488 Freeman . — Experiments on the Brown 
of doors, where the bell-jars were removed gradually to harden 
the plants to change of conditions. The spores used for 
infection were obtained from plants (both wild and cultivated) 
of B. sterilis and B. mollis from the University Botanical 
Gardens at Cambridge. Care was taken to prevent the ad- 
mixture of leaves of other species. The cut-off leaves were 
placed in closed glass chambers for twelve to thirty-six hours 
before the spores were used. The determining condition of 
germination of the spores in distilled water are still enigmatical 
and have evidently narrow limits. Spores from the material 
used for each separate infection were tested in distilled water 
twenty-four to forty-eight hours. In most cases germina- 
tion is abundant. Occasionally, however, apparently excellent 
spores may, under seemingly favourable conditions, almost 
utterly fail to germinate in distilled water, while spores from 
the same or neighbouring sori of the same leaf may success- 
fully infect a Brome plant. Experiment 77 q" is a very good 
example in point 1 , where no spores germinated in distilled 
water, while the infection of B. mollis with the same material 
was in each case (i. e. four out of four) successful. For this 
reason the negative results in the distilled water tests are not 
always an indication that the spores were incapable of germina- 
tion. Dry spores were used in most cases, and in rubbing 
them on to the leaves of the host there arises the danger of 
dropping or shooting spores on to other parts of the plant or 
upon the control plants. Such may explain the infections 
of control plants in Experiments 94-96, in which cases an 
abundance of spores was used. Such infection may also be 
caused by the rubbing together of two leaves upon one of 
which spores have been placed (Marshall Ward, 1. c.). The 
plants after infection were examined at intervals of several 
days from about the eighth to twelfth day onward. 
The weather throughout almost all of these experiments 
was colder than normal, and the long incubation period of 
most infections as compared with those noted by Marshall 
1 These numbers refer to the tables of details of the experiments not published 
here in full. 
