinations were arranged in the same manner as those employed by Morse and Jones in their work on this method. A small porcelain crucible on whose edge were placed three short platinum wires bent in the shape of the letter U, was placed in a larger porcelain crucible. The platinum wires prevented the lid from sticking to the crucible after heating and also allowed the products of decomposition to escape. The glaze was removed from the outside of the larger crucible with hydrofluoric acid to avoid sticking when heated to a high temperature. A second pair of crucibles arranged in the same manner was tared against the first one and in all cases treated like it. After the oxalate had been poured into the weighed crucible, it was decomposed by placing the crucible with its contents in a cylindrical asbestus covered air-bath, and slowly raising the temperature until the mass beams uniformly brown in color. In the last five determinations, the temperature was not allowed to exceed 300°C and after from forty to eighty hours the loss in weight was about ninety percent of the amount calculated for complete decomposition. In the first four the temperature was much higher and the time employed shorter. After the oxalate had been thus treated nitric acid was added and the contents of the crucible dissolved completely. The crucible was then transferred to a bath constructed by placing a larger porcelain crucible in a still larger one of iron and filling the intervening space with sand. It was slowly heated until the nitric acid had all evaporated and the dry nitrate began to give off red fumes. The crucibles were then removed to a similar bath containing iron filings instead of sand. This bath was heated by means of a single burner as long as red fumes were observed, and then for about five hours with a triple burner. Finally, the crucibles were transferred to a nickel crucible in the bottom of which a plate of unglazed porcelain was placed. The nickel crucible which had previously been set tightly into a hole cut in an asbestos board was then heated over the blast lamp for two hours. After this the porcelain crucible and contents were weighed and then reheated for half hour periods as before until three successive weighings remained constant. This usually required from three to four hours of blasting. In all determinations, the resulting product was tested for oxides of nitrogen with potassium iodide, starch and hydrochloric acid, but none was found. All weighings were reduced to the vacuum standard on the assumption of 8.4 for the Sp. Gr. of brass, 21. for platinum, 3.31 for the oxalate and 8.15 for cadmium oxide. The results are:
| Cadmium Oxalate. | Cadmium Oxide. | At. Wt. Cd. | |
|---|---|---|---|
| I | 1.97674 | 1.26414 | 111.73 |
| II | 1.94912 | 1.24682 | 111.82 |
| III | 1.96786 | 1.25886 | 111.77 |
| IV | 1.87099 | 1.19675 | 111.77 |
| V | 1.98941 | 1.27242 | 111.79 |
| VI | 1.37550 | .87994 | 111.85 |
| VII | 1.33313 | .85308 | 111.95 |
| VIII | 1.94450 | 1.24452 | 112.04 |
| IX | 2.01846 | 1.29210 | 112.09 |
A glance at these results shows that there is a variation of .36 of a unit and that the atomic weight in general increases with the number of determinations. In the first four determinations, there may have been loss of cadmium by reduction and subsequent volatilization, but in the later determinations this is not probable. It is believed that the greater part of the variation was due to imperfect dehydration of the oxalate. This and other sources of error in this method will be referred to later. The nickel crucible used gave a slight sublimate on heating, even after fifteen hours’ blasting. This condensed on the porcelain crucible as a brownish coating but, as both the crucible and its tare were blasted for the same length of time, it did not seem to change the difference of their weights. More than a dozen nickel crucibles were tried but none was found not to give a sublimate. The amount was so slight that no attempt was made to determine its nature.