FERTILIZER CONTROL STATION. 377 



ANALYTICAL* AND EXPERIMENTAL METHODS. 



In presenting the results of the analytical and experimental work 

 of the Station the purpose has been to avoid much reference to the 

 details of the methods adopted in securing these results, because 

 facts of this sort are chiefly of interest to the analyst and investi- 

 gator and have very little value to agriculturists who desire only to 

 know the practical bearings of an anal3'sis or experiment. At the 

 same time, it is proper for some statement to be made of the methods 

 used, so that those who desire so to do maj' intelligentl}- criticise the 

 work of the Station. This is briefly done in what follows, besides 

 which are some data and observations bearing upon certain methods 

 of investigation. 



■o' 



METHODS OF FERTILIZER ANALYSIS. 



The official methods as adopted by the Association of Official 

 Agricultural Chemists (see proceedings of third annual convention) 

 have been quite closely followed in the inspection of fertilizers. 

 The onl}' variation of any importance that has been made is in the 

 determination of potash. After the addition of barium hydrate in 

 excess, with subsequent heating, the solutions of ammonium 

 hydrate and ammonium carbonate are added without previous filtra- 

 tion, and without heating the mixed precipitates are allowed to 

 settle, when they are thrown upon a filter and thoroughl}' washed. 

 This causes a great saving of time as compared with two filtrations, 

 for not only is one filtration saved, but the mixed precipitates filter 

 much more rapidly and wash more easily' than the precipitate from 

 barium hydrate. Numerous comparisons of this method with the 

 one involving two filtrations have been made during the past two or 

 three years, with no appreciable diff^erence in results. 



Determination of Nitrogen. In all cases where admissible, the 

 Kjeldahl method of determining nitrogen has been used, and it has 

 been found to be time saving and accurate. It has repeatedly stood 

 the test of comparisons with the soda lime and absolute methods. 

 The modifications of the Kjeldahl method suggested by Arnold, As- 



* The anftly tical work of the Station has been entirely performed (luring the past year bj 

 Mr. Bartlett and Mr. Merrill. 



