On the Physiology of Parasitism. 113 
large number of plant structures. The following general conclusions 
are drawn:— 
Fleshy structures are very generally susceptible to the action 
of the extract. Among such have been tested the tissue of potato, 
swede, turnip, radish, apple, cucumber and a number of succulents, 
(species of stonecrop, Cotyledon, etc.). 
Petals and floral structures generally are rapidly destroyed. 
Upwards of thirty species of plants have been tested in this respect, 
and in all cases a rapid destructive action on the part of the extract 
has been established. It is highly probable that this behaviour is 
characteristic of floral structures generally. 
Among ordinary foliage leaves a strong action of the extract 
could be established in some cases, e.g., leaves of broad bean, lettuce, 
violet, Petunia, various speciesof Begonia, etc. In other cases the 
action was of a much less marked character; with leaves of a 
leathery or woody texture no definite action could be established in 
any case. 
The extract so far as could he seen had no action whatever on 
tissues of mosses or liverworts. It was also without action on 
filaments of algas and on bacteria. 
The detailed features of the action of the extract were followed 
out in the case of certain tissues (potato, turnip, cucumber, etc.), the 
plasmolytic criterion of death being here employed. It was shown 
that the first demonstrable action consisted in the solution of the 
middle lamella uniting contiguous cells, with the result that 
coherence of the tissue was destroyed. The attack was, however, 
not confined to this, and the general body of the cell disintegrated, 
though complete solution of all constituents did not take place. At 
a comparatively late stage in the process the cells lost their power 
of becoming plasmolysed in hypertonic solutions. Thus, of the two 
manifestations of the action of the extract, the toxic is subsequent 
to the enzymic. 
In dealing with the action of the fungal extract upon tissues, 
special attention was directed to the post-mortem features evinced. 
It was found in all cases that these were identical with those induced 
by the action of the fungus itself. Again, in all cases where a 
distinct parasitism of the fungus on a particular host could be 
established, it was found that the tissues of the latter were acted 
upon in a similar way by the fungal extract; and where the extract 
was without action (e.g. mosses and liverworts), no definite action 
could be established on the part of the fungus itself. These con- 
