7 2 
TlMEHRI. 
of 21 samples of cane juice. The samples marked with an 
asterisk are those to which a greater percentage of 
sugar has been assigned than sugar solutions of the same 
density could contain. Line viii. shows the approximate 
excess of total solids estimated from the density, and line 
ix. the result of multiplying the degree Bm. by 1*65, 
the figures thus obtained serving for comparison with the 
percentage of sugar in line iii as found by Prof. HAR- 
RISON. 
That the amount of total solids in cane juice is fairly 
well represented by the amount of solid matter (as sugar) 
indicated by the density is evident from the following 
experiments. They were carefully made by quickly 
weighing 5 cubic centimetres of juice in a broad platinum 
basin, and evaporating over a steam bath until the solid 
residue lost less than a milligramme in weight an hour. 
In some of the trials, alcohol-was added to the contents 
of the basin during the evaporation, taking care to 
decrease the heat to prevent loss from ebullition. No 
particular advantage seeming to be derived from the 
alcohol, its further use was stopped. The following are 
the results obtained : — 
Density of juice ... 
Degrees Bm. 
Per cent, solids (as sugar) 
indicated 
I 
1 
1 041 
5-7 
10*20 
2 
1 104 
13*6 
24-60 
3 
1083 
11. 1 
20*10 
4 
1073 
9'8 
1770 
5 
1058 
7*9 
14*20 
6 
1072 
97 
17*50 
7 
1073 
9-8 
1770 
Per cent, solids as 
mined ... 
deter- 
i 
IC31 
24*92 
20*l6 
17*87 
14*14 
1770 
1772 
It will be noticed that the presumed excess in Prof. 
HARRISON'S results shewn by line viii. is fairly uniform, 
and probably originates from some systematic error that 
can easily be explained and reftified. I am sure that 
