191«J AGKICULTTJBAL CHEMISTRY — AGBOTECHITT. 803 



entire sample, not on an aliquot portion of tbe solution as in Steel's method 

 and, as a consequence, the error in dealing with small amounts is thus ma- 

 terially decreased. (3) The tedium of washing the precipitates Ls to a great 

 extent avoided without any sacrifice of accuracy." 



IV. The detcrmdnatlon of aluminum in the presence of iron and orpanic mat- 

 ter, M. Steel (pp. 173-182).— The author concludes that his method yields ac- 

 curate results for aluminum when care is taken in its manipulation. " The 

 method proposed by the U. S. Government Committee on Research and Ana- 

 lytical Methods" is very similar to the method used by Steel. This method 

 yields accurate results for aluminum, both in aqueous solution and in the pres- 

 ence of large amounts of organic matter. The Schmidt and Hoagland method is 

 as accurate as the other two methods, ai»d has the advantage ... of involving 

 fewer manipulations." 



V. The deter mniation of aluminum in hiolofficml material: A comparison of 

 the method of Steel {modified hij Kahn) with the method of Schmidt and Hoag- 

 land, C. A. Smith and P. B. Hawk (pp. 183-188).— The method of Schmidt and 

 Hoagland was found by the authors to be more satisfactory than the method of 

 Steel as modified by Kahn in work previously noted (E. S. R., 27, p. 2G8) for the 

 determination of aluminum in blood and in gastric juice to which known 

 amounts of aluminum had been added. In determining the metal in gastric 

 juice by the Schmidt and Hoagland procedure it is indicated that the material 

 should be ashed before the determination is made. 



VI. Comment on the data in the preceding papers (II-V) on the best avail- 

 nhle method for the quantitative determination of aluminum in biological mate- 

 rials, W. J. Gies (pp. 189-194). — From a critical examination of the data pre- 

 sented in the previous articles the author concludes that " the Schmidt-Hoag- 

 land method is somewhat more accurate and serviceable than the Steel method 

 for the estimation of alimiinum in biological materials, and that the Steel 

 method gives low results." Further comments on the work (reported in the 

 previous papers) by the various authors are included. 



VII. A direct test of the degree of accuracy of the Schmidt-Hoa gland method 

 for the quantitative determination of aluminum, A. K. Balls (pp. 195-202). — 

 Experimental data submitted show that " the Schmidt-Hoagland method for the 

 determination of aluminum gave results which involved a loss of as much as 7 

 per cent of the available aluminum, but which was usually about 4 per cent. 

 The losses appear to have been due, in the main, to the formation of AUOs from 

 AlPOi, in the precipitate of the latter dtiring Ignition, but also partly to the 

 solubility of AIPO4 in the reagents and washings. The material, as finally 

 weighed, is not wholly normal orthophosphate of aluminum, but contains less 

 phosphoric anhydrid than does the same weight of orthophosphate. The indi- 

 cated error might invalidate the method for accurate determinations of rela- 

 tively large amounts of aluminum. For comparatively small quantities, how- 

 ever, the error appears to be negligible." 



See also the first paper of this series by Gies, noted on page 860. 



The determination of bromin and iodin in the presence of chlorids, L. W. 

 WiNKLEK {Ztschr. Angew. Chem., 28 {1915), Aufsatzteil, Nos. 96, pp. 477-^80, 

 figs. 2; 98, pp. 494-496, fig. 1). — It is shown that by the use of potassium per- 

 manganate bromin can be accurately determined in the presence of chlorids. 

 The procedure for the determination varies according to the amount of bromin 

 present. For the accurate determination of very small amounts of bromin the 

 free bromin extracted with carbon tetrachlorid is titrated with arsenious acid, 

 using an aqueous iodin solution as indicator. 



"Jour Indus, and Bngin. Chem., 7 (1915), No. 5, p. 446. 



