200 
R. H. Compton. 
has proved the existence in plants of antitoxin reactions quite parallel 
with those observed in animals. The visible effects of the develop¬ 
ment of the pollen-tube within the style, namely the enlargement 
of the ovary, the withering of the stigmas, etc., are strikingly 
suggestive of the phenomena of gall-formation (often so specific and 
stereotyped), hypertrophy and degeneration under the attacks of a 
fungal parasite. Both in gall-formation and in fertilisation the 
chemical nature of the stimuli and reactions has so far eluded 
investigation. 
If there be an analogy between fertilisation and infection, there 
is also one between sterility and immunity. There seem to be two 
main methods by which immunity is attained. In one the host is 
constitutionally resistant to the attacks of the parasite, this being 
the more frequent method. In the other the host is over-susceptible, 
and the attack of the parasite produces a quantity of decomposing 
matter which inhibits its further growth. Both these methods are 
paralleled in cases of failure of fertilisation. The first is the more 
usual: the pollen grains fail to germinate, or they germinate but 
the tubes do not flourish. The second occurs in certain cases of 
pollination from a distinct species, as recorded originally by Gartner 
(1849), where the pollen appears to act injuriously on the stigmas: 
and also in the remarkable cases of self-sterility in certain Orchids 
mentioned above, where the same phenomenon occurs. 1 
The immunity of certain races of Wheat to the rust disease 
apparently belongs to the second category: i.e., it is the result of 
over-susceptibility. The first attack of the fungus hypha causes so 
much decomposition of the tissues of the host in its immediate 
neighbourhood that its extension into the adjacent tissues is checked 
(Marryat, 1907): much in the same way that the wall of an abscess 
prevents the bacterial growth spreading far from the original seat 
of infection. The work of Biffen (1907) has shown that susceptibility 
to rust in Wheat behaves as a Mendelian dominant to immunity. 
Therefore, on the presence and absence theory, susceptibility is 
due to the presence of some substance which immune individuals 
lack. This substance thus seems to act as a partial check to the 
too energetic onslaught of the parasite in the susceptible plant. It 
might be expected that the same would obtain in the cases of self¬ 
sterility through decomposition ; but here evidence is lacking. 
1 In one case, that of Oncidium flexuosum , self-pollination and pollination 
from a species ot Notylia both produced the same injurious effect (Darwin 1905 
ii, p. 145). 
