230 Journal of Agricultural Research voi. xv, N0.4 



In diagram 2 both wheat and rye became infected when inoculated 

 with the rust from Hordeum juhatum. The rust developed on wheat 

 proved to be a pure Strain of tritici. That developed on rye, however, 

 consisted of both the tritici and secalis forms. Very clearly the original 

 rust on Hordeum jubatum was a mixture of the tritici and secalis forms. 

 The tritici form was isolated in pure form by transferring to wheat on 

 which the secalis form did not develop. But the first generation of the 

 rust on rye was still mixed, since tritici develops weakly on rye. Appar- 

 ently the second generation of the rust on rye was pure secalis, since it 

 did not develop on wheat, the leaves of which, however, died young. 

 But, after having passed four generations on barley, the rust infected 

 wheat normally and infected rye only weakly. This clearly looked like 

 bridging. The more probable explanation, however, is that .both the 

 secalis and tritici forms again developed the second generation on rye. 

 Since the wheat plants died young, the negative results recorded are not 

 significant. 



In the subsequent inoculations both secalis and tritici probably devel- 

 oped on barley, since a high percentage of inoculated leaves of Agropyron 

 repens became infected. The secalis form, however, was eventually lost. 

 This may have happened in two ways. The tritici form may have 

 developed more rapidly than the secalis; or material from the leaves 

 infected with secalis may not have been used in making inoculations 

 It is quite probable that if rye had been inoculated earlier both forms 

 would have been isolated. 



From the R^ — ^W^ material both biologic forms were isolated. This 

 was puzzling at first, because wheat is not a host for the secalis form. 

 The only plausible explanation seemed to be that spores of both biologic 

 forms were placed on the wheat during inoculation and not all germinated 

 in the moist chamber. A few viable secalis spores therefore remained 

 on the wheat, and when these were transferred to rye, they germinated, 

 causing infection. In order to ascertain whether this was possible, 

 wheat was inoculated with the secalis form, kept in a moist chamber for 

 48 hours, and then kept on a greenhouse bench for about 10 days. No 

 rust developed, but the inoculum was scraped from the leaves and used 

 to inoculate rye plants. Infection resulted on a relatively large number 

 of leaves, showing that the theoretical explanation advanced above was 

 probably correct. These facts show that extreme caution is necessary 

 in drawing conclusions when dealing with mixed forms of rust. 



The most convincing proof that the rust forms, after isolation in pure 

 form, could not be changed by host influences is furnished by the subse- 

 quent history given in diagram 2. The tritici form did not acquire new 

 parasitic ability on account of its association with barley; nor did it 

 increase in virulence on rye as a result of successive transfers. It will be 

 noted that repeated unsuccessful attempts were made to develop a 



