BATES AND JACKSON: QUARTZ-WEDGE SACCHARIMETER 27 



A study of the reaction velocities of the copper solution with 

 sucrose (a slow reaction) and with invert sugar (a rapid reaction) 

 showed that a great part if not all of the 0.006 per cent reducing 

 substance was due to pure sucrose. The recrystallized substance 

 was consequently concluded to be free from reducing sugar. 



In order to be certain of the elimination of impurities which 

 could not be detected by direct test, a quantity of sugar was 

 fractionally crystallized. In this process those fractions which 

 supposedly contained impurities of about the same solubility were 

 united, in order to prevent too great a subdivision of the original 

 substance. In this way five fractions were obtained after 32 

 crystallizations which upon test yielded essentially identical 

 rotations. This was concluded to be evidence of the purity of the 

 substance. 



In the elimination of moisture it was necessary to heat the 

 substance for such a time and at such temperature as to avoid 

 decomposition into caramel. The rates of decomposition were 

 measured in several series of experiments and the results plotted. 

 The curve showed the time at each temperature required to form 

 caramel equivalent in reducing power to 0.01 per cent invert 

 sugar. The data are: 100°C., 0.015 hrs.; 79?5, 0.57 hrs.; 66?6, 

 10.9 hrs. ; 50?0, 107.0 hrs. ; 39?0, 476 hrs. In preparing the sugar 

 for polarization care was taken to avoid sensible decomposition. 



In studying the elimination of moisture the substance was 

 subjected to high vacuum, to the highest permissible temperature, 

 and to various drying agents for long periods of time. The con- 

 clusion was reached that a few hours heating at 50° to 60? C in a 

 vacuum of 0.001 mm. of mercury and in the presence of quicklime 

 would eliminate all but negligible quantities of moisture. 



The solution used for polarization was never of exactly normal 

 concentration. For preparing it the approximate normal weight 

 was transferred to a weighed volumetric flask. Flask and sugar 

 were then subjected to the drying operations before the final 

 weighing. 



The volume of the total solution was found either by filling to the 

 graduation mark of calibrated flasks or was calculated from the 

 weight and density. Density was taken from published tables 



