364 



BROOKLYN BOTANIC GARDEN MEMOIRS 



general, he concludes that species of the same section of the genus as 

 the one serving as a source of uredospores were more fully infected 

 than the species of other sections. The evidence for Ward's con- 

 clusion is not very striking except in the case of the two hosts Bromus 

 mollis and B. sterilis. Some of Ward's data may be indicated as 

 follows : 



Host Inocalated 



Serrajalcus: 



Bromus arvensis 



Bromus brachystachys 



Bromus brizaeformis . 



Bromus Krausei 



Bromus macrostachys . 



Bromus moUiformis . . 



Bromus mollis 



Bromus pe?idulinus . . 



Bromus secalinus . . . . 



Bromus vestitiis 



Liber tia: 



Bromus arduennensis . 

 Stenobromus: 



Bromus gussoni 



Bromus madritensis . . 



Bromus maximus . . . , 



Bromus sterilis , 



Freeman (56) has made a further study of the brome rust, using 

 uredospores from Bromus mollis and B. sterilis. He states that twenty- 

 two different species of Bromus remained free from infection following 

 inoculation with rust from both hosts. Eleven species were infected 

 with uredospores from B. mollis but not with uredospores from B. 

 sterilis. Only one host, B. sterilis, was infected by uredospores from 

 B. sterilis, and not by uredospores from B. mollis. Five species were 

 infected with spores from both grasses. 



Puccinia Stipina Tranzschel. Under this name, Tranzschel (156) 

 groups the North American rust Puccinia Stipae Arth., with uredo 

 and teleuto on Stipa spartea and aecidial stage on Aster ericoides, A. 

 multiflorus, A. Novae-angliae and Solidago canadensis, and the Euro- 

 pean rust, Puccinia Stipae Bubak, with uredo and teleuto stages on 

 Stipa capillata and aecidial stage on Thymus and Salvia. Klebahn 

 (87, 91) finds evidence for two specialized races in the European rust: 



' The denominator indicates the number of leaves inoculated and the numerator 

 the number infected. 



