458 



Journal of Agricultural Research 



Vol. X, No. 9 



Table XXIV. — Results of inoculations with urediniospores from Phleum pratense L.- 



Continued 



The work with timothy rust had two main objects: (i) To determine 

 whether Pv^cinia graminis phleipratensis varied greatly in different re- 

 gions, and (2) to ascertain whether the rust could be easily changed. The 

 reason for the investigation was the fact that, whereas previous investi- 

 gators were unable to infect barley with the rust, Stakman and Jensen 

 (26) found no particular difficulty in doing so, thus suggesting either 

 different strains of the rust in nature, or an ability of the rust to change 

 rapidly. The attempts to increase the virulence of the rust will be dis- 

 cussed in a subsequent paper; the results of the direct inoculations are 

 given in Table XXIV. There was little if any real difference between 

 strains. The rust failed to infect barley in only a few cases. These fail- 

 ures, however, are not surprising, since conditions varied considerably in 

 the different trials. Variable results are often obtained when the same 

 strain of rust is used. It will be noticed that successful infection did not 

 occur to any extent in No. 14, 15, 16, and 17. The reason for this is very 

 probably the fact that the weather was extremely hot during this period. 

 The subsequent inoculations, with one exception, were successful. The 

 percentages of successful infection were especially high in late September 

 and early October, a period during which weather conditions were ideal 

 for rust development in the greenhouse. All rusts developed unusually 

 well during this period, while during the extremely hot weather in July 

 and early August none developed very well. 



