624 



STATE BOARD OF AGRICULTURE. 



highest. With the alkaline solutions "C," "D" and "L" (the last being 

 the laboratory solution and presumably approximating "C" in com- 

 position) there is no marked consistency, the relative values fluctuating 

 irregularly. In the majority of cases, the differences scarcely exceed 

 those possible from experimental error. We can conclude from these 

 results that the choice o'i solution to be used in this work must be based 

 upon considerations other than the effect of its reaction on the solubility 

 of calcium phosphate. 



This must not be construed as an argument in favor of laxity in pre- 

 paring this reagent. While in the majority of cases a difference of 0.4 

 Ph may make no significant difference in the results, it is probable that 

 some materials are more susceptible to the action of alkali than others 

 and consequently, in order that uniformity of results may be assured, 

 the specifications for the reagent must be rigidly adhered to. 



This question has another phase which was alluded to by Luck (36), 

 Eastman and Hildebrand (15), Bosworth (7) and others, namely the 

 constancy of the reaction of the reagent during the determination. In 

 order to get information on this point, tlie filtrates from the citrate 

 digestions in the above determinations were collected separately from 

 the washings and their reactions determined as follows : 



Approximately 5 c.c. portions of the more or less highly colored 

 filtrates were transferred to test tubes and diluted to about 20 c.c. with 

 distilled water. In most cases this dilution served to reduce the color 

 sufficiently to permit an accurate comparison with the standard solu- 

 tions by means of the comparator. In some cases, however, it was neces- 

 sary to compensate for it by placing a second tube of the sample diluted 

 as above but without indicator, in front of the standard and one of 

 distilled water in front of the sample. Since the standard solutions 

 differed by intervals of 0.2 Ph the intermediate values in the tables are 

 of course simply estimates. The results appear in the following table. 



TABLE V. 



Sample No. 



None. . 

 3393 I ^ 



3356 I g 



3354 I ^ 



3346 I ^ 



3252 I g 



3230 I 'f^ 



3216 If, 



3088 { ^ 



3086 1 15 



2871 |g 



2667 |g 



2666 I g 



Solution 

 D. 



7.3 



7.2 + 

 7.4 + 

 7.4 + 

 7.4- 

 4 



4 + 

 4 + 

 4 

 4 



2 + 

 2 + 

 2 — 



1 

 3 

 3 

 3 

 3 



4- 

 3 



2 + 



2 + 



.2 + 



7.2 + 



