330 
meating the inner cortex and bast almost to the first branches 
and downwards into the roots. Sclerotia were found on the out- 
side of the bark. 
These cases present a number of interesting comparisons 
with the diseased Aesculus Pavia. In the former the infection 
occurred at the soil level presumably by Botrytzs growing sapro- 
phytically in the soil and the mycelium spread slowly upwards 
destroying the cortex. In the diseased Aesculus the infection 
took place probably by a spore some twenty-six centimetres above 
the soil level and the mycelium spread rapidly in all directions, 
penetrating the medullary rays as far as the pith. In the goose- 
berry, lime trees and the larch seedlings there was no direct 
effect on the transpiration stream and the eventual death of the 
upper portion of the hosts was merely part of a general necrosis. 
In the Aesculus the cutting of the water supply to the upper part 
of the tree was rapid and complete, and although not a direct 
result of mycelial thrombosis, was due to a choking of the con- 
ducting elements by tyloses formed as a direct reaction to fungal 
stimulation. The diseased lime saplings normally showed only 
a form of conidiophorous stromata, whilst in the gooseberry and 
larch seedlings true sclerotia only were present; these, however. 
in the latter apparently very shortly giving rise to conidiophores. 
In the specimen of Aesculus both types of structure were present 
normally and bore a very definite and constant relation to the . 
lenticels. 
Thus the disease of lime trees, gooseberry bushes and larches 
have much in common, but apart from their similar etiology show 
little resemblance to the diseased Aesculus Pavia under considera- 
tion. 
ConcLuUsION. 
The points of interest in this specimen to which attention may 
e drawn are as follow :— 
The position on the tree at which infection occurred. 
The symptoms of the disease and the rapidity with which the 
host, a comparatively large woody tree, was killed. 
The distribution of the fungus in the tissues and especially its 
absence from the water-conducting channels. 
The induction of the formation of tyloses which created an 
effective barrier to the transpiration stream. 
The development by the fungus of conidiophorous stromata 
and of true sclerotia, and the constant relation of these to the 
lenticels. 
A number of very interesting features were presented by the 
‘morphology of the fungus in the tissues, and these will be 
described in a further communication. 
I am glad to record my indebtedness to Miss M. N. Owen, 
Temporary Technical Assistant in the Laboratory, for the pre- 
paration of many slides upon which much of the information 
in the present paper is based A ens 
