It also agrees fairly well with the figure 112.0706 which I obtained by the first method described. Advantages of the Method. The method possesses no advantage whatever over the one which involves starting with the element itself. The oxalate can however be obtained pure having pure metal. The salt is of definite composition when perfectly dry. The method as carried out avoided the contact of any foreign material with the salt after it was weighed. Disadvantages of the Method. 1 The avidity with which the dried oxalate takes up moisture from the air is an objection to its use for the determination of atomic weights. Even with the greatest care there is a slight element of uncertainty introduced from this source. 2 The oxalate is stated to decompose into a mixture of the oxide and metal. The temperature required for this decomposition is somewhat higher than the melting point of cadmium. The metal heated above its melting point possesses a vapor-tension and loss in weight must result, whatever precaution is taken in heating. This is the probable explanation why the results obtained by this method are lower than those of the preceding. A comparison of the two methods leads me to attach much more importance to the results of that one which establishes the relation between cadmium and cadmium oxide directly and I therefore regard the atomic weight of cadmium as very closely expressed by the figure 112.07 when oxygen = 16. Preparation of Certain Sub-compounds of Cadmium. Historical. Cadmium acts so generally as a bivalent element that it is usually regarded as entering into combination only where it can play this rôle. The only compound described, in which it has apparently a lower valence than two, was prepared by Marchand[10]. It was obtained by heating cadmium oxalate to the melting point of lead when a green powder remained behind which resembled chromium oxide. When heated on the air it appeared to be decomposed into metal and oxide. When treated with mercury the compound was not altered. An analysis showed it to have the composition represented by the formula Cd₂O. [10] Pogg. Ann. XXXVIII, 143. A. Vogel[11] has shown that the green powder described by Marchand consists of a mixture of the metal and oxide. When this mixture is treated with dilute acetic acid the metal remains behind as microscopic glistening globules. The lower the temperature at which the oxalate is decomposed the more oxide and the less metal were found in the product. There was then no compound known in which cadmium acted as if its valence was less than two when this work was undertaken.