GASES EVOLVED FROM GLASSES 



129 



drive off all the adsorbed gases as indicated by the curves in Fig. 1 



(200° for Nos. 1, 2, 3 and 4 and 300° for No. 5.) and the amount of 



carbon dioxide and permanent gases determined. The results are 



shown in Table III. 



TABLE III. 



Showing the Relation Existing Between Adsorbed Carbon Dioxide and Percentage 



Alkali in the Glass 



These data show in a remarkable way the relation existing between 

 the adsorbed carbon dioxide and the alkali content of the glass. This 

 experiment would seem to indicate that adsorbed carbon dioxide is 

 held by primary valence forces. Many other glasses must be tested, 

 however, before such a statement can be accepted as a fact. It seems 

 very plausible, however, to believe that carbon dioxide is taken up by 

 a film of sodium hydroxide which has been formed by the slow hydrol- 

 ysis of the glass. 



This coincides with the view of Kraus (17) who found, in the case 

 of a single glass which was high in soda, that no carbon dioxide was 

 adsorbed when glass and gas were both carefully dried and also that 

 on carbon dioxide is adsorbed by the glass if it has been washed with 

 boiling water to remove the film of alkali on the surface. The above 

 results do not necessarily indicate, however, that all gas held on the 

 surface of the glass is held by primary valence forces. Indeed a 

 determination that was made of the permanent gases given up by the 

 glass up to 200 °C. would indicate that these gases are held to the 

 glass primarily by secondary valence forces. Table III shows that 

 while appreciable quantities of these gases were evolved, no correla- 

 tion existed between these quantities and the chemical composition 

 of the glass. A similar determination was not made for water vapor 

 because the capacity of the McLeod gauge used in the experiment 

 was too small to take care of the large quantities of water vapor 

 evolved; if a determination had been made in this connection, it 

 seems highly probable that a definite relationship, existing between 

 the adsorbed water vapor and the alkali content of the glass, would 

 have been noted. 



