AGRICULTURAL CHEMISTRY. 413 



was u shortage in the percentage of fat. This shortage was almost uniformly 

 greater in the ether extraction than in the Babcock test. The Reichert-Meissl, 

 and iodin values and the melting points of mixed fats and the insoluble fatty 

 acids slu)wed that there was no loss of volatile fat by evaporation. 



In evaporated milk the casein is precipitated into insoluble curd and each 

 particle of curd locks up a part of the butter fat. In order to liberate this fat 

 larger amounts of sulphuric acid were used with the Babcock test but did not 

 affect the test materially, neither did potassium hydrate used as a solvent 

 befoi'e adding the acid. Borax as a solvent increased the percentage somewhat. 



The samples of evaporated milk were then exposed to the acid for 1 to 24 

 hours before being centrifuged, with the result that the samples in which the 

 milk remained exposed to the acid for 24 hours tested on an average within 

 0.1 per cent of the calculated percentage but the fat remained cloudy. Holding 

 the bottles for only 1, 4, or 8 hours did not appreciably influence the result. 



A one-fourth dilution of the evaporated milk, 1 part milk to 3 parts water, 

 increased the test to the correct percentage of fat but did not improve the 

 clearness of the reading. 



The addition to the test bottle of hot dilute sulphuric acid. 1 part acid to 1 

 part water, in the place of hot water after the first whirling prevented the pre- 

 cipitation of charretl and curdy masses and yielded a clear fat column, making 

 accurate reading possible. A series of comparative tests with this modified 

 Babcock test gave uniformly correct results. 



Experiments were then made to improve the official ether extraction method. 

 Prolonged ether extraction increased the fat yield slightly, but when continued 

 for 24 hours thei'e was still a shortage of 0.5 of 1 per cent of fat. The greater 

 the degree of dilution of the evaporated milk the more nearly complete was the 

 ether extraction. When the greater part of the proteids was removed by leach- 

 ing the paper coil with dilute acetic acid for 2 hours, with subsequent ether 

 extraction for S hours, all of the fat was liberated and yielded uniformly cor- 

 rect results. 



" The results of the modified ether extraction with 12 different brands of 

 evaporated milk averaged 1.69 per cent fat higher than those reported by a 

 commercial chemist. The minimum difference was 0.67 per cent, the maximum 

 difference was 3.28 per cent fat. . . . 



" The season of year when manufactured, the degree of concentration, and the 

 particular process of manufacture affect the properties of the caseous matter 

 in evaporated milk. These factors explain the difference in the results obtained 

 by the ordinary method of fat determinations with samples of different brands. 

 The Gottlieb method offers many advantages and the results are uniform, but 

 the percentage of fat is low." 



Full directions are given for using these modified Babcock and modifietl ether 

 extracting methods for testing unsweetened condensed milk. 



The detection of added water in altered milk, A. Kling and P. Roy ( Comiit. 

 Rmd. Acad. Sci. [Paris], l.',8 {1909), No. 16, pp. 1050-1052; Ann. Chim. Analyt., 

 l.'f (1909), No. (I, pp. 215-217; ahs. in Jour. Soc. Cheni. Indus., 28 (1909), No. 10, 

 p. 539). — The total nitrogen content of two samples of milk kept for 2 years 

 in storage in a closetl vessel was found to remain stationary, although changes 

 occurred in the nitrogen compounds. According to the authors this serves as 

 a basis for the detection of watered milk, by a method which is described, when 

 it is necessary to keep a suspected sample for some time. 



On the ash content of serum as a test for detecting watered milk, H. 

 Sprinkmeyeb and A. Dieuuiciis (Ztschr. Untvrsuch. Nahr. u. (JcnussmU., 17 



