ICE-CREAM AND ICES 639 



etc. go to the bottom. It follows that if ice-cream has softened 

 it should not be refrozen unless it is first thoroughly agitated in 

 the freezer. 



Refreezing, as a rule, should be discouraged, because the ice- 

 cream has been exposed to influences which may make it unsafe. 

 Softening usually occurs in retail stores where the can is fre- 

 quently opened and chance given for the entrance of infectious 

 material. Furthermore, poisonous products due to microbial de- 

 composition may form in softened ice-cream. 



The test for fat in ice-cream is usually made by a modified 

 Babcock method. The original Babcock method is not suitable 

 because the sulphuric acid attacks the sugar and possibly some 

 other substances, with the result that a clear fat column is not 

 obtained. Washburn gives the following method: 



" Carefully weigh 18 grams of a well melted (but not over- 

 heated) and mixed sample of ice-cream into a 30 per cent, cream 

 bottle. To this add 4 or 5 c.c. of lukewarm water. Now add 

 ordinary sulphuric acid, a little at a time, thoroughly mixing the 

 fluids with each addition. Little more than half and seldom as 

 much as two-thirds the usual amount of acid is required; and not 

 more than one-half of this amount should be used at the outset, 

 and some little time should be allowed for it to act. If the color 

 is not yet that of strong coffee, add a little more acid, shake, and 

 pause for a time. If still the color is too light, add yet more acid. 

 In this way the color is built up to the .desired point. When the 

 contents of the bottle have assumed almost the desired amber 

 color, add 4 or 5 c.c. of cool water to check the further action of 

 the acid. The test is thereafter conducted as would be an ordi- 

 nary cream test, care being taken that the machine does not 

 become too hot during whirling. If this scheme is carefully fol- 

 lowed, particularly in the matter of the slow and gradual addition 

 of the acid, the fat should appear in the neck of the test bottle 

 of a clear, light brown color and distinct from the solution below. 

 When this distinct, clean-cut condition has been obtained, the 

 tester may feel sure, provided the work has been in other respects 

 carried out in accord with the well understood details of the 

 Babcock method, that the results will be reasonably accurate." 



The Illinois Food Commission advises the following modifi- 

 cation of the Babcock method : 



" Weigh 9 grams of the melted and thoroughly mixed sample 

 into a 10 per cent. Babcock milk bottle. Add 10 c.c. of glacial 

 acetic acid, washing down the neck of the bottle with the acid. 

 Heat this mixture by putting the body of the bottle in hot water 

 until the contents appear smooth, uniform, and free from lumps. 

 Then, while still hot, add ordinary Babcock sulphuric acid about 



