LIFE HISTORIES OF RUSTS. 25 



Diagram 20. — Summary of inoculation experiments tvith leaf rv^t from oats. 



O 



Wtq (several leaves fleckefl). 



R 7.-, (several leaves strongly flecked), 

 s 7 



47 ^ 7 



44 



«44 



Althoiigli liii^lily specialized, the leaf rust of oats can be transferred 

 to barley, but it did not transfer to either wheat or rye. The effect 

 of barley on it was not determined, except to show that from barley 

 to oats the rust in a few trials transferred as easily as from oats to 

 oats. In Carleton's experiments (30, ]). 46) inoculation witli this 

 rust on barley gave negative results. 



In many of the experiments on biologic forms previously cited, it 

 was noticeable that the same rust species would not give the same 

 percentage of infection on various hosts if the inoculations were made 

 from rusts gathered in different locahties. This may account for the 

 diverse results of different investigators and leads to the belief that 

 there may be a large number of strains of a rust species, none of which 

 will act exactly like another toward the same hosts. Undoubtedly by 

 variation and ada])tation to varjang conditions a certain rust species, 

 widely distributed, may form a large number of strains or types 

 which, when this process has been continued for a considerable time, 

 differ widely in their physiological reactions. These may become 

 the physiological or "biologic" species. 



EFFECT OF CHANGE OF HOST ON THE MORPHOLOGY OF THE UREDOSPORE. 



In experimental cultivation of Puccinia graminis tritici from wheat 

 on barley and Puccinia graminis hordei from barley on wheat it was 

 found that there existed a slight morphological difference between 

 the uredospores of the stem rust of barley and the stem rust of 

 wheat. Upon closer examination this difference seemed to be meas- 

 urable — that is, the uredospores of barley measured (on a basis of 

 measurement of 50 spores, widely selected) considerably shorter and 

 very slightly narrower than those on wheat. An experiment was 

 therefore inaugurated to determine what the effect would be on the 

 size of the spores of the barley rust when gro^\^l on wheat and of the 

 wheat rust when grown on barley. Transfers were accordingly made 

 of the barley rust to two pots of wheat and of the wheat rust to two 

 pots of barley. The barley rust on wheat was then transferred to 

 wheat plants continuously for about a year, and the wheat rust in a 

 similar manner was grown on barley. The rust in each pair of pots 

 was transferretl to two other pots, so that two separate strains were 



216 



