^34 



Journal of Agricultural Research 



Vol. X, No. 9 



It is quite evident from Table I that Agropyron caninuni is a host for 

 both P. graminis triiici and P. graminis secalis. It will be observed 

 that in No. 2 only P. graminis secalis was obtained from the grass, while 

 in No. 4 and 5 probably only P. graminis iritici was present, since the 

 uredinia on rye were very small and were surrounded by dead leaf areas, 

 thus being typical of the uredinia of P. graminis tritici on rye.- It is 

 possible that both biologic forms were present in No. 3, although the 

 character bi infection pointed to the probable presence of P. graminis 

 tritici only. 



While Agropyron caninum is fairly common in the region covered 

 by the survey, it is probably of secondary importance. It is by no 

 means as common as some of the other species of Agropyron, and, 

 although often severely rusted, it can not be considered in the rank of 

 first importance as a means of enabling the rust to spread to cereals. 



Table II. — Results of inoculations with urediniospores from Agropyron cristatum 



The results of the one series of inoculations given in Table II show that 

 Agropyron cristatum may be infected with both P. graminis secalis and 

 P. graminis tritici. While only one uredinium developed on wheat, sub- 

 sequent events show clearly that both forms were present on the grass 

 when collected from the grass garden in which it was growing. 



It will be noted in Table II that barley became heavily infected. The 

 rust on the barley proved to be partly P. graminis tritici and partly P. 

 graminis secalis, mostly the latter. This was shown by a number of suc- 

 cessive transfers to barley, wheat, and rye. Both forms were eventually 

 isolated and, after isolation, remained pure; although, if extensive inocu- 

 lation experiments had not been made, it would have appeared that 

 barley might enable P. graminis secalis to infect wheat. 



It is quite apparent from Table III that the common form of stem- 

 rust on Agropyron repens is P. graminis secalis. Of the 307 leaves of 

 wheat which were inoculated only 3 developed uredinia, while not a 

 single one of the various trials on rye was unsuccessful, although in a 

 few cases not all of the inoculated leaves became infected. This is 

 not surprising, however, since many of the inoculations were made 

 during the hottest weather of the summer, when conditions for infection 

 were unfavorable. Rye plants vary considerably in their suscepti- 

 bility to P. graminis secalis, probably on account of their heterozygous 

 character, and this naturally accounts for some of the variability of 

 infection. 



