[Vol. 6 



66 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



parison by eye cannot easily be brought within 0.1 in the 

 hydrogen exponent. With a colorimeter it is perhaps possible 

 to obtain results agreeing to 0.01, but such an accuracy is 

 unnecessary and is not practicable in ordinary tests of acidity 

 or titrations." Our attention, however, was not directed to 

 this fact until after the completion of our method, and in- 

 deed only after a careful search of his book with the idea of 

 determining whether such references could be found in re- 

 lated literature. 



In this work a complete Kober ('17) nephelometer-colori- 

 meter was employed, as this instrument happened to be at 

 hand. In reality, it possesses two distinct advantages, 

 namely, uniform and effective source of light, and protection 

 from side illumination. It was realized that since the colori- 

 meter was only required in the study of colored solutions, the 

 important factor in this case was to apply effectively the 

 method of shield solutions. This was ultimately accomplished 

 so satisfactorily that the defects of the comparator method 

 were entirely obviated, while all the advantages of the colori- 

 meter were retained. 



The method consisted simply in arranging for each side of 

 the colorimeter a pair of cups slipping to a certain depth 

 (noted later) one into the other, as shown in fig. 1. The 

 method of procedure is then as follows : For the left-hand set, 

 or column, water (or colorless standard solution) is used in 

 the outer cup, and the colored test fluid plus indicator in the 



inner cup. After adjustment, this set is not removed from 

 the colorimeter during an observation. In the case of the 

 right-hand set the outer cup contains the colored test fluid, 

 while the inner cup is for the standard solution plus indi- 

 cator. This set is placed on the right for convenience, as it 

 may be necessary to compare with the test fluid a series of 

 standards until an exact match is obtained. A rough com- 

 parison is of course made before selecting the standard solu- 

 tion for comparison. In each case the column must contain 

 an equal depth of colored test solution and of standard or 

 colorless liquid, the indicator being in the standard in the 

 one case and in the test solution in the other. There are no 



