TABLE II. Results of cooperative work on the liter test for 1904 Dalican method of 



stirring Continued. 



A study of the results in Tables I and II leads to the following con- 

 clusions : 



(1) Different analysts do not obtain closely concordant results with 

 either method. This fact is doubtless due to several causes. Differ- 

 ences in room temperature affect results, as will be shown later in the 

 work of F. C. Weber, Thomas Gladding, and R. D. Oilar. The use 

 of different thermometers also affects the results, and the correction 

 given for the stem exposure is not strictly accurate for all thermome- 

 ters. But the average of all the results shows remarkable agreement, 

 indicating that the variations were as often high as low. 



(2) Individual analysts obtained very closely corresponding results 

 with the fatty acids prepared by either method, showing that under 

 the same conditions satisfactory agreement can be obtained, and that 



