742 EXPERIMENT STATION RECORD. 



Malt analysis, II. A. Hunicke {Jour. Amer. Chem. Soc, 26 (1904), No. 10, pp. 

 1211-1220). — Comparative tests were made of the effects of different amounts of 

 moisture and differenl degrees of fineness of grinding upon malt analysis, and pos- 

 sible sources of error in manipulation and analytical processes are spoken of. 



The detection of watered milk, A. E. Leach and II. C. Lythgoe (Jour. Amer. 

 Chem. Soc, 26 (1904), No. 10, /,j>. 1195-1208). — Determinations were made of the 

 refraction of the milk serum of commercial samples of milk and of milk of known 

 purity from Holstein cows. The Zeiss immersion refractometer was found the mosl 

 useful for this purpose, and a description of this instrument is given. In no instance 

 was the refractometer reading of the serum of pure milk less than 39 at 20° C, and 

 this is, therefore, considered a reasonable minimum figure, below which adultera- 

 tion with water would be indicated. Tables are given showing the indices of refrac- 

 tion on the Abbe refractometer and the equivalents in scale readings on the immer- 

 sion refractometer, and also for reducing scale readings at from 15-25° to 20° C. 



Determination of fat in milk, M.'Henseval (Rev. Gin. Lait, S (1904), No. 23, 

 pp. 5 29-535, /'As. 2). — The Gerber, Soxhlet, and Gottlieb methods are briefly described 

 and comparative determinations on whole milk, skim milk, and milk homogenized 

 by the process of Gaulin (E. S. R., 15, p. 714) are reported. In general, higher 

 results were obtained by the Gottlieb method, though, on whole milk, the 3 meth- 

 ods agreed closely. Considerable difficulty was experienced at times in applying 

 the Gerber method to homogenized milk, heating and centrifuging two or three 

 times being necessary. A period of 18 to 20 hours was required for extracting the 

 fat from such milk with ether, and then the results were generally lower than by the 

 Gottlieb method. Six samples of homogenized milk showed on an average 3.36 per 

 cent of fat by the Gerber method, 3.51 by the Soxhlet method, and 3.65 by the 

 Gottlieb method. 



The results of an examination of an emulsion of cotton-seed oil in skim milk are 

 reported, and attention is called to the possibility of adulterating milk with oils or 

 fats emulsified by the Gaulin process. 



Determination of fat in homogenized milk, A. Burr (Milchw. Zentbl., 1(1905), 

 No. 1, pp. 6-9). — Comparative determinations of fat in homogenized milk were made 

 by the Gottlieb-Rose, Adams, and Gerber methods, from the results of which it is 

 concluded that the Gottlieb-Rose method is the most accurate for this purpose. 



Investigations with sin-acid butyrometry, Hoitmkister (Milchw. Zentbl., 

 1(1905), No. 1, pp. 20-24). — Comparative determinations of the fat in milk, skim 

 milk, and buttermilk, with and without the addition of preservatives, were made 

 by the Gerber, Gottlieb-Rose, and the Sichler sin-acid-butyrometric (E. S. R., 16, 

 p. 506) methods. From the results the author concludes that the last method, while 

 not so suitable for numerous determinations as the acido-butyrometric method, may 

 be perfected to compare favorably with other methods. 



The relation of cows' milk to Schiff's reagent, and a test for formalin in 

 milk, E. Seligmann (Ztschr. Hyg. u. Infectionskrank., 49(1905), No. 2, pp. 325-328). — 

 Formaldehyde in milk may be detected by distilling and adding to the distillate a 

 solution of fuchsin bleached by sodium sulphite, when a violet-red color appears. 

 The same reaction is given by raw milk which the author finds can be prevented by 

 the addition of an acid to the milk, from which it is concluded that the reaction 

 depends upon some properties of the casein. 



To determine the presence of formaldehyde in milk without the necessity of dis- 

 tillation the author therefore adds 2 or 3 drops of sulphuric acid to 5 cc. of the 

 sample, and tests with Schiff's reagent. In dilutions of 1:40,000 the coloration is 

 most distinct at the end of about an hour. Control testsare recommended in all cases. 



The preservation of urine samples for analysis, M. S. McDowell (Pennsylva- 

 nia Sta. Rpt. 1903, pp. oS-41). — As formaldehyde did not prove entirely satisfactory 

 for preserving urine, the value of chloroform for this purpose was studied. Three 



