377 
Sugar-Cane in the West Indies. 
to all appearances identical with (4). When allowed to 
compete together for the same food-material the macro- and 
micro-conidia vanquish the more slowly growing Melanconium 
form. 
A corresponding result was obtained when sterile cane-slabs 
were infected in a similar way to the flasks above. Hence 
it is evident that if the results of flask-cultures of Melanconium 
spores are to be of any value in showing a genetic connexion 
between this form and the macro- and micro-conidia, such 
flasks must be infected from an undoubted pure culture, such 
as a hanging-drop colony developed from one Melanconiwn 
spore. 
The degree of parasitism of the West Indian Melanconium 
was next investigated. In my previous paper (17), infection 
experiments on healthy sugar-canes are described, in which 
spores of this form from a pure culture were found to invade 
the tissues of a cane to some extent when introduced into 
fresh wounds, and the conclusion was drawn that this fungus 
behaves as a parasite and causes the common ‘ rind * disease 
of the sugar-cane. This conclusion has been found on further 
study to be inaccurate. Some of the infected canes in the 
experiments referred to were kept under observation for three 
months, when it was found on examination that the fungus 
had not spread more than 3 inches immediately above and 
below the wounds, and that it had not penetrated the tissues 
of the internode to any great extent. In no case could it be 
traced beyond the internode in which the wounds were made, 
neither did the canes in question show any of the well-known 
appearances of the ‘rind’ disease, such as the drying up of 
the leaves and the discoloration and shrinkage of the affected 
areas. On examining the mycelium it was found to have 
passed into a resting condition. The affected tissues were 
uniformly bright red in colour. 
These results suggested the necessity of further infection 
experiments with this form. These were made, and the 
results were as follows : — 
1. On November 37, eighteen healthy Bourbon canes were 
