LEAF AND SEEDLING DISEASES 177 
by the uredo-stage, which can spread from leaf to leaf 
throughout the summer, but by the aecidial stage, which 
cannot spread by itself. Some sporidia will blow up from 
the ground and infect a few larch needles, and these needles, 
which bear spores as early as three weeks after expanding, 
soon wither and die. But that is the end of it as far as the 
larch is concerned. Once the aecidiospores have attacked 
the birch, the fungus can spread from birch to birch as long 
as the summer lasts, and no limit is set to its distribution. 
Probably this is the reason why the fungus is found so much 
more commonly on the birch than on the larch. The disease 
on the larch is only serious when seedlings are attacked. 
Massee (1903) reports a case where rows of larch seedlings 
were severely damaged by this rust. The owner, believing 
in mixed planting as a panacea for plant diseases, put in 
rows of birch seedlings between the larch, with the result 
that any one familiar with the life-history of the fungus 
would have anticipated. Of course birch is the worst con- 
ceivable tree to bring into the neighbourhood of this rust, 
and where the disease is troublesome in a nursery it is 
advisable to destroy.the birch as far as possible in the 
immediate vicinity. As the mycelium spreads. entirely 
inside the leaf and not on the surface, spraying can only 
be effective if the mixture is on the leaf at the time of 
infection, so as to prevent the germination of the sporidia. 
This requires such thorough and constant application to be 
successful that it is outside the scope of practical sylvi- 
culture. 
Melampsora tremulae, Tul, &c. ( =Coeoma laricis). Coeoma 
is another aecidial form which closely resembles Peri- 
_ dermium, except that it possesses no peridium, and is 
consequently far less conspicuous. The small red spots or 
cracks which it produces in the epidermis of larch needles 
and cotyledons are no doubt often overlooked. The uredo- 
stage is borne on species of willow and poplar, and four 
distinct species, as well as many sub-species, are recognized 
in the uredo- and teleuto-forms, though there are no features 
by which these species can be distinguished in the coeoma 
1388 N 
