CHEMISTRY. 819 



Wagner metbod and 18.48 and 18.45 by tbe author's method — thus show- 

 ing the great importance of manufacturers supplying slag in the finest 

 possible form. 



Humic acid dissolved 05.8 per cent of the phosphoric acid in very fine 

 ground slag, 30.38 per cent of that of Eedonda phosphate, and 43.3 per 

 cent of that in bone precipitate. 



The determination of potash as perchlorate, F. S. Shiver 

 {Jour. Amer. Gliem. Soc, 21 (1899), No. 1, pp. 33-42).— The solution of 

 potash, after freeing from sulphuric acid and nonvolatile acids, is evapo- 

 rated in a porcelain dish with an excess of perchloric acid until all 

 odor of hydrochloric or other volatile acids has disappeared, and the 

 white fumes of perchloric acid begin to appear. The sirupy solution is 

 then treated with alcohol containing 0.2 per cent of perchloric acid, the 

 precipitate washed by decantation, transferred to a Gooch crucible, 

 washed with pure alcohol, dried, and weighed. The method is pro- 

 posed as a check upon the commonly employed platinum chlorid 

 method. — l. h. merrill. 



A source of error in the determination of nitrogen in Chile salt 

 peter by the Ulsch method, L. Brandt (Chem. Ztg., 23 (1899), No. 5, 

 p. 22).— The author discovers an impurity in the iron used for his deter- 

 minations in the form of an organic nitrogen compound, the nitrogen 

 of which is converted into ammonia when the iron and sulphuric acid 

 are brought together in the process. The error thus occasioned, he 

 finds, amounts to as much as 0.8 per cent, when 0.5 gm. of Chile salt- 

 peter is used in the determination. He announces this discovery to put 

 other analysts on their guard. — j. t. Anderson. 



Some errors in the determination of nitrogen, C. G. Hopkins 

 (Jour. Amer. Chem. Soc., 20 (1898), No. 12, pp. 961-905).— Two common 

 sources of error were investigated, (1) the loss of ammonia in titrating 

 in an open vessel, and (2) the loss of ammonia from the receiving flask 

 in case there is not sufficient acid above the end of the delivery tube to 

 neutralize all of the ammonia distilled over. In the first case it was 

 found that when titration requires from 3 to 5 minutes the error from 

 loss of ammonia from the standard solution may become an important 

 factor, the total variation in the two series of experiments amounting 

 to 0.8 cc. or 3 per cent of the ammonia required. In tbe second case the 

 loss of ammonia was found to vary with the depth to which the deliv- 

 ery tube of the condenser dipped into the standard acid, in some cases 

 amounting to a considerable proportion of the amount of ammonia 

 theoretically required. 



Report on an investigation of analytical methods for distin- 

 guishing between the nitrogen of proteids and that of the simpler 

 amids or amido-acids, J. W. Mallet. Separation of proteid 

 bodies from the flesh bases by means of chlorin and bromin, 

 H. W. Wiley ( U. S. Dept. Agr., Division of Chemistry Bui. r>l, pp. 30).— 

 In the first part the classes of nitrogenous constituents in food and the 



