﻿94 BOTANICAL GAZETTE [February 



were put in the bottle, which was then corked; lo^^ of the solu- 

 tion to be tested were run from D, whose outlet was stuck 

 through the cork; then the outlet of the burette containing 

 oxalic acid was stuck throuQfh the cork and enougfh run in to 

 just neutralize. The acid was run in cautiously enough so that 

 the color disappeared with the final addition of a single drop; 

 this made it unnecessary to titrate back, though a burette of 

 KOH of the proper strength was always ready for use if needed. 

 As the diagram [^fig. 2) shows, every vessel of any kind con- 

 taining an alkaline solution can communicate with the air only 

 through the tower £, containing pumice soaked in KOH. 



For the idea of this method of titrating I am indebted to 

 Symons and Stephens-.'*^ When, as in this work, extreme accu- 

 racy is necessary, titrating in a flask closed with the thumb, as 

 is directed in manuals of quantitative analysis, yields serious 

 irregularities; and when the liquid to be tested is measured by 

 a pipette Into which it is sucked, the results are not even tolera- 

 bly uniform. Using the common method I often made five 

 determinations of the titre to get a '* reliable" average. The 

 first two are likely to agree entirely, and almost never vary by 

 more than one drop (0,05''''); when they agree I call any more 



tests superfluous. This accuracy is the more notable because 



N 



the acid used was . The usual method of determining- the 



100 ° 



CO2, in milligrams, has no recommendation except custom ; 

 expressing it in volume of a normal solution is in much better 

 harmony with modern chemical methods. My acid was practi- 

 cally 0.44,^ as strong as the most dilute one commonly used, of 

 which i*^'^ is the equivalent of i ™^ CO^. It permits therefore 

 more than twice the accuracy. 



The limit of positive accuracy in volumetric analysis is one 

 drop of the strongest solution in use. The directions extant for 

 making up the baryta water provide for its being several times 

 as strong as the acid it is titrated with. To get the benefit of 



the use of very dilute acid I have also used weak baryta water, 



N 



not more than twice as strongf, at the most, as the acid, 



^ ' 100 



4' Symons, W. H., and Stephens, F. R., Carbon dioxide. Its volumetric deter- 



mination, Journ. Chem. Soc. London 69: 869-881, 1896. 



^ ' H 



