Aug. 27, 191 7 Puccinia graminis on Cereals and Grasses 475 



experiments, are P. graminis tritici and P. graminis secalis, both of which 

 infect H. juhatum, easily and both of which are commonly found on the 

 grass in the field. Freeman and Johnson (14, p. 21, 28) showed that 

 stemrust of rye infects both rye and barley, but did not work with the 

 grass hosts. Pritchard (22, p. 181), as a result of work done in North 

 Dakota, concludes — 



that one form of P. graminis is common to Hordeum jubatum, Agropyron tenerum, A. 

 repens, Avenafatua, oats, and rye, but is incapable of infecting either barley or wheat. 



Pritchard very probably worked with two forms — viz, (i) P. graminis 

 secalis, which probably caused the infection of Hordeum jubatum, Agro- 

 pyron ienerum, A. repens, and rye; and (2) P. graminis avenae, which 

 infected the oats, and A. fatua. 



In Sweden Eriksson (11, p. 601) did considerable work on the grass 

 hosts of rye stemrust, and gives the following as hosts: 



Secale cereale, Hordeum vulgare, H. jubatum, H. murinum, H. comosum,, Triticum 

 repens, T. caninum, T. desertorum, Elymus arenarius, E. sibericus, and Bromus seca- 

 linus. 



It is quite apparent that this rust in Sweden is very similar, if not 

 identical, with t*hat in this country. Eriksson mentions Triticum {Agro- 

 pyron) desertorum as a host; the writers were not able to infect this host 

 in the greenhouse, but the number of trials was too small to justify 

 definite conclusions. While the species of grasses with which Eriksson 

 and the writers worked are not all the same, the same genera were investi- 

 gated, and some species of these genera were found as hosts for the rusts 

 in both countries. It is very probable, although by no means certain, 

 that most of the species of Agropyron, Hordeum, Elymus, and Hystrix 

 are susceptible to rye stemrust. 



Jaczewski's results in Russia do not agree closely with those of Eriks- 

 son nor those recorded in this paper. He mentions Secale cereale (18, 

 p. 353) as the host, and states that the rust is capable of infecting Triti- 

 cum repens, T. caninum, and Dactylis glomerata. The w^riters were unable 

 to infect D. glomerata with stemrust of rye, although only 46 leaves were 

 inoculated. It is entirely possible that the rust can infect D. glomerata 

 weakly, but it is improbable that it does so commonly enough to be of 

 practical significance. Jaczewski states that Hordeum vulgare is immune 

 to the rust. In this country it is very susceptible. 



The stemrust of rye in this country easily infects rye and barley, but 

 very rarely, or scarcely at all, wheat and oats. Of the wild grass hosts 

 Agropyron repens is by far the most common. It is nearly always 

 severely affected. Although A. repens can be infected weakly by P. 

 graminis tritici also, it is doubtful if the wheat stemrust occurs on it, 

 unless quite exceptionally, in the field. In general the various species 

 of Agropyron, Elymus, Hordeum, and Hystrix are hosts for the rust. 

 It is found commonly on at least some species of all these genera in the 

 field. 



