210 PROCEEDINGS OF THE AMERICAN ACADEMY 



The Oxides of Cadmium, Mercury, Lead, and Bismuth. 



These oxides, the only other suspected ones which could be easily 

 analyzed by the method in hand, all yielded negative results. The 

 oxide of cadmium was distinctly crystalline, and contained only the 

 merest trace of gas. The oxides of mercury, lead, and bismuth ob- 

 tained by the ignition of the corresponding nitrates also appeared to 

 contain no occluded gaseous impurity. 



Unfortunately, the oxides of antimony, iron, and a number of other 

 metals, are not sufficiently soluble in acids to test with ease their power 

 of occlusion by this method. It seems probable that interesting results 

 might be obtained from them : hence in the near future other methods 

 will be tried here, with the hope of determining if these oxides also 

 occlude gaseous impurity. 



Theoretical Considerations. 



From the fact, observed vvith both copper and zinc, that oxides 

 which still contain a ti'ace of nitrates, as well as those made from the 

 carbonate, retain no imprisoned gas, it is readily inferred that the 

 decomposition of a trace of nitric acid is alone responsible for the 

 impurity. It is natural that this last trace of nitric acid should be 

 confined below the sui'face, whence the gases resulting from its ulti- 

 mate decomposition would find it hard to escape. On this supposition 

 it is not unnatural that zincic oxide which has been partly reduced, 

 and hence somewhat disintegrated, should contain less occluded gas 

 than that which has not been thus reduced. Moreover, since magnesic 

 nitrate is harder to decompose than the other nitrates, and the oxide 

 is more compact, we should expect to find more gas occluded in this 

 case than in the othei's. All these inferences agree with the facts. 

 The difference in the rate of expulsion of the oxygen and of the nitro- 

 gen is interesting, and less easy to explain. 



The negative results observed with a number of metals lead one to 

 conclude that the physical condition of the oxides in these cases was 

 so porous that even the last traces of nitrogen were allowed to escape. 

 Indeed, cupric and zincic oxide made from very finely divided basic 

 nitrates, obtained from aqueous solution, contained much less gas than 

 samples which were obtained in a more compact condition by the 

 direct ignition of the normal nitrate. This fact shows how much 

 depends upon physical conditions. 



It must be borne in mind that the occlusion of gases noted in this 



