2 Jones, Action of Alkalies on Glass and on Paraffin. 



The method is now generally attributed to Petten- 

 kofer,* who in 1858 described a process identical in 

 principle with Dalton's. Pettenkofer, who appears to 

 have had no knowledge of Dalton's work, employed a 

 special form of bellows, used lime water to absorb the 

 carbon dioxide, and estimated the excess by means of 

 standard oxalic acid, using turmeric paper as indicator. 

 Subsequently he substituted baryta for lime water. l?oth 

 these methods involve contact for .some time between the 

 glass bottle and the alkaline liquid, and it is obvious that 

 if the latter diminishes in strength owing to its action on 

 glass, the accuracy of the carbon dioxide determination 

 must be interfered with. 



On this point there is considerable difference among 

 chemists who have used the method, some affirming that 

 the action is immaterial, others, that it greatly affects the 

 accuracy of the method. Thus, Reisetf states that the 

 glass is attacked, but that the titre of the solutions is not 

 affected, and Dr. T. E. Thorpe informs me that he " has 

 records of the titre (taken once a week) of baryta solutions 

 which have been standing in stock bottles for upwards of 

 two years, which show absolutely no indication of such a loss 

 of alkalinity as is made out." On the other hand, Spring^ 

 maintains that the glass is attacked and that the strength 

 of the baryta diminishes. Ebermeyer§ states that baryta 

 is removed from the solution and combines with the silica 

 of the glass, and Messrs. Letts and Blake,'; in an elaborate 

 and important memoir, point out that the action on glass is 

 so serious that it is necessary to use vessels coated with 



* Che»i. Soc. Journal, lO (1858), p. 292. 



iA7Wales Je Chiwie et de Phys. [5] 26 (1882), p. 175. 



+ Men:. Acad. roy. BeJgiqiie, t. 37, p. 73. 



%Die Beschaffenheit det Waldluft, Stuttgart, 1885. 



II Proc. Roy. Dublin Soc, Vol. IX. (N.S.), Part II., No. 15. 



