Brown. — Studies in the Physiology of Parasitism. III. 403 
principle in the infection drop at all approximate to its concentration in the 
extracts employed in these experiments. 
The above results are contrary to the view of de Bary according 
to which killing takes place previous to penetration of the cuticle by 
the fungus. The discrepancy can only be explained by assuming either 
that the conclusions of de Bary do not hold for Botrytis cinerea , or that the 
method of extraction here employed has failed to obtain some toxic 
substance which plays an important part in the early phases of natural 
infection. All question of the presence of a cuticle-dissolving enzyme may 
be put on one side, as no evidence has yet been obtained that any fungus 
produces such an enzyme ; and certainly no such enzyme is present in 
the extract in question. 
For the application of de Bary’s result to the case of Botrytis attack, 
there is no direct evidence apart from Nordhausen’s statement that brown- 
ing and killing of the cells of a moss leaf took place before the hypha had 
penetrated and even before the spore had germinated. This statement 
has been criticized in another place, 1 and it is very doubtful if any weight 
can be attached to it in this matter. Again, the experiment with stained 
epidermis of Allium discounts the excretion by the fungus of any acid sub- 
stance, at any rate of such a comparatively strong acid as oxalic acid. The 
experiment in which the toxic products of the fungus were removed or con- 
siderably diluted by means of a stream of water can only be readily inter- 
preted according to the view that there is in the infection drop no toxin 
which is capable of passing through the cuticle. The incompatibility 
with one another of certain of Nordhausen’s conclusions renders further 
criticism of his work unnecessary. 
The hypothesis of killing in advance of penetration of the cuticle 
necessitates a crystalloidal toxin. The maximum concentration of this 
toxin would occur in the neighbourhood of the hypha ; but owing to the 
high diffusibility which must be postulated of the toxic substance, it is 
impossible to believe that any considerable concentration gradient of toxin 
could subsist within the limits of the infection drop. A toxic substance 
capable of diffusing through the cuticle on the one side would freely diffuse 
into the spore-free region of the infection drop on the other. The action of 
such a toxin should therefore not be confined to the immediate neighbour- 
hood of the hyphae, and the existence of such a substance should also 
be demonstrable in the general liquid of the infection drop. 
The remainder of this paper is devoted to the examination of the 
infection drop from the point of view of the presence of a crystalloidal toxin 
and in particular of a soluble oxalate. 
1 No. I of this series. Ann. Bot., vol. xxix, 1915. pp. 329, 330. 
