THE ANALYSIS OF MILK. 83 



apparatus. Care must be taken that the coil should not be filled with 

 cotton wool to its highest surface, and that the ether which comes from 

 the condenser attached to the apparatus when the extraction is going 

 on should always drop in the middle of the coil. After the coil is placed 

 in the extraction apparatus, a wide-necked weighed flask containing 25 c.c. 

 of pure ether is attached to the lower end of the extraction apparatus. 

 The porcelain dish which has been used, along with the glass stirrer and 

 pestle which have been used, are repeatedly rinsed out with ether, which 

 is then poured on to the coil in the extraction apparatus. Sufficient 

 ether is then added to the extraction apparatus till the syphon is almost 

 full, a condenser is then fixed on above, the wide-necked flask placed in 

 a sand-bath, the temperature of which is maintained at about 60 C., 

 and the extraction is started. As a rule, it is ended in about three hours. 

 Whether this is long enough, or whether the extraction requires to be 

 continued for a longer period, can be proved by the watch-glass test. 

 After the extraction has been finished the flask is taken off, and after the 

 ether has been slowly distilled it is placed in the drying-bath, dried for 

 45 minutes at 100 C., and then for 15 minutes at a temperature of from 

 105 to 110 C., cooled in the desiccator, and weighed. The flask is again 

 introduced into the drying-bath, dried for 30 minutes at 100 C., allowed 

 to cool, and weighed again; and this is repeated until the two last 

 weighings are found to show no greater difference than 1 milligram. 

 Nearly always from 60 to 90 minutes is sufficient to effect thorough 

 drying. If the fat has to be determined in skim-milk, sea sand is not 

 used, but gypsum. A larger quantity of this is used than is necessary to 

 absorb the liquid, and the extraction lasts for at least four hours before 

 the watch-glass test is applied for the first time. The limits of experi- 

 mental error for milk may be stated at, for whole milk, plus or minus, 

 -05 per cent, for skim-milk, plus or minus, '03 per cent. The determi- 

 nation of the fat by the Soxhlet method gives equally exact results. The 

 extraction apparatus must be firmly connected with the fat flask, and 

 the condenser to the apparatus. The three pieces of apparatus should not 

 be attached to each other with cork. 



A much simpler method, and perhaps even a more accurate one in its 

 results, for the estimation of fat, is Adams' process, in which the milk is 

 dried on blotting-paper. 



A coil of filter-paper, 56 cm. long, and 6*5 cm. broad, which has been 

 previously treated with ether to remove any trace of fat it may contain, 

 is allowed to absorb from 8 to 10 grams of milk, weighed out from a 

 beaker by difference as above described. After a few minutes, and when 

 the milk has thoroughly soaked in, the coil is hung on to a peg in the 

 drying-bath and allowed to dry for an hour at 97 to 98 C. The coil is 



