BRIDGING SPECIES 71 
case out of many experiments could B. madritensis be infected 
directly from B. mollis. B. sterilis therefore acted as a 
“bridging” species, and enabled the parasite to pass from 
B. mollis to B. madritensis, though it could not do so without 
this intervention. The same existence of “bridging” species 
has been demonstrated in Erysiphe graminis, and no doubt 
will be found in numerous other instances. 
Ward further mentions (1903) that he found B. ardwennensis 
var. villosus to be infectible by the spores from B. sterilis, B. 
mollis, and B. patulus, as well as by those from B. arduennensis, 
and therefore easily able to serve as a bridging species between 
these others. Nevertheless, that such intermediary species 
do not exist in all cases is proved by the fact that when he 
grew “more than 200 species and varieties of Bromus side by 
side or intermingled in contiguous beds, certain species invari- 
ably caught the disease and became rusted, while others close 
by showed no sign of infection.” 
The occurrence of these abnormal spores, i.e. mutations, 
which is proved in the case of Bromus, is of great significance 
and gives us the clue by which we can understand how a 
gradual or sudden passage can take place, and has taken place, 
from one host to another, so that now an appreciable percentage 
of the modern vegetable world have parasites more or less 
specialised to themselves. 
Another important consequence follows from this fact of 
specialisation. If the parasite is so narrowly adapted to its 
particular host, it may be expected that varieties of the host 
can be found or bred which will be able to resist attack, that 
is, will be immune. A great deal of research has been devoted 
of late years, especially by Professor Biffen of the Agricultural 
School at Cambridge, to this subject of breeding a race of 
wheat which will be immune to Rust, and a certain amount of 
progress has been made. Immunity depends chiefly (perhaps 
entirely) upon the ability of the cytoplasm of the host-cells 
to resist infection by secreting anti-toxins which will kill the 
mycelium of the fungus. Immunity and susceptibility (which, 
however, seem to be always relative only and not absolute) have 
been proved to be inherited, and in fact to be Mendelian 
