REED: SPECIALIZATION OF PARASITIC FUNGI 393 



It has repeatedly been suggested by many investigators that special- 

 ized races of parasitic fungi may extend their normal host range by 

 passing through certain so-called "bridging hosts." There are many 

 cases known where a particular host plant can be infected b}- two or 

 more races of a parasite. Such hosts may enable a specialized race 

 to infect a wider range of plants. 



Ward (172) apparently was the first to emphasize this possibility 

 as a result of his studies of the behavior of Puccinia dispersa on various 

 bromes. Ward supposed that Bromiis arduennensis , which is readily 

 infected with uredospores of Puccinia dispersa from Bromus mollis 

 of the section Serrafalcus, as well as by uredospores from B. arduen- 

 nensis of the section Liber tia, served as a means for the rust on bromes 

 of the section Serrafalcus to pass over on to bromes of the section 

 Libertia. The following data indicate Ward's results as bearing on 

 this point. 



Uredospores Uredospores 



from Bromus from Bromus 



Section Libertia: arduennensis mollis 



Brcmtis arduennensis 8/7'- 13/14 



Bromus arduennensis var. villosus 10/10 1/14 



Section Serrafalcus: 



Bromus mollis I/8 1 19/154 



Bromus secalinus 8/8 31/61 



Stenobromjis: 



Bromus maximus 0/6 1/74 



Bromus sterilis 0/8 4/148 



An examination of the above data leads one to conclude that the 

 rust on the two hosts, Bromus arduennensis and B. mollis, are practi- 

 cally identical in their capacity for infecting other bromes. The 

 rust on Bromus arduennensis does not have any wider host range than 

 the rust on B. mollis. Both grasses appear rather to be hosts for the 

 same strain of rust. 



Ward's evidence that Bromus Krausei and B. pendulinus may serve 

 as bridging hosts is perhaps stronger. The following data indicate 

 the relation of these bromes to the rust on Bromus sterilis and B. mollis. 



Uredospores Uredospores 



from Bromus from Bromus 



Serrafalcus: sterilis mollis 



Bromus Krausei 14/29^^ 27/27 



Bromus pendulitius 17/65 50/50 



Bromus molliformis 1/25 2/26 



Bromus mollis i/i37 II9/I54 



Bromus vestilus 1/4 3/4 



Slenobromus: 



Bromus sterilis 126/146 4/148 



Bromus gussoni 37/6o 6/53 



'-The denominator of the fraction indicates the number of leaves inoculated 

 and the numerator the number infected. 



'^ The denominator of the fraction indicates the number of leaves inoculated 

 and the numerator the number infected. 



