424 PROCEEDINGS OP THE AMERICAN ACADEMY. 



oxide was investigated by Rogers, probably occupies its expected place, 

 near cadmium, as far as its oxide's capacity to retain oxygen is con- 

 cerned. In some recent experiments in this laboratory, Cushman 

 succeeded in partially decomposing nickelous oxide in a stream of 

 nitrogen below 1200°. All these considerations are concordant with 

 the hypothesis devised to explain the escape of occluded oxygen for 

 metallic oxides. 



The only point regarding this occlusion of gases left unexplained is 

 the fact that in the experiments of Morse and Arbuckle the total amount 

 of gas was in every case nearly the same, although the proportion 

 of oxygen and nitrogen varied. This is a very interesting circumstance. 

 The experimenters ignited their oxide to constant weight, in other words, 

 until two successive ignitions caused no appreciable change ; but it does 

 not follow from this that the evolution of gas had ceased, — one can 

 only maintain that it had become so slow as to be inappreciable. Evi- 

 dently, then, the question to be answered is, — Why has the oxygen 

 more difficulty in escaping when a small amount of nitrogen is present 

 than when a large amount is present? The reason of this eccentricity 

 may be because those S|)ecimens which contain the most gas are natu- 

 rally the most porous and loosely held together, although usually formed 

 in larijer individual asreregations : and the freer atomic and molecular 

 motion allowed by the looser structure permits the oxygen to leave more 

 rapidly. Hence in each case the sum of the oxygen and nitrogen 

 amounted to about the same volume. This hypothesis within an hy- 

 pothesis is rather too uncertain to deserve much emphasis, however.* 



In the course of some of my old experiments on the impurities in 

 cupric oxide, it became necessary to reduce the substance in a current 

 of carbon monoxide. Referring to this work, Dr. A. Scott, in the 

 paper already cited,t expresses surprise at the existence of an admix- 

 ture of hydrogen in this gas. His experience agrees with that of 

 Stas in finding only a small amount of such gaseous impurity in car- 

 bonic oxide obtained from oxalic acid, while my results indicated a 



* In thisconnectionit is a matter of interest to note that Ramsay (Phil. Mag., (5.), 

 XXXVIII. 206 [1894]) has found that the presence of nitrogen in a palladium Inilb 

 seems to prevent to some extent tlie diffusion of hydrogen into it. This observation 

 is not unlike Morse and Arbuckle's; but if as a matter of fact an indifferent gas 

 can produce such an effect, our present notions concerning semipermeable septa 

 demand some revision. 



t Loc. cit., p. 563. 



