244 INTRODUCTION TO THE STUDY OF FUNGI 



certain species of Rumex, that of Puccinia Trailii on Rumex 

 acetosa only, and that of Puccinia Mcignusiana on Ranunculus 

 repens and bullosa. If the Puccinia itself, which corresponds 

 to the imago stage, in being the ultimate and most perfect 

 development of a cycle — is insufficient fur the determination, 

 or discrimination, of Puccinia p)hragmitis, Trailii, and Magnus- 

 iana, tlien it may be contended that the basis of classification 

 is greatly defective for all practical purposes. Yet it is not 

 uncommon to meet with observations associated with one 

 species, so called, of Puccinia, that " it cannot be distinguished 

 morphologically" from another species, but that "its life-history 

 is different." This is the crucial test of the system, for if the 

 perfect stage of one species is not to be distinguished from 

 another, or, it may be, from two others, although the system 

 may be very philosophical, it is nevertheless impracticable. 

 Let us proceed, however, to the method set forth in Saccardo's 

 Sylloge, and now generally adopted. 



The primary divisions are again based upon the spore 

 and its septation. The Amerosporae includes all the genera in 

 which the teleutospores are unicellidar. Of these, two genera 

 have no pseudoperidium to enclose the teleutospores, namely 

 Uromyces and Hemileia. The latter is represented by the 

 destructive coffee -leaf disease, the assumed teleutospores of 

 which have half the surface smooth, and half warted. Its 

 association here is scarcely more than provisional. Uromyces 

 is the great genus of the section, and the teleutospores 

 germinate by one pore. In typical species the spermogonia, 

 aecidia, iiredospores, and teleutospores are all present on the 

 same host-plant. In others the spermogonia and aecidia are 

 unknown. In another section the spermogonia, aecidia, and 

 teleutospores are present on the same host, but the m^edospores 

 are unknown. In the final section only the teleutospores are 

 known, the spermogonia, aecidia, or uredospores never having 

 been discovered or recorded ; this is, in effect, the section which 

 includes the imperfectly known species. The two genera in 

 which the uredospores are included in a pseudoperidium are 

 Melampsora and Melampsorella, very closely related — the sori 

 of the teleutospores being crust-like, blackish, and determinate 

 in the former ; flattened, indeterminate, and pallid in the 



