36 FARLOW ON THE GYMNOSPOEANGIA 



sporidia of Q. glohosum had become mixed with those of the two other species. Such 

 a supposition is possible in the case of G. macrojms which often grows in company with 

 G. glohosum, but it can hardly be true of the G. biseptatum in question, which grew in 

 a deep swamp remote from G. glohosum, and the specimens of which were collected and 

 covered with care to prevent a mixing of the spores with those of other species. 



Whether we consider the distribution of our species or the results of the cultures made, 

 there is nothing to confirm the views of Oersted as to the connection of particular species. 

 In this connection, I would refer to a paper by Rathay known to me only by the abstract 

 given by Magnus in Bot. Zeit., 1880, p. 798. The method of culture adopted by Rathay is 

 unknown to me, but he came to the conclusion that R. penicillata belonged not as a form 

 of R. lacerata to G. davariaeforme, but to G. fuscnm. If then our G. fuscum var. gloho- 

 sum. be really a variety of G. fuscum., and if R. penicillata be a form of R. cancellata as 

 supposed by Rathay, then the spermogonia on C. tomentosa, which so frequently followed 

 the sowing of the sporidia of G. glohosum,, might be supposed to belong to what I have 

 called R. penicillata, which does occur on Crataegus in the United States. One could not 

 be at all certain, however, without seeing the fully developed aecidia, but it must not be 

 forgotten that those who are fully imbued with the belief that the different aecidial genera 

 as Aecidium, Roestelia, etc., are stages of Puccinia, Gymnosporangium, etc., accept the ap- 

 pearance of spermogonia alone, without having seen the aecidia, as strong proof of a con- 

 nection between different forms. In fact the instances where the aecidia themselves have 

 been produced by cultures of teleutospore forms are very few in number. But even if we 

 admit that the spermogonia following the sowing of G. fuscum belonged to R. penicillata, 

 what are we to say of those which followed the sowing of G. m,acropus and G. hisepta- 

 tum 1 It is absolutely impossible to consider G. hiseptatum a form of G. fuscum, nor, in 

 my opinion, is there any reason to suppose that G. macrcpus is a form of that species. 



Sp ermogonia followed sowings of G. macropus on both C. tomentosa and Amelanchier, 

 and accordingly they might have belonged to R. lacerata or R. aurantiaca. R. cornuta 

 may be excluded as belonging, according to Oersted, to G. conicum, which is not in the 

 least related to G. macropus, and the distribution of R. hyalina makes it very improbable 

 that it is connected with the ubiquitous G. macropus. R. lacerata should be connected 

 with G. davariaeforme and, as has already been remarked, Schroeter has suspected that 

 G. macropus may be a form of the last named species, but I have already stated my rea- 

 sons on structural grounds for not considering them two forms of the same species, and I 

 do not think that that belief should be altered in consequence of the results of my cultures. 

 There remains then R. aurantiaca which might possibly be connected with G. macrojms. 



The case of G. biseptatum is still more desperate. It certainly cannot be connected 

 with R. penicillata, or R. lacerata, and if we assume it probable or even possible that 

 there is a connection between G. macropus and R. aurantiaca, there is only left R. cor- 

 nuta to be matched with G. hiseptatum, and this would imply that G. conicum and 

 G. hiseptatum were forms of the same species, which I presume that few botanists are 

 willing to admit, for excellent anatomical reasons. 



The reader has probably in the last few pages been surfeited with ifs and or's, and a 

 choice of rather bewildering alternatives. There is only one more point to be suggested 

 in this connection. That is, that the appearance of the spermogonia after sowhig the 



