Report of Society's Meetings. 225 
cent, with sample 6, to 79*3 per cent, with sample 56, thus showing a 
difference of 28*9 per cent, in the crushing, although the proportion of 
juice in the two samples of cane differs by less than 5 per cent. Such a 
difference as these two samples afford is no doubt very exceptional, 
and may even be due to an error in the determination, although I can 
find no evidence of that being the case. 
Apart from these examples, however, numerous others will be found 
in the table, where the proportion of fibre is nearly or quite the same, 
yet the crushing results vary largely as, for instance, with samples 
1 and 2, 11 and 12, 17 and 19, 28 and 29, 58 and 67. In each case the 
proportion of juice yielded by grinding varies between 10 and 15 per 
cent, and the lower result is given by the cane containing the greater 
amount of juice. 
I meant to have brought this matter to the notice of the Society at 
the last meeting, but as the results had been obtained only in the 
regular way of analysis, and were not intended to prove any such 
point as the one in question, I thought it better to postpone my 
communication until I had made some special experiments with as great 
care and under as similar conditions as possible. Owing to the long 
time required to dry the fibre, only a few of these have been completed, 
but as far as they go, they fully support the conclusions previously 
arrived at. 
In making these special experiments the adjustment of the rollers of 
the mill remained unaltered throughout. The same two men were 
employed at the cranks (I being one), and the megass was squeezed and 
re-squeezed until no further juice could be extracted — the final squeezing 
being done with the megass in layers thick enough to require the utmost 
available power to get it through the mill. The crushing was per- 
formed in duplicate by using only half a cane for each experiment. 
Table II. gives the results obtained. 
Being so few in number the results do not exhibit the striking 
differences met with in the previous Table, still they go a long way towards 
substantiating them. Thus, by comparing samples 1 and 3, and 4 and 7, 
it will be seen that both pairs differ to about the same extent in the 
amount of juice they contain, but, in the crushing, the former differs by 
only 2 per cent, and the latter by no less than 11 per cent. Samples 
4 and 5, also, while containing praftically an equal amount of juice, 
differ by about 10 per cent, in the amount they yield. 
I am not prepared to say definitely why different canes containing the 
FF 
