702 EXPEETMENT STATION EECORD. 



various factors which influence the results in nitrogen determinations in gen- 

 eral, and especially those in milk analysis, are discussed very completely. 

 Special stress is laid on the error due to the nondecomposition of the vola- 

 tile fatty acids, the adhering of these bodies to the sides of the oxidizing flask, 

 and their final transmission into the distillate. The author prefers using 

 copper sulphate instead of mercury, which gives low results in milk analysis. 

 Sodium hydroxid is discarded in favor of potassium hydroxid, as the latter 

 solution boils more freely, very seldom bumps, and does not necessitate the 

 addition of zinc. 



The method proposed is as follows: Weigh off 10 gm., or thereabouts, of 

 milk in a flask and heat with 25 cc. sulphuric acid. When the frothing ceases, 

 add 10 gm. of potassium sulphate, and oxidize until no color is present. Dilute 

 the solution with 50 cc. of water and filter through a hardened filter paper 

 to remove any unoxidized fatty acids. After this add just so much 60 per 

 cent potassium hydroxid as will neutralize the acid, allow the mixture to 

 cool, and then add the remainder of the alkali, which in all is about 175 cc. 

 Finally add 100 cc. of 17.5 per cent potassium sulphid solution. 



For the distilling process the author describes a new apparatus, in which the 

 distillation is carried out in about 45 minutes. The distillate is collected in 

 60 cc. of decinormal sulphuric acid, and subsequently accurately diluted to 

 250 cc. The nitrogen is estimated by adding 15 cc. of a 10 per cent potassium 

 iodid solution and 15 cc. of 4 per cent potassium iodate solution, closing the 

 flask with a ground glass stopper, allowing it to stand for 45 minutes, and 

 titrating back the iodin with decinormal thiosulphate solution. 



The determination of nitric acid with nitron, C. Paal and A. Ganghofer 

 (Ztschr. Anahjt. Clicm., J,S {J909), No. 9, pp. 5 J,5-55 5). —This is a verification 

 of Paal and Mehrtens' work (E. S. R., 18, p. 525) on the accuracy of the method 

 for the determination of nitrates in meat. The method is also found efficient 

 for solutions which contain dextrin, peptone, or gelatin. 



Detection of nitrates in the presence of bromids, Villedieu {Jour. Pharm. 

 et Chim., 6. ser., 30 {1909), No. 2, pp. 66, 67; abs. in Analyst, Sit {1909), No. 

 ^02, pp. 418, Jtl9). — Although the coloration produced by nitrates with sulphuric 

 acid and ferrous sulphate is masked by the color of the iron bromid if a bromid 

 is present, good results can be obtained even with solutions containing a small 

 amount of nitrate and much bromid by neutralizing the liquid, shaking it with 

 an excess of basic lead acetate, and allowing it to stand. The upper liquid is 

 then decanted and filtered. To the filtrate an amount of sodium sulphate is 

 added sutficient to precipitate all lead present, and this is then again filtered. 

 If 1 cc. of the filtrate is mixed with 1 cc, of sulphuric acid and 1 drop of a 

 reagent prepared by mixing sulphuric acid with a very small quantity of ferrous 

 sulphate, even minute quantities of nitrates give a reddish-brown coloration. 



Contributions to the micro-chemical analysis of the alkaline earth metals, 

 N. ScHOORL {Ztschr. Anahjt. Chem., .'/S {1909), No. 7-8, pp. JfO 1-4 15). —Many 

 general and special micro-chemical methods are described for barium, strontium, 

 and calcium. 



A new titrametric method for the determination of barium, E. A. Hill 

 and W. H. Zink {Ahs. in Chem. Ztg., 33 {1909), No. 70, Repert., p. 305).— The 

 method is based upon the precipitation of barium as an iodate and the reaction 

 of this substance on iodide of -potash solutions of linown strength. Sodium, 

 potassium, magnesium, and ammonium salts present in slight amounts do not 

 influence the reaction, but this is not the case with calcium and strontium. 



The estimation of lead in tin vessels, F. Knopfle {Ztschr. Untersuch. Nahr. 

 u. Genussmth, 17 {1909), No. 11, p. 670). — Methods such as Busse's do not yield 



