50 • Journal of Agricultural Research voi.xvi.no.z 



the change is a gradual one, it is reasonable to hope that some evidence 

 of this change may be obtained by the methods used in the present work. 

 Tv/o lines of work were pursued: (i) Attempts were made to develop 

 a number of morphological strains of a given biologic form by culturing 

 it for fairly long periods of time on several different hosts, and 

 (2) attempts were made to unify spore sizes of different biologic forms 

 by growing them on the same hosts. For instance, P. graminis tritici 

 develops well on common wheat, barley, and on various species of 

 Agropyron, Hordeum, and Elymus. The writers tried to ascertain 

 whether these hosts exerted an appreciable effect on the rust when it had 

 been confined to them for considerable periods of time. Again the 

 secalis, tritici, and triiici-compacti forms grow about equally well on 

 barley. Theoretically, therefore, it could be assumed that they ought 

 to become morphologically similar if grown on barley long enough. In 

 fact, all of the biologic forms discussed in this paper develop at least 

 weakly on barley. It could be assumed that if they could all be grown 

 on barley long enough, they would eventually become similar morpho- 

 logically. The results of the effect of hosts are given on Tables IV to XII. 



KEY TO TABLES IV TO XII 



In Tables IV to XII the host from which the rust was originally cultured is given 

 in the second column. Intermediate hosts refer to the hosts on which the rust had 

 been grown up to the time the plant was inoculated. The term "intermediate host" 

 is not used here in the sense of bridging. W, O, B, and R refer to wheat, oats, barley, 

 and rye, respectively. Other symbols are explained when used. The number of 

 " urediniospore generations" (successive transfers) on the host is indicated by the 

 figure immediately following the symbol for that host. Thus, R2B4R1W2B5 indicates 

 that the rust was transferred to rye twice, then to barley four times, followed by one 

 transfer to rye, two to wheat, and five to barley. The degree of infection is self- 

 explanatory. The result of inoculation is given in the usual manner in the form 

 of a fraction, the denominator showing the number of plants inoculated and the 

 numerator the number which became infected. 



Attempts to Develop Morphologic Strains of Biologic Forms by Culturing 



ON Different Hosts 



To determine whether or not a given biologic form of P. graminis has 

 a tendency to split up into a number of different morphological strains 

 on account of confinement for fairly long periods of time to a number of 

 different hosts, a series of experiments was conducted with the six 

 biologic forms indicated above. The host plants employed were very 

 frequently of distant taxonomic relationship, but, unless they were 

 equally congenial to the parasitic attack of the fungus, their effect was 

 not considered when the final conclusions were drawn. The result of 

 this phase of the work, which extended over a period of two years, is 

 given in Tables IV to IX. 



