166 Transactions British Mycological Soctety. 
The above results clearly show that the optimum tempera- 
tures for the growth of the mycelium in the three rusts are 
different. Black rust undoubtedly flourishes better at higher | 
temperatures and cold has a pronounced inhibitory effect upon 
its growth. The brown rust occupies an intermediate position, 
whereas the yellow one can flourish at low temperatures, 
Further the results of the first experiment suggest a possible 
explanation of the sequence in the appearance of the yellow and 
the brown rusts. It has been pointed out that in this country 
yellow rust usually appears before the brown one. The reason 
why black rust does not appear till late in the season is that 
the source of fresh infection of cereals lies in the aecidiospores 
on barberry and the aecidial stage does not appear till late in 
spring. 
g. INFECTION FROM COLLATERAL Hosts. 
Eriksson* has pointed out that the spread of rust from one 
kind of cereal or grass to another is considerably restricted on 
account of specialization of parasitism. 
Similarly Butlert has remarked that wild grasses play no 
part in carrying rust in the main wheat growing tracts of India. 
Stakman and Piemeisel} have stated that grasses aid materi- 
ally in the spread of black rust on cereals. These authors have 
also remarked that, even if overwintering of this rust on grasses 
is infrequent, grasses may still be important because they may 
be nearer barberry bushes than are the grain fields. 
As far as this locality is concerned the overwintering of black 
rust, whether on cereals or grasses, is very doubtful. The possi- 
bility of the fresh outbreak of the rust each year due to infection 
from other cereals or grasses is therefore out of the question. 
Besides it has been stated above that in wild grasses it has been 
observed only on Couch grass. 
Yellow rust has been observed by the writer only once on 
Agropyron repens and Agropyron camnum but Miss Sampson 
sent some specimens of Dactylis glomerata from Aberystwyth 
which were found to be infected. 
Speaking generally one might state that on account of a more 
rigid specialization of parasitism in the case of both brown and 
yellow rusts the possibility of other hosts playing a part in their 
perpetuation is limited. A detailed discussion of this question 
as far as this locality is concerned will be taken up in the second — 
part of this paper. 
* Eriksson, J., Bot. Gaz. xxv, p. 26 (1898). 
+ Butler, E. J., Fungi and Disease in Plants (1918). 
+ Stakman, E. C. and Piemeisel, F. J., Journ. Agr. Res. x, p. 429 (1917). 
