218 



THE DETECTION OF ADDED SUBSTANCES. 



cream are not of great accuracy. The fall in specific rotatory 

 power of milk-sugar is by no means constant, as milks sterilised 

 side by side may show very appreciable variations in this 

 respect. 



Tests other than the estimation of albumin must be considered 

 merely as corroborative and of qualitative value only. 



It must be remembered, however, that no sharp distinction 

 can be drawn between milk which has been raised to a tempera- 

 ture over 70 C. for a short period, and which is naturally not 

 sterilised in the true sense of the term, and the milk which has 

 been heated for a sufficient length of time to destroy all microbial 

 life. For this reason, a milk should not be reported as sterilised, 

 solely on the result of a very low percentage of albumin, if neither 

 the " creamometer " nor the para-phenylene-diamine nor " milk- 

 sugar " tests give corroborative indications. It is probable that 

 the milk, in this case, has been pasteurised slightly above 70 C. 



The following analyses (Table XL.) will show to what extent 

 the methods above described can be depended on : 



TABLE XL. COMPARATIVE ANALYSES OF MIXED FRESH 

 AND STERILISED MILKS. 



Diluted Condensed Milk. This will behave as sterilised 

 milk ; the cream rises with even less readiness from diluted con- 

 densed mik than from sterilised milk. 



