SUMMARY. 17 



cereals. From the physiological specialization of the rust, from its 

 distinct method of uredospore infection, from the numerous negative 

 results of inoculations with tlic teleutospore on harberries, and from 

 the delicacy of the uredo mycelium, as compared with typical ^mmwi^ 

 foi-ms, it is evident that this rust has many distinctive characteristics. 

 Unless further experiments should show that the rust can produce its 

 a?cidium on the barberry and until such experiments have been per- 

 formed, the writer favors the use of the specific name Puccinia phlei- 

 pratensis Erikss. and Henn. 



At the Arlington Experimental Farm the rust mycelium lives 

 through the winter. It is very probable that it lives over winter in 

 the uredo stage much farther north than the latitude cited. As the 

 fecidial stage, if present, is undoubtedly rare in the United States the 

 teleuto stage is of doubtful importance in the wintering and dis- 

 semination of the rust. 



The rapid distribution of timothy rust in recent years is undoubt- 

 edly due to the dissemination of the uredospores by the usual agencies, 

 namely, insects, birds, animals, man, surface winds, and upper air 

 currents. Its ultimate distribution over all timothy-growing sec- 

 tions favorable to it is to be expected, and methods of preventing 

 any considerable damage to the timoth}^ crop become necessary. 



In a study of the resistance of varieties of timothy to rust in 1908 

 and 1909, W. J. Morse found that the resistance "varied greatly^ 

 ranging from zero to 98 per cent." The resistance of difFerent strains 

 relative to each other varied little during the two years. Under a less 

 severe rust attack in 1910 these strains in many instances failed to 

 retain the same relative position as in previous yeare. This tends to 

 show that dependable data can be obtained only when vigorous rust 

 attacks occur. 



In greenhouse experiments where strains which had been tested in 

 the field were subjected to similar inoculation and identical condi- 

 tions during the period of incubation, the percentage of inoculated 

 leaves wliich did not become infected did not correspond \\dth the 

 figures on rust resistance obtained in the field. When the severity 

 of infection and not the percentage of leaves developing pustules was 

 considered, however, the relative resistance of strains as obtained in 

 the field was fairly well maintained in the greenhouse. 



As there are wide differences in timothy strains with regard to rust 

 resistance, the problem in timothy-rust prevention becomes one of 

 breeding. Tliis may not be as difficult as it appeai-s at first. Such 

 work, however, to be of value must be carried on when the rust is 

 abundant and where either naturally or artificiall}^ a rust attack 

 occurs every year. 



224 * 



