The Artificial Infection of Plants. 117 
with a dry camel-hair pencil, taking care to use only those 
which will readily brush off. The contents of the watch- 
glass may then be applied to the wheat plants. By this 
method the accidental infection of the control plants is 
avoided, because the bell-glass is never removed from 
them. 
Personally, I have found infection with promycelial 
spores more certain than with the zcidiospores, because 
we can see that they are actually germinating at the time 
they are used, while with the ecidiospores this cannot 
be done so certainly ; moreover, the zcidiospores being 
generally brought from a distance, they are apt to lose 
this germinative power, unless used immediately, from 
becoming too dry on the one hand, or on the other, if kept 
in too moist an atmosphere during their transit, from many 
of them having already germinated before they are em- 
ployed. It is only by attention to these minute details 
that we can ensure uniform success. 
The Gymnosporangia are very easy to cultivate. A 
few seedling hawthorns can be obtained anywhere, and it 
is necessary only to soak the Gymunosporangium clavarie- 
JSorme for twelve hours, when the golden promycelial spores 
will be visible to the naked eye. 
In producing R. cancellata on pear, it is necessary to 
infect two-year old plants, because if seedlings be infected 
the spermogonja alone will be produced, because seedling 
pears lose their foliage before the Reestelia has had time 
to develop ; with these plants it is essential to success 
that they should be thoroughly established before they are 
made the subiect of experiment. 
In working out the life-history of the allied species 
duplicated cultures are very valuable. Suppose we wish 
to produce the ecidia of P. magnusiana and P. phragmitis. 
Having provided the proper material and a number of 
