LIFE IIISTOEIES OF RUSTS. 



21 



but slightly changed forms of the same species can not be doubted. 

 Although they infect barley and wheat ])lants with almost equal ease, 

 they behave differently in the other inoculations. That this differ- 

 ence may be attributed to the influence of host on the parasite is 

 clear from the fact that wheat rust after passing to barley behaves 

 in a similar manner to the barley rust, although the latter retains a 

 more versatile character even after passing to the other host plants. 

 Table II (p. 26) throws further light on the relationships of wheat 

 and barley rusts. Diagrams 11 and 12 summarize the inoculation 

 experunents with Puccinia graminis secalis (stem rust) from rye. 



Diagram 11. — Summary of inoculation experiments with stem rust from rye. 



R 



-W -ir, QQ flecked. 



-R 



17 

 25' 



^ 31 



■Bi 



-o 



22 



^ 20 



-r|)- 



"^ 18 



^ 16 



-I- 



-B 



20 



Bi- 



R 12' 12 fl^'^k^d. 



-B 



-Bi. 



Diagram 12. — Summary of the important results shown in diagram 11. 



17 



R 



25 



B 



23 



31 





R 



20 



-i- 



^ 19 



16 



The rye rust infects rye and barley with about equal ease; in fact, 

 the proportion of successful inoculations indicates even greater 

 preference' for barley than for rye, thougli a much larger number of 

 inoculations would be necessary to decide this pohit conclusively. 

 Wlieat is rarely directly uifected. Derr reports having obtained the 

 successful infection of rye rust to wheat in a few instances. These 

 infections were not obtained directly, but 1 out of 22 inoculations 

 was successful after the rye rust had passed to barley. In our own 



216 



