2<M 



Arthur: Nineteen Years of Culture Work 18 



could possess urediniospores partly with one basal pore and partly 

 with three or four equatorial pores. The answer involved the 

 value and application of pore characters in defining species. 

 After special search, 8 which led to both kinds of urediniospores 

 being found repeatedly in the same sorus, although for the most 

 part they occurred in separate sori, it was concluded that only 

 one species of rust was under consideration, but with morpho- 

 logical as well as physiological races, not well delimited. 



It seemed probable, furthermore, that the previously described, 

 one-pored form of Carex rust, known as Uromyces uniporultts . 

 Kern, was a race also belonging to the Carex-Ribes species, but 

 it was not possible to test the matter by cultures. In this con- 

 nection it is interesting to note, and provocative of speculation, 

 that there is no form yet known with three- and four-pored 

 urediniospores belonging under Uromyces in the Carex-Ribes 

 aggregation, to make the parallelism with its Puccinia form 

 complete. 



In 191 7, the last year of the culture series, the principle of 

 basing species upon morphological characters, with a greater or 

 less degree of mobility in interpretation, was further illustrated 

 by the case of the Spartina rust, Uromyces Polemonii (Peck) 

 Barth., which it was found could be segregated into four races, 9 

 separable by small but appreciable differences in morphological 

 characters of both aeciospores and teliospores, and by wholly 

 unrelated aecial hosts, and further reinforced by some differences 

 in habitat and geographical distribution. The correlated Puc- 

 cinia-form for this common and widely distributed American rust 

 is that of Puccinia Distichlidis, so-called because the type collec- 

 tion was incorrectly labelled as on Distichlis instead of on Spar- 

 tina. Its range and aecial hosts, so far as known correspond to 

 only one of the four Uromyces races. 



At the time the culture work began the subepidermal rusts 

 occurring on wild grasses in America with few exceptions, passed 

 under the name of Puccinia rubigo-rcra, along with part of the 

 similar leaf rusts of cereals. No criteria had been found for 

 distinguishing them, not even those which had received special 



8 Mycologia 7: 67-69. 1915. 

 Mycologia 9: 309-312. 19 17. 



