AGEICULTUKAL CHEMISTRY — AGROTECHNY. 505 



chlorid. Pure flour settles to the bottom of the tube as a flesh-colored mass with 

 a supernataut yellow fluid, while flour containing alum gives a slatish to blue- 

 violet color. 



For baked goods a microscopic examination of the ash Is necessary. The ash 

 is dissolved in hydrochloric acid, precipitated with ammonium hydrate, and 

 redissolved in hydrochloric acid. To the resulting solution some potassium 

 bisulphate is added. If the resulting precipitate contains octohedral crystals, 

 which are characteristic of the salts of aluminum, alum is present. The 

 crystals must be detected with the microscope. 



A report on the determination of the amount of tin in tinplate used 

 for canning preserved food, W. W. O. Beveridge (Jour. Roy. Army Med. Corps, 

 16 (1911), No. 2, pp. lJiO-160, figs. 2).— This work presents the results of some 

 examinations of tinplate destined for the manufacture of cans for holding pre- 

 served fruits, etc. As a preliminary test for the amount of erosion in a given 

 can the author utilized the Walker gelatin-potassium ferricyanid test and the 

 following method for determining the amount of tin present in the tinplate : 

 Pieces of plate 1 sq. in. in size are subjected to the action of a warm solution 

 of dry chloriu gas in carbon tetrachlorid or in fuming stannic chlorid to remove 

 the tin. then washed and reweighed, the tin present in the sample thus being 

 determined by difference. 



For determining the variations in the coating of tin the author describes a 

 method, from which the following conclusions are drawn: "(1) For cans to 

 hold preserved meats, such as corned beef, the amount of tin per base-box of 

 112 plates measuring 14 by 20 in. should be at least 2.5 lbs., or an amount not 

 falling anj^where below 0.0361 gm. to the square inch (both sides), to insure a 

 fair distribution of the tin. (2) For cans intended for such foods as fruits, 

 jams, etc., owing to their acid nature, and for meat extracts, essences, and 

 highly salted foods, a layer of tin equivalent to not less than 3.5 lbs. per base- 

 box, or 0.050G gm. per square inch (both sides). (3) For dried foods such as 

 biscuits, meal, etc., probably 2.25 lbs. or 0.0325 gm. per square inch would be 

 quite suflicient. (4) In every case one-twentieth of the surface of the top, 

 bottom, and sides respectively should be taken for analysis from the same can, 

 and the lowest amount found on a given surface quoted. (5) In all cases the 

 layer of tin should be continuous and as evenly laid as possible. (6) There 

 should be no fissures and any surface extensively riddled with pin-holes as 

 shown by the Walker test should lead to rejection of the plate. It is fair to 

 state, however, tliat at tlie present time a certain number of pin-holes appear 

 to be inevitable, but with improved methods of plating they are likely in the 

 future to be almost entirely eliminated. (7) The custom of painting the outside 

 of the cans to prevent rust is a good one, and should be continued." 



Detection of saponin in beverages and foods by the hemolytic method, 

 C. SoRMANi (Ztschr. Untcrsuch. Nahr. u. Genussmtl., 23 (1912), No. 11, pp. 

 561-566). — The author concludes that the hemolytic methods, especially Rus- 

 coni's procedure, yields far better results than the usual chemical methods for 

 detecting saponin in foods, etc. 



Detection of saponin, J. RtJHLE (Ztschr. Vntersuch. Nahr. u. Genussmtl., 23 

 (1912), No. 11, pp. 566-577). — 'Vamvaka's method for the detection of saponin 

 and for differentiating saponin from glycyrrhizin is not considered accurate. 

 The color reactions with concentrated sulphuric acid and Frohde's reaction 

 (100 cc. of concentrated sulphuric acid -fl gm. of ammonium molybdate) are 

 not always reliable, while the reactions obtained with a-naphthol, thymol, and 

 purified oxgall, which are dependent upon the presence of sugar, are of no 

 value at all. In these tests the hemolytic test was the only one which gave 

 entirely satisfactory results. 



