ANALYSIS OF SOUR MILK. 147 



established that, if milk is adulterated with added water, the 

 percentage added can be deduced from an analysis of the sour 

 milk, and that the figure thus obtained does not differ by more 

 than 3 per cent., and usually by much less from that estimated 

 by the analysis of the fresh milk. 



The author and ililler have submitted the Government Lab- 

 oratory method to a critical examination, and generally endorse 

 thi' above conclusions. 



Table XVI. gives the results of the. analysis of 18 samples of 

 sour milk, corrected according to the foregoing scheme. 



Carbonic acid can only be estimated in koumiss or kephir 

 contained in a corked bottle. The worm of a champagne tap is 

 carefully turned off to leave a perfectly smooth stem ; the tap 

 is also carefully reground to make sure that it fits. 



\ drying and absorbing apparatus is fitted up, consisting of 

 (a) a U-tube containing pumice and sulphuric acid, (b) a U-tube 

 containing soda lime immersed in a beaker of cold water, (c) a 

 U-tube filled half with soda lime and half with calcium chloride. 

 Th(>se are connected in the order named, and the end of («) is 

 connected by a short piece of india-rubber tubing to the cham- 

 pagne tap. 



(b) and (c) arc weighed, and the tap (closed) carefully forced 

 through the cork of the bottle ; the tap is slightly opened and 

 the carbon dioxide allowed to slowly escape ; when the escape of 

 i'.as becomes slack the bottle may be slightly warmed, by placing 

 it in warm water, and shaken to promote further escape. When 

 no more gas comes oft' the tap is disconnected, a soda lime 

 tube suljstituted, and a current of air drawn through the 

 apparatus. 



(b) and (c) are dried, cooled and again weighed ; the increase 

 represents the amount of carbon dioxide which has escaped from 

 the bottle. The total contents of the bottle are now weighed 

 and the percentage calculated. 



There still remains a little carbon dioxide dissolved ; this can 



N 



be best estimated by titrating a weighed amount with ^^ baryta 



water, using phenolphthalein as indicator ; the diSerence between 



the acidity thus estimated and that estimated as previously 



described will represent, without great error, the carbon dioxide 



N 

 (1 CO. -— - alkali = 0'0022 gramme CO^). This should be added 



to the amount estimated by absorption. 



Buttermilk and whey are analysed by the methods given 

 for milk ; the total proteins of whey cannot be determined by 

 Ritthausen's method, and the total nitrogen must be estimated. 



