io8 
XV. G. Freeman. 
centage, a comparatively few a very high percentage, but that the 
greater number would more or less approximate to the mean richness 
of sucrose for the particular variety. Supposing now that the plants 
richest in sucrose are selected to form the parents of the next 
generation it would appear probable that their sugar contents 
would be found to show similar differences, a few would be high, 
a few low, but the greater number would again approximate to the 
mean of this series, and the mean of this the second generation 
should, theoretically, be slightly higher than that of the first 
generation from which it has been raised by selection of the richest 
canes. Imagine this process repeated for many generations, 
resulting each time in slightly raising the mean sucrose contents, 
and we see that there appears to be at hand a method of increasing 
the yield of sugar from a given variety of cane. The method is a 
vegetative or asexual one exactly as bud variation, but whereas in 
the latter case all the experimenter can do is to put to the test any 
cane which shows unusual characters of colour, etc,, which may or 
may not be correlated with variations in the richness of the cane in 
sucrose, in chemical selection he is enabled to proceed along well 
defined lines, and to work steadily towards a definite goal, the 
production of a race of canes of increased sugar content. Any 
success in bud variation depends on taking advantage of naturally 
occurring large differences, in chemical selection reliance is placed 
on increasing the sucrose yield by the integration of small 
differences, which are only recognizable by the chemical analysis 
of the juice of the individual canes. 
The mode of conducting experiments, outlined above, appears 
easy, but in actual practice many difficulties are encountered and it 
is by no means a simple matter to carry on work in chemical 
selection of sugar canes so to obtain results which shall be 
absolutely reliable and above criticism. In the first place different 
portions of an individual cane yield juice varying greatly in 
composition ; at any particular time the younger internodes are 
relatively rich in glucose and poor in sucrose, whilst the older or 
basal points are relatively poor in glucose and rich in sucrose. 
Again a sugar-cane plant comprises a number, say ten to twenty, 
of separate canes or stalks which may vary, as will be shown later, 
very considerably in the richness of their juice when examined 
simultaneously, and there is no reliable criterion to enable one or 
other cane to be selected as representative of the whole plant. 
Finally, to mention any one other difficulty, it is essential that the 
