402 Brown. — Studies in the Physiology of Parasitism. III. 
principle of the extract through the cuticle was not observed to take place. 
Such a negative result was invariably found in experiments with leaves 
and petals (both surfaces) of Viola , Petunia , Dahlia , leaves of Vicia Faba , 
Begonia heraclaefolia , &c. In a number of cases a certain amount of 
variation in the results obtained was noticed. Thus, with some specimens, 
no entrance whatever took place ; with others of the same species a certain 
amount of action was observed. This took the form of discoloured spots, 
varying in number and extent, but in no case did anything approximating 
to a general discoloration spreading over the surface of contact make 
its appearance. Such discoloured spots were occasionally obtained in 
.tests on petals of Tropaeolum and Geranium ; and more frequently in 
the case of petals of Rosa and Fuchsia . Nevertheless, even in rose 
petals which showed many discoloured spots after twenty-four hours’ 
treatment with the extract, portions could be found which were quite 
unaltered. 
Bearing in mind the possibility of the existence of previously un- 
detected injuries of the cuticle (such as insect bites, &c.) 1 as well as the 
possibility of accidental contamination with living spores which is not 
always readily detected, we may state that there is no action on the part 
of the fungal extract when applied to the cuticle ; in other words the 
cuticle, apart from accidents which may occur in cases with some fre- 
quency, constitutes an impassable barrier to the passage of the active 
principle of the fungus. The experiments with petals of Viola , which 
are readily attacked by the fungus, show, moreover, that the germ tubes 
can easily infect a tissue the cuticle of which is completely impervious to 
the fungal extract. 
In the preceding experiments the extracts employed were such that on 
injection into the various tissues the latter were disintegrated in the 
striking manner described in the first paper of this series. Furthermore, 
there was no question of the extract having lost its activity during the 
period of the experiment. One particular experiment may be cited in 
illustration of this point. 
A fungal extract when laid on a rose petal for twenty-four hours pro- 
duced no change whatever. After this lapse of time, injection of the same 
liquid into the same petal produced complete disintegration within half an 
hour. 
Finally, in a parallel series of experiments with infection drops con- 
sisting of a dilute suspension of spores in a dilute nutrient, discoloration 
was found to take place in 12-24 hours from the time of sowing. An 
investigation shortly to be published has shown that at no time during 
the period from sowing to discoloration can the concentration of active 
1 Such injuries may occur very frequently, even in apparently healthy leaves. See in this 
connexion a paper by Barker and Gimingham, Annals of Applied Biology, vol. i, 1914, p. 13* 
