GASES IN METALS 185 



Stubbs also believed that Sieverts' data showed that the solubility of 

 sulphur dioxide in copper varied as some higher root of the pressure 

 (2.4 root) than the square root and that it supported his theory of 

 partial reaction of the gas with the copper. 



Because very little is known of the state in which dissolved gases 

 exist in metals, certain data reported by Franzini '^ are interesting. 

 Franzini believed that, if ionized gas existed in metals, it could be 

 displaced by the action of an electric field. His data on the variation 

 in electrical resistance, caused by application of an electrical field to 

 iron and nickel wires which previously had been saturated with 

 hydrogen, show that the absorbed gas can be displaced toward the 

 negative pole. Thus, evidence of the presence of ionized gas in a 

 metal was obtained. 



Illustrative Examples of Gas-Metal Absorption Studies 

 The Solubility of Oxygen in Silver 



Steacie and Johnson's ^■' careful experiments on the solubility of 

 oxygen in silver form a good illustration of the problems involved in 

 this type of study. Since their work yielded some of the best data 

 available on any gas-metal system, it is reviewed here in some detail. 



The principle of their method of determining solubility is a variation 

 of the method ordinarily used. A known weight of silver, contained 

 in a silica bulb, is heated to a given temperature. All traces of gas 

 are removed by evacuation of the apparatus, and then a known 

 amount of purified oxygen is admitted to the silica bulb. After 

 equilibrium between the oxygen and the silver is reached, the pressure 

 of the gas is measured by a permanently connected manometer. The 

 theoretical pressure of oxygen in the system, assuming none is absorbed 

 by the silver, is calculated from the quantity of gas introduced, its 

 temperature, and the volume of that part of the apparatus it occupies. 

 The difference between calculated and observed pressures is a measure 

 of the amount of gas absorbed by the metal. 



In order to determine the solubility of gases in metals accurately, 

 sufficient time must be allowed for equilibrium to be reached. Al- 

 though equilibrium was reached very quickly at the highest tempera- 

 ture (800° C.) used by Steacie and Johnson, several days were required 

 at the lowest temperature (200° C). It is obvious that this long time 

 greatly increases the possibility of errors from leakage of gas into the 

 apparatus. Since Steacie and Johnson were especially interested in 

 solubility at low temperatures, to be certain that no errors were 



1* Franzini's work was reviewed in Natwe, March 12, 1932, p. 404. 

 '^ Steacie and Johnson, Proc. Roy. Soc. London, 112, 542 (1926). 



