MILK TESTING. 71 



which have been melted at the necessary temperature, thoroughly incor- 

 porated with each other, are enclosed in test-tubes, and subjected to centri- 

 fugal force in the lactocrit. The percentage of fat is estimated by the 

 observed volume of melted fat. Originally this method was only utilized 

 in the investigation of whole milk. Subsequently, in 1890, the mixture of 

 acids was replaced by a quantity of ethylidene-lactic acid and solution of 

 hydrochloric acid, which perfectly dissolved the nitrogenous matter in the 

 milk, without attacking the fat to any extent. The result was a great 

 improvement, both in accuracy and convenience, in determining the fat, 

 and a more extensive application of the method ensued. It enables a 

 determination of fat to be made in skim-milk and butter-milk, as well as 

 whole milk, if not directly, yet with great accuracy. 



In the method devised by Marchand de Fecamp, which was investi- 

 gated and improved in 1878 by Schmidt and Tollens, the milk is treated 

 in a lacto-butyrometer with alcohol, ether, and a little potash. The fat is 

 dissolved and almost entirely separated in the surface layer of the ether. 

 From the volume of this layer the percentage of fat is calculated by means 

 of a table. 



A fact which militates against the Marchand method is the retention, 

 in the lacto-butyrometer, in a dissolved condition, of a certain proportion 

 of fat. This amount, although generally the same, may vary under 

 certain conditions. In this method, therefore, conditions have to be 

 reckoned with which are not perfectly under control. All the improve- 

 ments made up to the present time in this process affect only the details, 

 such as greater convenience in working it, more exact methods of reading 

 the degree, &c., and do not affect the accuracy of the process. With 

 milk containing from 3 to 3J per cent of fat this method gives good 

 results, the variations from gravimetric methods being generally less than 

 2 per cent. It is well suited for practical use in agriculture generally, 

 and is useful, for many purposes, in large dairies. For scientific work, 

 however, or for legal purposes, and for the determination of the commercial 

 value of milk, it is not sufficiently accurate to be relied upon. It cannot 

 be used for solutions containing more than 1'339 per cent of fat. The 

 methods for fat determination already described, and more especially the 

 Soxhlet and lactocrit, are thoroughly accurate, delicate, and reliable 

 scientific processes. 



Where it is impossible to estimate the percentage of fat, and the 

 above methods of milk -testing are consequently inapplicable, a 

 milk-test devised by Muller may be found to serve the purpose. 

 The specific gravity at 15 C. is determined, and the milk is allowed 

 to stand for 24 hours in a Chevalier cremometer at a temperature 



