379 
Sugar-Cane in the West Indies. 
were doubly inoculated — at wounds in an upper and a lower 
internode — with actively growing mycelium of the fungus 
from pure cultures. Four other canes were used as controls. 
On January 22, the results were almost identical with those 
obtained above. 
4. On December 21, four healthy White Transparent canes 
were doubly inoculated, at wounds made at an upper and 
a lower internode, with Melanconium spores from a pure 
culture. On January 27, it was found that the fungus had 
not penetrated beyond the column of tissue containing the 
wound and bounded by the nodes above and below. The 
control canes showed no infection. 
These experiments point to the non-parasitic character of 
the West Indian Melanconium towards the sugar-cane. The 
behaviour of this form, therefore, closely resembles that of the 
Java Melanconium studied by Went (13), and the whole of the 
evidence points to its being a saprophyte. 
A careful examination of a large number of canes attacked 
by the ‘ rind * disease showed that after the outer leaves begin 
to dry up at their margins — the first indication that canes 
are attacked by this disease — they gradually die in from 
four to eight weeks. When the leaves are about half dried 
up, the black filaments of Melanconium can be detected burst- 
ing through the discoloured areas of the stem. In many 
cases this is the only fungus which at this stage can be 
seen on the affected stems. The inoculation experiments 
described above, however, point to the conclusion that Melan- 
conium cannot be regarded as the cause of the ‘ rind ’ disease. 
The difficulty was solved by the study of another fungus 
which was found on the affected canes, and which in previous 
studies of the West Indian ‘rind’ disease had been over- 
looked. Before considering this fungus it remains to deal 
with the macro-, micro-conidial, and the ascigerous phase of 
Trichospheria Sacckari, Massee. 
C c 2 
