174 Transactions British Mycological Society. 
made several inoculations with a pure culture of the yellow 
rust of wheat and found that barley and rye are not infected 
by it. Some experiments with the yellow rust of barley were 
conducted in the summer of 1921, but on account of very 
hot weather the infections were unsuccessful even on barley 
itself, and the pure culture was lost. Yellow rust was noticed 
on rye in this locality last summer (1921) but an attempt to 
establish a culture failed on account of the great heat, nor was 
it possible for any work to be done with the form on Couch grass. 
I wish to express my warmest thanks to Mr F. T. Brooks, 
under whose supervision these investigations have been made, 
for advice and the interest he has taken in their progress. I am 
also very grateful to Mr F. L. Engledow and Mr S. F. Armstrong 
of the School of Agriculture for having provided me with seeds 
of the different varieties of cereals used in the experiments. 
August 1922. 
Io. SUMMARY. 
Pan I. 
(a) Black rust (P. graminis). 
(x) Fresh uredo-sori of this rust are not found after the com- 
mencement of winter. Uredospores in old pustules exposed to 
winter cold soon lose their viability. 
(2) Observations and experiments show that this rust cannot 
overwinter in this locality even as mycelium inside the host 
plants. 
(3) Direct infection of wheat by sporidia is not possible. 
(4) There is no evidence to show that this rust can originate 
from “‘mycoplasm” or from intra-seminal sori. 
(5) The annual outbreak of this rust can be explained only 
by fresh infection through aecidiospores produced on barberry. 
(b) Brown and yellow rusts (P. triticina and P. glumarum). 
(6) Fresh uredo-sori of the brown rust of wheat and the 
dwarf rust of barley are found during the greater part of winter. 
Uredospores from the open during winter always show good 
germination. 
(7) Fresh uredo-sori of the yellow rust of wheat were found 
during the greater part of the winter of 1920-21. Uredospores 
from such pustules always germinated well. 
(c) Incubation period. 
(8) For the annual recurrence of rusts the factor of the 
greatest importance is the occurrence of plenty of uredospores 
on self-sown plants and tillers at the time when the autumn sown 
crop appears. The infection of young seedlings is followed by 
