Some Experiments on Sugar Cane. 71 
the number of degrees Bm. A comparison of the 
figures in the column headed " B° * 1*65" with those under 
the heading " per cent, of total sugar " will show the 
degree of accuracy obtainable. As before mentioned it 
is likely that even more accordant figures would be 
obtained by determining a fa6lor for a particular 
kind of cane instead of using a single one for the 
juice of many kinds. Such a fa6lor affords a very sim- 
ple means of ascertaining the approximate value of 
juice when a polariscope is not at hand, especially 
if supplemented by the copper test for glucose. Even 
when a polariscope is employed it is useful for check- 
ing the analytical results arrived at. For example, 
in Professor HARRISON'S report already mentioned, the 
quotients resulting from the application of the rule to his 
analyses of cane juice are in most instances much too 
large. The quotient obtained by dividing Bm. de- 
grees into pure sugar solutions of a corresponding 
strength is i'8o, but a higher quotient than that is ob- 
tained in no less than five instances from Prof. HAR- 
RISON'S results, while many others closely approach it. 
Therefore, either the determinations are incorreft, or, 
what is absurd, there must have been more sugar in the 
cane juice than would be present in pure sugar solutions 
of the same density, notwithstanding the presence of 
large quantities of non-saccharine matter. Examining 
the analyses further, it appears that other constituents 
of the juice have been over-estimated, for the per- 
centage of total solids present in every case is consider- 
ably greater than juice of the given density would con- 
tain in solution. This^is shown by the table on page 74, 
which contains a summary of Prof. HARRISON'S analyses 
