Arthur: Nineteen Years of Culture Work 21 



other species of rust." Additional genera in the same families 

 were added from time to time for aecial hosts, until in the cul- 

 tures of 1916 the species was grown on Abronia and Polygonum, 

 thus adding two more families. Mr. E. Bethel, of Denver, Colo- 

 rado, who made the field observations and suggestions for the 

 later additions, has continued the list since the culture series 

 stopped and brought the number up to 76 species, belonging to 

 19 families, 12 a truly astonishing showing, and all the more so as 

 no clearly defined races have so far been detected. The only 

 other species of rust with such a remarkably extended series of 

 aecial hosts at all approaching P. subnitens Diet., is that of P. 

 Isiacae (Thiim.) Wint. from the dry trans-Caspian region of 

 western Asia, as reported by Tranzschel. 13 This species with 

 telia on Phragmites communis has aecia on 19 species of hosts 

 belonging to 9 families, the aecial families being the same as for 

 P. subnitens. 



In still another way the conception of species was modified 

 when in 1905 teliospores from Ruellia ciliosa were grown on the 

 same host and also on R. strepens. The latter host, with loose, 

 watery tissues, gave rise to aecia fully ten per cent, larger in every 

 way than did the former host with its firm, woody tissues, thus 

 showing that the forms recognized by the Sydows under Puccinia 

 lateripes B. & Rav. and P. Ruelliae (B. & Br.) Lagerh. 14 repre- 

 sent only a host influence upon one and the same species, this 

 influence being traced not only in the aecia, but also in the other 

 spore-forms. 



Thus it will be seen that while the main work of the cultures 

 was effective in completing the life cycles for many species, and 

 in some cases extending and defining the range of hosts, it was 

 at the same time most profoundly modifying the current concep- 

 tion of species among the rusts. Instead of a rigid ideal of a 

 few invariable characters and a limited range of nearly related 

 hosts to be determined by cultures, we have substituted a com- 

 plex of somewhat variable morphological characters as the basis, 



12 Bethel, Phytopathology 9: 193. 1919. 



13 Beitrage zur Biologie der Uredineen. Trav. Mus. Bot. Acad. Sci. St. 

 Petersb. 3: 40. 1906; 7: 14. 1909. 



14 Sydow, Monographia Uredinearum 1: 235. 1902. 



