389 



All rusts at present known fall into one of these two classes : the 

 sporophytic generation gives rise either to a single spore-form, or else to 

 initial and final spore-forms, with usually intermediate repeating forms. 

 Whether one or more than one spore-form arises between the cell fusion 

 and final nuclear fusion, constituting the sexual period, all such spores, of 

 whatever morphological structure, are of a sexual nature, the initial form 

 (whether of the secidiuin-type, cseorua-type, primary uredo-type, or when 

 none of these is produced, the teleuto-type) being the one which most 

 clearly shows the stimulus of fertilization. 



The above facts, especially when taken in connection with the highly 

 differentiated structures associated with the initial and repeating spores, 

 often being quite equal or superior to those of the teliospores, show every 

 reason that may be based upon morphology and development for consider- 

 ing the initial and repeating spores as practically of equal taxonomic 

 rank with the teliospores. To illustrate, a genus founded upon a repeat- 

 ing stage, like the genus of imperfectly known fern rusts, Milesia, should 

 be as valid as if founded on the telia. This genus has recently been re- 

 christened Milesma on the ground that the original name, given in 1870, 

 is invalid because it was only applied to the uredinia and not to the telia. 

 Again, now illustrating with a specific name, the heteroecious rust which 

 was first specifically called poculiforme was described in its secial stage 

 under /Ecidium, and according to the preceding argument on the impor- 

 tance of the initial spores, this name having priority, although not at the 

 time made to include the telia, should be used, whatever genus name be 

 considered the best, as e. g., Dicwoma poculiforme or Puccinia poculiformis, 

 not Puccinia graminis. 



From the foregoing it will be seen that for purposes of taxonomy 

 names applied to the pycnia (spermogonia) may properly be ignored, on 

 the ground that they apply to asexual or conidial structures, but that 

 names applied to reeia and uredinia (SEcidium, Ccooma, Peridermium, 

 livedo, and other such forms) should have the same standing as names 

 applied to telia (teleutospore stage). 



I have tried to show that the main features in the life cycle of all 

 rusts exhibit essential uniformity, there being two large groups, one with 

 a single form of spore (teliospore) in the sporophytic generation, and the 

 other with additional initial and (usually) repeating spores, and that the 

 great diversity lies in, the details of their structural development. It is 



