CHEMISTRY. 821 



of the deteriniiiation of nitrates, but tlieoretical results wi're obtained when tlie 

 chlorids were precipitated in the manner described. 



Synoptic tables for the chemical analysis of water, P. ( Joupil ( Tableaux 

 si/ni>j)lii/U('s ]>niir ridiali/xr (■hlitii(jnr ilf rrtiii. I'arls: ./. B. BaiUiere ct Si»i, 1901, pp. 70, 

 .//;/s. J0\. 



The quick determination of nitrates in soils, C. Montanari [Staz. Spcr. Agr. 

 Ital, 34 [1901), pp. 690-693; abx. in Chcm. C'entbl., 1901, II, No. 13, p. 793).— The 

 author adapts the method of Grandval and Lajoux, referred to above, for the deter- 

 mination of nitric nitrogen in air and water to this purpose. 



Methods of analyzing commercial fertilizers, F. Kretschmek (Ztxchr. Angew. 

 ( 'hnn .,U{ 1901 ) , X<>. 45, pp. 11.16-1138; ahx. in Clwin . Centbl. , 1901, II, No. 24, p. 1277 ).— 

 A brief account of metliods conuuonly employed in (Tennany for the determination 

 of nitrogen, pliosi)horic aci<l, ]»otash, and lime. 



The solubility of phosphatic manures in some organic acids, W. F. Sutherst 

 (Clirm. Nrirs, S4 {1901), No. :.'1S7, j>p. 199, 200). — Tests are reported of the action of 

 1 per cent acetic, tartaric, and citric acids on Christmas Island coj)rolite, basic slag, 

 basic superphosphate, and precipitated phosphate. In the method followed 1 gm. of 

 the phosphate was digested for 24 hours with frequent agitation in 100 cc. of the 

 solvent, phosphoric acid l)eing determined in 50 cc. of the filtered solution thus 

 obtained by the molybdic method. The solubilities obtained were very variable. 

 Acetic acid gave much lower results than the other two acids tested. The author 

 states that " though citric acid seems to be now fixed upon as the reagent in testing 

 phosphates, it is not at all a true representative, . . . for the most satisfactory results 

 as a general solvent are given liy tartaric acid." 



A suggestion regarding the modification and simplification of phosphate 

 analysis, M. 1'asson [Ztuchr. Auyctr. t'Jicin., 14 (1901), No. 45, pp- 1134-1136; nhg. in 

 ('hem. Centbl., 1901, II, No. 24, p. 1277) . — It is recommended that such weights of mate- 

 rial be used (12.8 gm. per liter or 6.4 gm. per half liter) that each milligram of the mag- 

 nesium pyiophosphate obtained shall correspond with a definite percentage (0.1) of 

 phosphoric acid, thus obviating the necessity of calculating the percentage. INIodifi- 

 cations l)ased on the same principle are suggested for the volumetric methods. 



On the separation and determination of small amounts of potassium in 

 salt mixtures, F. II. van Lkent (/.l.srhr. Analijt. CIk'iii., 40 [1901), \n. 9, itp. 569- 

 573). — For this purpose the author recommends, on the l)asis of his experiments with 

 sea water, a modification of the cobalt nitrate method of de Koninck and tiilbert, 

 converting the potassium finally into perchlorate or the platinum salt and weighing 

 in these forms. The conclusion is reached that the method is better adapted for use 

 with mixtures of i)otassium with other salts, especially when the proportion of potas- 

 sium is small, than the direct platinum or perchlorate method. The reagent 

 emplo}'ed is i)rei)ared by dissolving 9.58 gm. cobalt chlorid wUh 25 cc. of glacial 

 acetic aciil in 500 cc. of water. Just before use equal parts of this solution ami of a 

 solution of i)0 gm. of sodium nitrate in 500 cc. of water are mixed. 



Estimation of sucrose and lactose in condensed milk, S. II. K. and C. N. 

 Ruber [Ztsrhr. Anah/I. Chnn., 40 [1901), \o. 2, pp. .'y7-7/r>).— The metiiodof Kjeldahl 

 for e.stimating two sugars at the same time and in the same solution depends upon 

 whether the copper yielded by each sugar present is in a correct ratio to the total. 

 This method is correct when the two sugars reduce Fehling's solution with equal 

 rapidity. This is not exact with others, like invert sugar and lactose, and the authors 

 propose a formula for calculating an ai)i)roximate result. The figures they obtained 

 by the calculation agreed very closely with those appiie<l to :> samples of milk 

 prei)ared with known amounts of sugar. 



Determination of milk sugar by the Wollny refractometer in comparison 

 with analytical and polariscopic methods, K. Hhai x [Milrh /!</., .10 (1901), Nus. 

 37, pp. 578, 579; 38, pp. 596-599, Jig>i. 4; 39, pp. 613-616). 



