﻿The effect of the host on the morphology of certain species 

 of Gymnosporangium 



The large number of rather closely related species of Gymno- 

 sporangium which have been described as infecting such host 

 plants as the red cedar, Chamaecyparis , the juniper, etc., on the 

 one hand, and the hawthorns, apples, shad-bush, etc., on the 

 other, raises the question as to the relation of the particular host 

 to the specific differences in the parasite. 



It is commonly accepted (17, 18) that certain species of 

 Gymnosporangium gain entrance into the tissues of the coniferous 

 host through leaf infection. In the case of G. macropus on the 

 red cedar, strictly foliicolous galls are often quite numerous. To 

 what extent species of Gymnosporangium are perennial and 

 whether in such cases sori may develop on the leaves as well as 

 on the branches or trunk does not appear to have been definitely 

 determined in many species. Reed and Crabill (18) claim that 

 the mycelium of G. macropus never penetrates beyond the base 

 of the leaf into the tvvig. The leaf forms which Farlow (3) 

 referred to G. clavipes were later found to belong to G. nidus-avis, 

 now believed to develop on all parts of the red cedar. When the 

 trunk or branches are attacked the bark is split open or deeply 

 furrowed. The same species may produce the witch 's-broom or 

 bird's-nest malformations, and isolated sori may even be found 

 at the base of the leaves or possibly on the leaves themselves. 

 Farlow and Thaxter, in their later publications (4, 20-22), cleared 

 up the points relating to the identity of G. nidus-avis, so that one 

 is not left in doubt as to their conceptions of the variations and 

 limitations of this "multiform species." 



I am studying a somewhat similar case in connection with G. 

 biseptatum, which is perennial in the branches or trunk of Chamae- 

 cyparis, The teleutospores are commonly three- or four-celled, 

 519 



