﻿538 Dodge: Effect of host on Gymnosporangium 



of spores with more than two cells in the sori of G. fraternum, and 

 the overwhelming predominance of three- to six-celled spores in 

 G. biseptatum; (2) leaves of certain marked cedar trees at New- 

 field, to my knowledge, have borne sori of G.jraternum in the spring 

 season of three years, the first being in 191 1, with no indication of 

 there being in the process of formation the fusiform swellings 

 characteristic of G. biseptatum on any of the branches in 

 1913-1915; (3) it is possible to infect both Aronia and A melan- 

 chier with G. fraternum, while G. biseptatum regularly infects A melan- 

 chier only, so far as known at present; (4) we may have here a 

 case similar to that described by Long (11). The two rusts, 

 Puccinia EUisiana and P. Andropogonis have the common teleuto- 

 spore host Andropogon. The aecidial stage of the former is found 

 on Viola, that of the latter on Pentstemon. Starting with aecidio- 

 spores on the violet and going back through the Andropogon he is 

 able to infect Pentstemon very easily; but when he begins with P. 

 Andropogonis on Pentstemon he finds that it is much more difficult 

 to infect the violet. He believes that P. EUisiana is continually 

 passing over from the violet through Andropogon to Pentstemon 

 by natural infections, in which case P. Andropogonis is being created 

 over and over again. Once on the Pentstemon the rust is probably 

 unable to get back in nature to the violet and therefore becomes 

 fixed as a new form of the old species, P. EUisiana. In addidon 

 to the marked changes in the aecidium which I have previously 

 mentioned, Long also finds that the uredospore characters are 

 modified by changing the regular sequence of hosts. Gymno- 

 sporangium fraternum being able to infect both Aronia and Ame- 

 lanchier would correspond to Puccinia EUisiana; G. biseptatum, so 

 far as known being confined to the aecidial host A melanchier, is 

 the new form. In other words, the changes wrought in the mor- 

 phological characters of G. fraternum during its life on Amelan- 

 chier would be reflected as physiological changes to such an extent 

 that the fungus on going back to the Chamaecyparis becomes peren- 

 nial in the tissues of the branches where further morphological 

 changes are induced. (5) Cedars inoculated by me with aecidio- 

 spores of Roestelia transformans in July developed sori of G. frater- 

 num the following spring, while controls (not inoculated) did not. 

 Dr. E. W. Olive transplanted a small cedar from Newfield to the 



