Some Experiments on Sugar Cane. 
By E. E. H. Francis. 
N June 1S85, I had the honour of reading a 
paper before the Royal Agricultural and Com- 
mercial Society showing that the proportion 
of sugar present in sugar cane was greatly overstated 
in books because the statements were based on mere 
calculations and not on a6lual determinations. Instead 
of average sugar cane containing "18 to 21 per cent- 
of sugar", as the standard text books say, it was 
intimated that the amount of sugar in canes of the 
richest quality w 7 ould seldom be found to exceed 16 per 
cent. On the same grounds, obje6lions were made to 
certain " analyses" of canes made in this colony, and 
particularly to one of cane from Barbados, in which the 
proportion of sugar was given as 20*23 P er cent* 
At the time the paper was written I had had no oppor- 
tunity of analysing Barbados canes, and going on leave 
* The error arising from the substitution of calculation for " deter- 
mination" was subsequently fully acknowledged by the analyst in 
this case. Also Professor Harrison of Barbados, in his report for 
1886 on the " Results obtained on the Experimental Field at 
Dodd's Reformatory," says, (p. 5):— " Another important point will 
be noticed upon examining Table No. 7 in which the distri- 
bution of the constituents in the juice and the megass is shown. 
It will be seen that the proportion of total sugar (crystallisable 
sugar and glucose) left in the megass is very much less than has 
usually been assumed. It has been an almost universal practice 
amongst chemists to calculate the proportion of sugar in the whole 
cane from the amount present in the juice extracted, by assuming that 
that left in the megass possessed the same composition. There exists 
