716 EXPEEIMENT STATION RECORD. 



The detection of coconut fat in butter and lard, E. Polenske (Arb. E. 

 Gsndhtsamt., 38 {1911), No. 3, pp. 402-JilJ,) .—It appears that the method of 

 Ewers (E. S. R., 23, p. 417) and Fendler (E. S. R., 23, p. 418) can not be used 

 for this purpose. The Polenske method, on the other hand, will give trust- 

 worthy results. The results of examining 34 fresh and 8 old butters are given 

 in detail. 



Determination of fat in feeds by extracting in the cold, C. Gbimme (Chem. 

 Rev. Fett v. Harz Indus., 19 {1912), No. 1, pp. 1-5). — As considerable danger 

 from fire exists when using ether for extracting fats, tests were conducted 

 with chloroform, carbon tetrachlorid, and di-, tri-, and per-chlorethylene, for 

 determining their value as fat solvents for air-dry cotton-seed meal. 



Carbon tetrachlorid and tri-chlorethylene were the only reagents which 

 gave results comparing well with those obtained with ether. The method 

 finally adopted was as follows: Ten gm. of the air-dry sample is shaken 

 with 100 cc. of either carbon tetrachlorid or tri-chlorethylene for | hour in a 

 150 cc. flask in the cold and with the aid of the shaking machine. After 

 allowing the insoluble particles to settle for 30 minutes, 50 cc. of the filtered 

 fat solution is freed from the solvent by distillation and then dried for about 1 

 hour (by occasionally blowing air over it) at a temperature of 105° C. Multi- 

 plying by 20 the figures obtained gives the percentage of fat in the feed, provid- 

 ing the material did not contain more than 10 per cent of fat. If more fat is 

 present corrections should be applied, viz, for 15 per cent fat -f 0.2 per cent, and 

 for 20 per cent fat -fO.4 per cent. The method gave good results with cotton- 

 seed meal, peanut cake, peanut meal, rape cake, sunflower cake, coconut cake, 

 soy-bean meal, linseed meal, bran, rice-feed meal, fish meal, corn meal, wheat 

 germs, and rye bran. 



Grading rosin at the still, F. P. Veitch and C. F. Sammet ( U. 8. Dept. Agr., 

 Bur. Chem. Circ. 100, pp. 4, fig. 1.) — A method is proposed whereby a pro- 

 ducer can grade rosin at the still when it is made and w^hich does away with 

 the cutting of rosin samples. The method consists of pouring the liquefied 

 rosin with an ordinary tin dipper into a mold made of ordinary roofing tin. 

 When cool the samples can be compared with a set of standard types of rosin. 



" By far too much low-grade rosin is being made. There are three chief 

 causes for this : First, the presence in the gum in the still of considerable 

 quantities of wood, bark, pine needles, charcoal, and dirt which, through care- 

 lessness, accident, or both, have gotten into the gum in the woods and the 

 larger part of which remains in the still during distillation. . . . Second, 

 excessive heating or prolonged heating of the rosin in the still at the close of 

 distillation, largely for the purpose of removing all tun^entine and water 

 which it may contain. Third, the presence in the finished rosin of much 

 finely divided foreign material, such as charcoal and earth mentioned above, 

 due to faulty straining of the finished rosin. . . . Rosins which had been 

 graded A, B, C, D, and E, when dissolved in turpentine or alcohol, carefully 

 strained, and the solvent evaix)rated with heat, all grade higher, many of 

 them F or better, thus showing that these rosins were graded low because 

 of the foreign material they contained, which had not been removed in strain- 

 ing, rather than because of the color of the rosin itself," 



Oil from plants belonging to the order Papilionacese, C. Gbimme {Pharm. 

 Zentralhalle, 52 {1911), No. 1,3, pp. 1141-11^9; ahs. in Chem. Zentbl, 1911, 

 II, No. 23, pp. 1139, 17J,0; Jour. Soc. Chem. Indus., 30 {1911), No. 2Jf, p. 

 1460). — ^The following seeds were examined for their oil content: Cicer arie- 

 Unmm, Pisum sativum, Vicia faha, V. sativa, V. sepiitm, Lens esculenta, Cajanus 



