160 



OSMOTIC PRESSURE OF AQUEOUS SOLUTIONS. 



the solutions, was warmed, but not to a temperature above 60°. The 

 solution (sometimes thinned with a little alcohol) was filtered, and from 

 the filtrate the sugar was precipitated by alcohol which had been distilled 

 from lime — a few crystals of the purest sugar being used to start the precip- 

 itation. The precipitated sugar was collected on a perforated porcelain 

 disk in the bottom of a glass funnel, and freed as perfectly as possible from 

 mother liquor by means of the filter pump. The material was then trans- 

 ferred from the funnel to a porcelain dish and mixed to a thin paste with 

 85 per cent alcohol. Finally it was again filtered, and then nearly dried 

 by drawing through it filtered air. The original rock candy and the 

 product of the first crystallization will be designated hereafter by the 

 letters A and B. The yield of B was 32 kilograms. 



The various portions of B were thoroughly mixed and then resubjected 

 to the treatment which has already been described, except that the prod- 

 uct of the second precipitation was washed first with diluted ethyl alco- 

 hol and afterwards with warm methyl alcohol. The yield of the twice 

 recrystallized sugar, which will be designated by the letter C, was about 16 

 kilograms. 



A portion of C was again dissolved, reprecipitated, and washed with 

 both ethyl and methyl alcohols. The product of the third precipitation 

 will be designated by the letter D. 



Combustions were made of all four products, namely A, the original 

 rock candy; B, which had been precipitated once; C, twice; and D, three 

 times. The results are given below in percentages of hydrogen and 

 carbon. 



Table 46. 



The differences between the percentages of hydrogen and carbon which 

 were found and the theoretical values are all within the unavoidable errors 

 of analysis, and there was, therefore, no reason to be discovered in the 

 figures given above for regarding any one sample of the sugar purer than 

 another. A determination of carbon and hydrogen does not, however, 

 suffice for the detection of glucose or invert sugar in cane sugar; and 

 evidence of the probable presence of reducing sugars could be discovered 



