MILK. 309 



wash-liquor are precipitated by tannic acid (see page 21). 63# of 

 the tannic acid precipitate is considered as albumin, and this must 

 be added to the albumin found directly. This method gives exact 

 and good results. 



According to the method of SEBELIEN, 3-5 grms. of milk are 

 diluted with an equal volume of water, a little common-salt solution 

 added, and precipitated with an excess of tannic acid. The precipi- 

 tate is washed with cold water, and then the quantity of nitrogen 

 determined by KJELDAHI/S method. The total nitrogen found 

 when multiplied by 6.37 (casein and lactalbumin contain both 

 15.7$ nitrogen) gives the total quantity of albuminous bodies. This 

 easily -executed method gives very good results, but until the quan- 

 tity of nitrogen in the albuminous bodies of other varieties of milk 

 has been exactly determined this method cannot be used except 

 for the analysis of cow's milk. 



To determine the casein and albumins separately we may make 

 use of the method first suggested by HOPPE-SEYLER and TOLMAT- 

 SCHEFF, in which the casein is precipitated by magnesium sulphate. 

 According to SEBELIEN, the milk is diluted with its own volume 

 of a saturated magnesium-sulphate solution, then saturated with 

 the salt in substance, the precipitate filtered and washed with a 

 saturated magnesium-sulphate solution. The nitrogen is deter- 

 mined in the precipitate by KJELDAHI/S method, and the quantity 

 of casein determined by multiplying the result by 6.37. The quan- 

 tity of lactalbumin may be calculated as the difference between the 

 casein and the total albumin found. The lactalbumin may also be 

 precipitated by tannic acid from the filtrate containing MgSO* from 

 the casein precipitate, diluted with water, and the nitrogen deter- 

 mined by KJELDAHI/S method and the result multiplied by 6.37. 

 SEBELIEN'S method is only suited for cow's milk. 



The quantity of globulins in milk cannot be exactly determined. 

 A minimum result can be obtained by first precipitating the casein 

 completely by NaCl in substance, and then precipitating the glob- 

 ulins in the filtrate by magnesium sulphate (SEBELIEN). The 

 casein may also be precipitated from the diluted milk, and the 

 globulin precipitated after neutralization by means of MgS0 4 . In 

 these cases we obtain somewhat high results, because of the pres- 

 ence of traces of casein which remain behind. 



The fat is determined directly by thoroughly extracting the 

 dried milk with- ether, evaporating the ether from the extract, and 

 weighing the residue. The fat may be determined by aerometric 

 means by adding alkali to the milk, shaking with ether, and deter- 

 mining the specific gravity of the fat solution by means of SOXH- 

 LET/S apparatus. In determining the amount of fat in a larsre 

 number of samples the lactokrit of DE LAVAL may be used with 

 success. The milk is first mixed with an equal volume of a mix- 



