CONVENTION OF OFFICIAL AGRICULTURAL CHEMISTS. 275 



pota«h, sulphate of potash and iiiaguesia, and kainit(with and without 

 removal of liuie). The results are calculated both by the old and new- 

 factors.^ The results were uniformly favorable to the Lindo-Gladding 

 method, which apparently gave results Very closely approximating- 

 theoretical amounts of potash in the salts used, especially if the (looch 

 crucible was used. Apparently there was no advantage gained by 

 removal of lime in the case of kainit. 



No important changes in the method were adopted. The reporters 

 for the ensuing year, however, were instructed to carefully revise the 

 phraseology of the method, to test the Stassfurt method on pure salts 

 with and without the addition of the impurities ordinarily found in 

 commercial potash salts, and to test the proposed modification of the 

 optional method in which sulphuric acid is precipitated in acid solution. 

 It was decided to omit precipitation of lime in the case of kainit, and it 

 was recommended that in the inspection of fertilizers a test be made for 

 chlorin. In view of recent criticism by Vogel and Haefcke^ of the 

 Lindo-Gladding method the reporters Avere instructed to prepare a 

 paper for publication in some leading foreign journal, carefully explain- 

 ing this method as adopted by the Association. 



A paper on " Some sources of error in the determination of potash in 

 fertilizers" was presented by M. A. Scovell, A. M. Peter, and H. C. 

 Curtis. The investigations reported in this paper indicate that in pre- 

 paring the water solutions of potash in mixed fertilizers containing 

 various porous substances a sutticient amount of the potash is occluded 

 to affect the accuracy of the determination. This is true to a greater 

 extent when there are bulky iirecipitates of phosphate. It also 

 appears that on evaporating solutions of i)otash containing soluble 

 phosphates to dryness in porcelain dishes a certain amount of the pot- 

 ash combines with the phosphate to form compounds insoluble in water. 



ISoils and ash. — A. Goss iiresented the report on tests of methods of 

 analysis of soils and ash. Four chemists took jiart in this work, and 

 the results of determinations of i)hosphoric acid and potash in soils of 

 known history by digestion with } normal hydrochloric acid, calcium 

 chlorid, citric acid and oxalic acid ; ordinary citrate solution diluted 1 

 part to 30 of water; and ammonium oxalate(16gra. per liter)are reported; 

 together with the results of comi)arative tests of digesting 200 gm. of 

 soil with 1, 2, 3, 4, and 5 liters of i normal citric acid. The use of ^ 

 normal hydrochloric acid, calcium chlorid, etc., was referred to the 

 reporter for further study. 



The following modification of the method of determining total nitro- 

 gen, proposed by the reporter, was adopted : Place 7 to 14 gm. of the 

 soil in an ordiiiary Kjeldahl Hask and boil 1 hour with 30 cc, or as much 

 as is re(iuired, of concentrated sulphuric acid and 7 gm. of yellow oxid 

 of mercury. Cool, add permanganate until the color is permanent, 



' The new factors jn-oposed are for KCl, 0.3069 ; for K.:SO^, 0.3587 ; for K,0, 0.1939. 

 ^Landw. Vers. Stat., 47 (1896), No. 2-3, p. 97. 



