g8 BOTANICAL GAZETTE [august 



eU' form is the highest in complexity and a late stage in the develop- 

 ment of the rusts. The peridium is undoubtedly a late modification 

 for protection, and cuplike aecidia are very likely less primitive than 

 the cacomas. This would argue that a rust of the Puccinia graminis 

 type is in structure very far removed from the primitive rust, a view 

 which is quite opposed to the stand taken by Blackman. 



I entirely agree with Dietel (12) that the presence or absence 

 of the secondary uredospores in the life-cycle is of secondary impor- 

 tance w^hcn considering the history of the rusts. The essential stages 

 are those of cell fusion and chromatin fusion and reduction. The 

 uredospores simply prolong the sporophytic phase as the conditions 

 demand, and it is likely that the -opsis forms may result from the 



+ 



elimination of the uredo stage in some rusts; while in others they 

 represent an earlier less complete stage in evolution than the eu- forms. 



The curious condition found in Endophyllum may well have 

 arisen by the elimination, for some cause, of the teleuto and uredo 

 stages. The cuplike structure and short intercalary cells all point to 

 its once having been a very highly specialized form from the stand- 

 point of evolution. It is quite conceivable that as a result of some 

 change in condition the aecidiospores should have developed the 

 capacity to reinfect the aecidium host. 



The facts of alternation of generations in the rusts are in harmony 

 with Dietel's view that the micro- and lepto- forms are the most 

 primitive types of rusts. Besides the arguments advanced above, 

 we have in the diverse structures of the teleutospores of the different 

 genera a further support for this view. Assuming the teleutospores to 

 be the oldest spore form, it is easy to see how they might have become 

 more markedly modified than the newer uredo- and aecidiospores. 

 The loss of function of the spermatia as gametophytic conidia, too, 

 is exactly what might be expected with the development of the sporo- 

 phytic spores. 



The writer wishes to acknowledge his indebtedness to Professor 

 R. A. Harper for his many helpful suggestions and other assistance, 

 and also to Dr. E. W. Olive for permitting him to use the data con- 

 cerning Puccinia transformans in advance of its publication by him. 



Madison, Wisconsin 



