Oct. 28, 1918 Plasticity of Biologic Forms of Puccinia graminis 235 



This strain of P. graminis secalis was obtained (diagram 6) from 

 Agropyron repens at St. Paul, Minn., on October 15, 1914, and was kept 

 until October 15, 191 7. It was confined to intermediary hosts for three 

 years, during which time about 60 different sets of inoculations were made 

 on wheat. About i ,800 leaves were inoculated, some of which are not re- 

 corded in diagram 6, and only six became infected. There is strong reason 

 to suspect that these were accidentally infected with the tritici form, all 

 occurring during the earlier period of work. Barley, Elymus rohustus, 

 E. canadensis, E. virginicus, Agropyron tenerum, and Hordeum juhatum 

 were all used as intermediary or bridging hosts, but none of them enabled 

 the rust to transfer to wheat, which remained practically immune 

 throughout the entire period. Flecks very seldom developed. Since 

 these hosts, which ought to act as bridging hosts, if any hosts act in this 

 manner, did not enable the rust to transfer to wheat after three years of 

 continuous culture, it seems quite improbable that such a change would 

 take place quickly in nature. 



In Table II a number of miscellaneous experiments are included. 

 The most conclusive is No. i, in which the secalis strain was kept 

 continuously on barley for 16 successive urediniospore generations, 

 covering a period of eight months. During this time 1 1 sets of inocula- 

 tions were made on wheat, but none of the 264 leaves inoculated became 

 infected, except one, which was very clearly shown to be accidental. 

 In all of the other trials with the other strains of rust no infection what- 

 ever resulted on the wheat. It is quite clear, then, that under the con- 

 ditions of these experiments neither barley nor Elymus rohustus enabled 

 the rust to bridge over to the normally immune wheat. This result could 

 hardly be expected, because if such changes did take place in nature the bi- 

 ologic forms isolated from these hosts could not be as uniform as they are. 



The last two experiments (No. 6 and 7), on the other hand, show the 

 inability of barley to serve as a bridging host for the tritici strain, or to 

 induce this rust to infect rye more vigorously than it ordinarily does. 



Table II. — Results of attempts to transfer Puccinia graminis secalis to wheat and 

 P. graminis tritici to rye by using barley and Elyinus robustus as intermediary hosts 



»■ One leaf accidentally infected with P. graminis tritici. 



6 Four flecks in 2 out of 4 trials. 



