138 VOLUMETRIC AND INDIRECT ESTIMATION OF FAT. 



should be cut up into small pieces, which can be dropped in. 

 Add sufficient water to make up the weight to 15-25 grammes 

 and proceed as in analysing milk. Cheese requires rather longer 

 shaking than other products, but gives equally good results. 



If desired, cream may be weighed instead of being measured. 



The calculation is performed as for sour milk. 



The above directions differ in some respects from those given 

 by Leffmann and Beam. The author has had. however, some 

 years practical experience of the methods described and is con- 

 vinced of their accuracy. A stand for the bottles is to be recom- 

 mended ; this may conveniently be made of wire rings into which 

 the bottles fit, with a flat plate for a bottom ; the bottles can 

 then be easily carried about. 



To clean the bottles : empty whil^ hot in a convenient recep- 

 tacle, and wash twice thoroughly with hot water ; if necessary, 

 run a brush down the neck. They are washed conveniently in 

 the stand. Never leave pipettes dirty. 



Failures and their Probable Causes. The only failures 

 likely to happen are : 



1 . Dark layer of fat. 



2. Fluffy layer under the fat. 



1. If the acid be too strong, or the temperature too high, or 

 the mixture left too long before whirling, the fat may be dark. 

 The remedy is obvious. 



2. A fluffy layer under the fat is often caused by allowing the 

 milk and acid to stand too long unmixed. It may sometimes be 

 due to a bad quality of amyl alcohol. 



Grit on the bottom of the bottles may cause fracture while 

 in the machine. Fracture may also occur from too sudden a 

 stoppage after the whirling is completed. 



Modifications of the Lefifmann-Beam Method. The 

 Leffmann-Beam method has been subjected to considerable 

 modification ; thus Paterson and, later, Gerber have used amyl 

 alcohol alone without hydrochloric acid. 



G-erber's Acido-butyrometric Method. This is essentially 

 the Leffmann-Beam method, the chemical principles of which 

 have been adopted. The use of hydrochloric acid as a solvent for 

 the amyl alcohol has been, however, discarded, following Paterson. 



Gerber employs a test bottle, which he terms an " acido- 

 butyrometer," which differs from that employed by Leffmann 

 and Beam ; it is a modified form of Marchand's lacto-butyrometer, 

 and, like this, is closed with a cork. 



A definite strength of sulphuric acid is prescribed (90 to 91 per 

 cent.), and rules for testing the acid and amyl alcohol used are 

 laid down. 



