102 ANALYSIS OF MILK. 



froof spirit till the filtrate measures 120 c.c. 60 c.c. are evapor- 

 ated in a dish on the water-bath and the residue weighed ; the 

 residue is then ignited and the ash weighed, the weight of the 

 residue less that of the ash being the total sugar. In the other 

 60 CO., the milk-sugar is estimated by Fehling's or Pavy's solu- 

 tion ; the difference between this and the total sugar is cane 

 sugar. A deduction should be made from this of 



0-5 per cent, when the difEerenec is 0-5 or under, 

 0-2 „ „ „ 0-5 to 1-0, 



0-1 „ „ „ 1-0 to 1-5, and 



none per cent, if the difference exceeds 2-0. 



This method is stated to give fair results with percentages of 

 cane sugar exceeding a half per cent. 



Shenstone's Method. — The cane sugar may also be estimated 

 by determining the total polarisation of the sample as directed 

 for milk-sugar, and by estimating the milk-sugar gravimetrically 

 or volumetrically by Pavy's or Fehling's solution. The difEerence 

 between the percentage of anhydrous milk-sugar found by reduc- 

 tion of copper and that deduced from the polarisation divided 

 by 1"217 will give the percentage of cane sugar. This method 

 yields excellent results with mixtures of fresh milk and cane 

 sugar, and with many samples of sweetened condensed milks, 

 but is apt to lead to figures below the truth, if the milk has been 

 much heated, owing to the reduction in the rotatory power of 

 milk-sugar under these conditions. 



Griinhut and Riiber have shown that the estimation of milk- 

 sugar and cane sugar in condensed milk by reduction with 

 Fehling's solution does not give accurate results ; the figures 

 are always high for milk-sugar, and the cane sugar correspondingly 

 low. 



By polarisation before and after inversion, using Herzfeld's 

 formula for cane sugar, they obtain satisfactory results. 



Stokes - Bodmer Method. — Stokes and Bodmer prefer to 

 estimate the milk-sugar by titration with Pavy's solution (Feh- 

 ling's solution can be substituted for this), to invert the cane 

 sugar by boiling with 2 per cent, of citric acid for ten minutes, 

 and then to estimate the combined milk-sugar and resulting 

 mixture of glucose and fructose by titration ; the difierence 

 between the two figures will be due to the products of hydrolysis 

 of cane sugar. In this case it is advisable to standardise the 

 solution on a mixture of milk-sugar and inverted cane sugar in 

 about the same proportions as found in the milk. The deter- 

 minations may also be made gravimetrically. 



Watts and Tempany have proved that in the presence of milk 

 constituents cane sugar is not completely inverted by boiling 



