﻿Dodge: Effect of host on Gymnosporangium 529 



spores in the field at the time of collection were so great that, 

 likewise, I should have attached no importance to these results 

 were it not for the suggestions by Farlow (3) and others of a 

 possible relationship between the ordinary perennial caulicolous 

 form on Chamaecyparis and the "annual" leaf form on the same 



Comparing "Gymnosporangium fraternum on Aronia" and 

 "G.fraternum on Amelanchier" we see two noteworthy differences: 

 (i) the effect on the host; (2) the characters of the aecidia 

 themselves. Roestelia transformans {"Gymnosporangium frater- 

 num on Aronia'') sometimes entirely transforms the terminal 

 buds so that a large number of aecidia are produced on a massive 

 gall; the same type of swelling may occur on young branches 

 (Figs. 7-9). The leaf may be more or less bent or deformed (Fig, 

 10). The aecidia arise in long horn-like projections of the host. 

 In a majority of cases, however, in my cultures, as noted above, 

 the Aronia leaves are not deformed and the primary or basal 

 hypertrophy is not always prominent (Fig. 12). 



Each aecidium of "G. fraternum on Amelanchier" arises from 

 a gall which is more wart-Hke or ovoid. Fig. 20 shows such galls; 

 the spot in the center of each indicates where the peridium will 

 grow out. Figs. 21-25 show further stages. The peridial cells are 

 strikingly different in the two forms. They are very long in both 

 cases. The cells from Amelanchier are irregularly bent or hypha- 

 like and smooth on both surfaces (Fig. 43), while in those from 

 Aronia the inner surface is invariably roughened with several series 

 of warts (Fig. 41). The forms of the galls from which the aecidia 

 arise are such as might be due to the reaction of different hosts 

 to the stimulus of the same fungus. 



The published infection experiments with G. hiseptatum are 

 very few. Certainly the possibility remains that Aronia can 

 be infected with sporidia from the perennial caulicolous form. 

 Arthur (la) reports the infection of Aronia with G. effusum, and 

 Kern (10) suggests that G. effusum and Roestelia transformans are 

 possibly connected. Tubeuf, Saccardo, Sorauer, Pammel, and 

 others have published more or less as a fact the connection between 

 G. Ellisii and Roestelia transformans. I have repeated Fromme's 

 (6) experiments with G. Ellisii and easily succeeded in infecting 



