184 BELL SYSTEM TECHNICAL JOURNAL 



This shows that the square root relationship found by Sieverts can 

 be explained by assuming a dissociation of the molecules of gas some- 

 where in the process of solution. 



Donnan and Shaw ^^ applied this type of analysis to the solubility 

 of oxygen in silver and have shown that the solubility is proportional 

 to the square root of the gas pressure not only if dissociated gas is 

 dissolved as described above, but also if this dissolved gas reacts 

 with the silver to form a compound containing one atom of gas per 

 molecule, in this instance, Ag20. This conclusion can be reached 

 easily by extending the analysis in the preceding paragraph to include 

 an application of the law of mass action to the reaction between the 

 dissolved atomic gas and the metal. If this is done, it is found that 

 the concentration of compound is directly proportional to the concen- 

 tration of atomic gas, which has been shown to be proportional to the 

 square root of the gas pressure. Hence the concentration of compound 

 is proportional also to the square root of the gas pressure. 



It is apparent from this discussion that data showing the effect of 

 pressure on gas solubility do not show whether gases are dissolved in 

 metals as atoms or as compounds. In order to learn the state in 

 which gases exist in metals, Sieverts '^ studied the solubility of sulphur 

 dioxide in copper. He expected that changes in the pressure would 

 affect the solubility of this triatomic gas differently from that of a 

 diatomic gas. He was surprised, therefore, to find that for this 

 system also the solubility was proportional to the square root of the 

 gas pressure. In this instance, adherence to the square root law 

 could not be explained by assuming dissociation of the gas molecules 

 into atoms. 



In this field, Stubbs " made some interesting contributions after 

 those of Sieverts. He showed that the freezing point of copper 

 saturated with sulphur dioxide was depressed 2.54 times as much from 

 that of pure copper as should be expected from van't Hoff's freezing 

 point formula if the gas remained molecular in solution. If there 

 were complete reaction between the gas in solution and the metal, 

 the freezing point depression should be three times that for the exist- 

 ence of molecules only, and the amount of gas absorbed should vary 

 as the cube root of the pressure. Stubbs took the discrepancy in 

 these figures to indicate that about 70 per cent of the dissolved sulphur 

 dioxide reacted with the copper according to the equation, 



SO2 + 6Cu = CU2S -f 2CU2O. 



16 Uonnan and Shaw, Jour. Soc. Cheni. hid., 29, 987 (19tO). 

 13 Loc Clt 



1' Stubbs,' Jo«r. Chem. Soc, 103, 1445 (1913). 



