412 EXPERIMENT STATION EECORD. [Vol. 41 



must be employed, however, if the soils are lower than 0.5 to 0.6 per cent of 

 organic carbon." 



Aids to the analysis of food and drugs, C. G. Moor and W. Partridge ( Lon- 

 aon: BaiUivre, Tindall it Cox, 1918, J,, ed., pp. X I -\-268).— This book contains in 

 compact form information dealing with the analysis of foods, the more common 

 drugs, disinfectants, and soaps. In this fourth edition the subject matter has 

 been revised, and several new sections have been added. An appendix enu- 

 merates the alterations in the composition of food and drugs in England necessi- 

 tated by the war. 



An improved method for the quantitative determination of caffein in vege- 

 table material, F. B. Power and V. K. Chesnut {Jour. Amer. Chem. Soc, J^l 

 {1919), No. 8, pp. 1298-1306). — Inaccuracies in previous methods of determining 

 caffein are pointed out, and an improved method developed at the Bureau of 

 Chemistry, U. S. Department of Agriculture, is described and its accuracy 

 indicated by control experiments. The essential features of the method, the 

 technique for which is described in full, consist in the extraction of the caffein 

 by hot alcohol, treatment of the extract after evaporation of the alcohol with 

 magnesium oxid to liberate the caffein from its combinations and to remove the 

 tannin and other acidic substances from solution, hydrolysis of the hot water 

 extract of the above acidified with 10 per cent sulphuric acid to decompose 

 saponins, and finally extraction of the acid filtrate with six successive portions 

 of chloroform and recovery of the caffein by distillation of the chloroform after 

 the addition of a 1 per cent solution of potassium hydroxid to remove coloring 

 matter. 



The method is said to give results in close agi'eement with those obtained by 

 the U. S. Pharmacopoeia method of assay, and to possess some advantages over 

 that ofhclally adopted. 



Analytical data are presented on the yields of caffein by the gravimetric 

 method and as calculated from nitrogen determinations in samples of tea, 

 coffee, coffee leaves and pulp, and guarana. The pulp of the coffee berry 

 yielded about 0.89 per cent of caffein, an amount which is thought to render its 

 utilization for the production of caffein worthy of consideration in view of the 

 large amounts of material available. 



Process for the estimation of butter fat, coconut fat, palm kernel fat and 

 their mixtures, S. H. Blichfei.dt {Jour. Soc. Chem. Indus., 38 {1919), No. 11, 

 pp. 150T-152T, fig. 1). — Certain alterations in the previously noted process for 

 estimating butter and coconut fats in margarin (E. S. R., 24, p. 213) are de- 

 scribed, and analytical results obtained by the method at different laboratories 

 are reported. The development of the process and some of the analytical data 

 supplied are attributed to P. S. Arup. 



The iodin value (Wijs) of palm kernel oil, R. H. Ellis and E. IVI. Hall 

 {Jour. Soc. Chem. Indus., 38 {1919), No. 10, p. i^ST).— Tables are given of the 

 maximum, minimum, and mean values for the iodin nimiber (Wijs) of a large 

 number of samples of both crude and refined palm kernel oil, obtained by ex- 

 pression from kernels crushed in the mills inider ordinary works conditions. 



The mean iodin value of the expressed oil was in every case outside the range 

 of the values usually given in standard texts. The authors conclude that the 

 normal range of iodin value (Wijs) is from IG to 23, and that any oil giving a 

 value outside this range would be of doubtful purity. 



The cuprous chlorid-iodin method for reducing sugars simplified, F. M. 

 Scales {Jour. Indus, and Engin. Chem., 11 {1919), No. 8, pp. 7^/7-750).— Certain 

 improvements in procedure in Clark's modification (E. S. R., 40, p. 114) of the 

 author's earlier method (E. S. R., 34, p. 611) for determining reducing sugars are 

 described which are said to simplify the method and render it more accurate. 



