812 EXPERIMENT STATION EECORD. 



18-2Jt). — The method suggested tentatively in this preliminary communication 

 depends upon the formation, separation, and saponification of the esters. 



To 5 gm. of the melted fat is added 30 cc. of a tenth-normal alcoholic solu- 

 tion of sodium hydroxid. The mixture is thoroughly shaken, allowed to stand 

 for 8 minutes, and neutralized with dilute sulphuric acid. It is then made up 

 to 145 ce. with water and distilled, the alcoholic distillate of 30 cc. being kept 

 separate from the aqueous distillate of 100 cc. The two fractions are washed 

 into Erlenmeyer flasks, alcohol is added to the aqueous fraction, the free acids 

 in both are neutralized, and saponification is effected by the addition of 40 cc. 

 of tenth-normal sodium hydroxid and heating with a return-flow condenser for 

 45 minutes. The mixture is then titrated against tenth-normal hydrochloric 

 acid, the difl'erence between the number of cubic centimeters required for neu- 

 tralization and 40 showing the number of cubic centimeters of the tenth-normal 

 sodium hydroxid required for saponifying the esters in 5 gm. of fat. 



The highest number of cubic centimeters required for saponifying the esters 

 in the aqueous fraction was 10 in the case of pure butter, while it was very 

 much in excess of that for cocoanut oil. 



Detection of cocoanut oil in butter, F. von Morganstern and W. Wol- 

 r.RiNG (Ztschr. Vntersuch. Nahr. u. Genussmtl.. IS (1901), A^o. 7/, pp. 18J,, 185). — 

 The method of Wijsman and Reijst (E. S. R., 17, p. 834) was applied by the 

 authors to 20 samples of pure butter, the results showing no constant relation 

 between the first and second silver numbers. In the majority of cases, how- 

 ever, the second silver number was higher than the first. This, according to the 

 conclusions of Wijsman and Reijst, would indicate adulteration with cocoanut 

 oil, which was not the case. 



The silver index method of Wijsman and Reijst, II. Svobooa (Ztschr. Vnter- 

 such. Nahr. n. GciiiinsintL, l.i [1907). .Vo. 1. pp. 1.5-18). — This method proposed 

 for the detection of cocoanut oil in butter (E. S. R., 17, p. 834) was applied by 

 the author to 80 samples of pure butter. The Reichert-Meissl number with a dis- 

 tillate of 110 cc. averaged 27.29 and with a distillate of 300 cc, 31.03. The first 

 silver index averaged 3.64 and the second 3.91. In 57 of the 80 determinations 

 the second silver index was higher than the first, which was contrary to the 

 results obtained by Wijsman and Reijst. This method is, therefore, considered 

 of no value as a means of detecting the adulteration of butter with cocoanut 

 oil. 



On the determination of fat in cheese by the Gerber method, A. Scala 

 {8tdz. Sper. Agr. Ital., 39 (1906), No. 8, pp. 73Jf-737). — Comparative determina- 

 tions were made with the gravimetric method, the Gerber method, and the Sieg- 

 feld modification of the Gerber method. The results by the Gerber and Siegfeld 

 methods were above those by the gravimetric method in 1(3.6 and 45.21 per cent, 

 respectively, of the 42 determinations made and below in 83.33 and 54.76 per 

 cent. 



The chemistry of flesli. V, Methods for the determination of creatinin 

 and creatin in meats and their products, II. S. Grindley and II. S. Woods 

 ■(.four, liiol. Chc)ii.,2 (1907). No. .'/, pp. 309-315). — Using the colorimetric method 

 described by Folln (E. S. R., 17, p. 165), -determinations of the creatin and 

 creatinin in a number of samples of meat and commercial meat extracts are 

 reported. According to the authors, the colorimetric method employed has been 

 successfully used with uncooked and cooked meats, meat products, and drip- 

 pings. Recorded data show marked differences in the total amount of creatin 

 and creatinin contained in different commercial extracts now on the market. 



" Experiments upon meat extracts prepared upon a small scale in the labora- 

 tory, from fresh meat, are now under way to find out, if possible, the conditions 



