The Diffusion Process. 57 
The centrifugal force then throws the chips of cane out 
to the largest diameter, where they escape over the 
edge and are collected by a suitable guard. I believe 
that this admirable machine is actually at work at the 
Pin. Mon Repos in the island of Guadeloupe, with ex- 
cellent results. 
As regards the durability of the knives, we learn that 
in M. Robert's machines they are sharpened every 24 
hours, or more often if the canes are very hard ; ( but in 
any event it takes but a few minutes to remove the dull 
knives and replace them by others kept in reserve.' 
This, observe, in a machine with only six knives, and 
cutting up a minimum of 6,ooolbs. of cane per hour; so 
the operation of sharpening does not appear to be very 
formidable. I confess I do not like this frequent chang- 
ing of knives : it is evident that the knife edges must 
lie accurately in one plane in the case of the disc 
machine, or accurately in the surface of a cone if a drum 
be employed, otherwise the slices will not be uniformly 
thin : and to obtain this exactness by hand would be 
troublesome if not impossible, and would involve 
unnecessary manual labour. Had I the control 
of such machinery I would certainly sharpen all 
the knives in place by an emery wheel traver- 
sing on a slide-rest, which would rapidly make 
every knife-edge true to a hairs-breadth. This might 
be done with the wheel in place, or two wheels might be 
used alternately, the idle one being sharpened on a 
suitable headstock. The change of wheels would only 
take a few minutes, about as long as it takes a turner 
to change his chuck for his face-plate in a larger 
lathe. The knife should project, so far beyond the 
H 
