(408) 



also be related to a Gymno sporangium on some 1 

 to the Juniperaceae. The writer has explained in some detail 

 in another paper (Kern, 1910 1 ) that there is within the range of 

 Aecidium Sorbi no known Gymno sporangium to which it may belong, 

 but that there is in very close association with it a cedar-rust in 

 the form of Uredo nootkatensis on Chamaecyparis nootkatensis 

 and that there is strong probability of a relationship between 

 these two forms. This hypothesis is supported by inferences 

 drawn from analogy, homology, and certain features in their 

 geographic distribution. Uredo nootkatensis is an undoubted 

 uredinial stage and from its host affinities its connection with a 

 Gymnosporangium-like telial form is naturally to be anticipated. 

 In this way we come to the belief that a uredinial stage does 

 occur within this group and that there is the strongest kind of 

 evidence that it occurs in a species which has the cupulate-type 

 of aecium, or, in other words, in a species which must be looked 

 upon as primitive. There is also evidence indicative of the pos- 

 session of other primitive characters. Since G. Libocedri, whose 

 aecial stage is of the cupulate-type, has a small foliicolous telial 

 stage, not causing any hypertrophy, we may by analogy predict 

 a similar telial stage for the Uredo nootkatensis- Aecidium Sorbi 

 combination. In considering the telial characters, it should be 

 said, there is every reason to regard the small foliicolous forms which 

 cause no hypertrophy as the most primitive, and the forms which 

 occur on the branches with fusiform swellings and those which 

 induce special gall-like outgrowths, as relatively more and more 



This very probable appearam 

 species which possesses other chai 

 stages of an undoubted primitive nature, the writer interprets 

 as a support for the view that the presence of the uredinial stage 

 is in itself a primitive character. According to this theory, the 

 possession of the cupulate-type of aecium, foliicolous telia, and 

 presence of uredinia are characters of the most primitive species 

 of Gymnosporangium. The suppression of the uredinial stage 

 and the assumption of some of its functions by the aecial and telial 

 stages appear to have been the first step in reduction, while the 

 reduction in the hosts, or autoecism, has taken place much later. 

 There is but one autoecious species known, G. bermudianum, and 

 according to the foregoing view it would stand as the most highly 



