RICHARDS. — RETENTION AND RELEASE OF GASES. 421 



copper from a cupric solution, aud since neither copper nor an oxifle 

 appeared adhering to the platinum negative electrode, it is safe to infer 

 that the conductivity is not electrolytic. While these oxides thus 

 conduct electricity without transferrence of material, it is undoubtedly 

 true that many substances, even glass for example, which are non- 

 conductors when cold, become capable of carrying a current eJectrolyti- 

 cally when heated. Melted salts are well known to be good electrolytic 

 conductors, and Graetz* has found that this conductivity begins before 

 the salts have melted. Our knowledge of the rationale of these facts 

 is altogether too incomplete to afford satisfactory explanation of all 

 the phenomena attending the passage of electricity through solids, 

 liquids, or gases at high temperatures; and it is perhaps unsafe to 

 record the conductivity of cupric oxide as a definite indication of the 

 temporary presence of metallic copper in the case under discussion. 



Following is a more certain proof that this dissociation is to be con- 

 sidered as analogous to the decomposition of limestone by heat. Here, 

 as there, the gas, in the present case oxygen, must have a definite 

 pressure of dissociation ; but no appreciable amount of metal or lower 

 oxide can form when the tension of the oxygen in the surrounding 

 atmosphere is greater than this definite pressure. On the other hand, 

 cupric oxide should lose not only its occluded oxygen, but all the 

 oxygen that normally belongs to it, leaving metallic copper (or at least 

 cuprous oxide), when heated in a perfect vacuum or in an inert atmos- 

 phere, if the hypothesis under consideration is tenable. This begins 

 even in the air at about 1,000°. The search in chemical literature 

 revealed a statement by Hilditch that cupric oxide ignited at lower tem- 

 peratures in a vacuum weighs less than it did before, as well as one by 

 Morley that the same substance slowly evolves a gas when heated in 

 a vacuum. f Since it is hard to tell whether this evolution might not 

 be merely that of the occluded gases, it was thought worth while to 

 repeat the experiment. 



Two grams of cupric oxide, which had been ignited for a long time 

 in pure air until constant in weight, were found to evolve a gas steadily 

 when heated in a vacuum to about the melting point of common salt 

 (790°), provided that the gas was removed by a Spreugel pump as fast 

 as it was formed. When allowed to attain equilibrium, this gas (which 



* Wied. Ann., XL. 18 (1890). 



t Hilditch, Chem. News, XLIX. 37 (1884). Morley, Am. J. Sci., XLI. 231 

 (1891). 



