17 



differences noted, therefore, in the case of tho mill juices :ire 

 than in the juices of diffusion. This mutter will be referred to sixain iu 

 the Louisiana analyses to follow. In Table No. 14 the differences are 

 given : 



TABLE No. 14. Comparison of total solids by spindlr icitlt results obtained by direct es- 

 timation. 



Dr. Crampton makes the following observations on this work : 



These results arc very interesting and important. They show that while the spin- 

 dles give results Tint slightly below the actual determination by drying in the case of 

 mill juices, the results with the diffusion juices were, on an average, .85 too high. 

 The spindles used were tested afterwards with a standard solution of pure sugar, and 

 found to give results about .2 too high. They corresponded closely with a delicate 

 sacchari meter tested by Scheibler. The different results given by them in the case 

 of tho mill and diffusion juices I am unable to explain, as it would seem more rational 

 that the diffusion juices, being more nearly pure solutions of sugar than the mill, 

 would give results approximating more closely to the standard upon which the spin- 

 dles were based. It is possible that the large amount of suspended solids in the mill 

 juices may in some way account for the discrepancy. At all events the direct de- 

 termination doubtless gives more reliable results. Correcting the average results on 

 tho basis of the samples in which a direct estimation was made we have : 



Total solids in tho mill juices for the season 15. 66 



Co-efficient of purity based on above 60. 9 



Total solids in diffusion juices for the season 10.23 



Co-efficient of purity (ir>. :' 



Showing an increase in the purity of the diffusion over the mill juices of 4.4 points. 

 The ratio of glucose to sucrose in the two juices for the season was as follows : 



Milljuice 1 : 2.80 



Diffusion juice 1 : ;>. 95 



This would seem to show one of two things: Either there was absolutely no inver- 

 sion in tho battery, and the slight difference in favor of tno diffusion juice was due 

 to error of analysis, or that the glucose in the eane was not so readily diffusible as t In- 

 sucrose, and thus a greater proportionate amount of the latter wMsobta-innl by diffusion 

 L>;r7C Bull 18 2 



