Oct. 28, 1918 Plasticity of Biologic Forms of Puccinia graminis 241 



The tritici strain used in this series of inoculations (diagram 9) was 

 isolated originally from Hordeum jubatum at St. Paul, Minn., on Septem- 

 ber 29, 1914, and was kept in uredinial culture until June 15, 1917. 

 Barley, Elymus canadensis, Bromus tectorum, and Hordeum jubatum 

 were all used as intermediary hosts, but none of them enabled the rust 

 to do anything which it could not do at the begirming of the experiment. 

 The results of inoculations with the rust developed on Bromus tectorum 

 are particularly interesting, since this grass is a host for the six common 

 biologic forms of P. graminis in the United States. Theoretically when 

 the triticiiorm was transferred to it, since it is a congenial host also for 

 the secalis and avenae strains, it should have transferred to rye and oats. 

 However, it did not. 



Diagram 10. — Results of inoculations with Puccinia graminis avenae from Dactylis 



glomerata. 



36 



Puccinia graminis avenae from Dactylis 

 glomerata, St. Paul, Minn., Aug. 22, 1914. 



Brotnus tectorum 



67 



Bromus tectorum. 



Bromus hordeaceus - 



Oi— Bromus tectorum — — 



' Minute uredinia. 



Bromus erectus—^ 

 June 14, 1917. 



The avenae strain represented in diagram lo was isolated originally 

 from Dactylis glomerata at St. Paul, Minn. It has been kept on oats for 

 about three years. A large number of inoculations were made on various 

 hosts during that time, but it has never performed differently from the 

 rust on Dactylis glomerata. Bromus tectorum was used as a bridge in 

 attempts to induce the rust to pass more readily to barley and rye, 

 since it is a host for both the tritici and secalis forms as well as for the 

 avenae form, but it is clear that it did not do this. The uredinia developed 

 on barley were minute, and the percentage of infection was no greater 

 than that which is obtained from transfers made directly from oats or 

 any other host for the awwae form, a large number of which have been tried. 



EXPERIMENTS ON THE ADAPTION OF BIOLOGIC FORMS TO SEMI- 

 CONGENIAL HOSTS 

 In order to determine whether or not biologic forms are capable of 

 adapting themselves to semicongenial hosts through constant associa- 

 tion, many attempts were made to increase the virulence of the rusts on 

 these hosts by means of continuous, successive transfers. The biologic 

 forms of P. graminis tritici, P. graminis avenae, P. graminis phleipra- 

 tensis, and P. graminis agrostis were studied and the results are given in 

 Tables III to VI, inclusive. 



