OF TlfK UNITED STATES. 



33 



From the account already given, it will be seen that I recognize the following species, 

 and the question which I have unsuccessfully tried to answer is, what species with us are 

 genetically connected. To sum up the species, we have: 



G YMNOSPOKANGIUM. ROESTELIA. 



G. Ellisii. R. botryapiles. 



G. clavariacforme. R. transformans. 



G. macropus. R. ccmcellata. 



(r. J)/scH))i. R. cornuta. 



G. fuscum var. globosum. R. lacerta. 



G. biseptatum. R. penicillata. 



G. clavipes. R. hyalina. 



G. conicum f R. aurantiaca. 



It will be seen that I have mentioned eight species of each genus which could on 

 anatomical grounds alone be considered distinct. Of the species enumerated, G. conicum 

 is given as doubtful, because it seemed to me that sufficiently abundant material was want- 

 ing to enable any one to speak with certainty. G. fuscum var. globosum, it will also be 

 borne in mind, is not by several writers considered distinct. With the possible exception 

 of the two species just mentioned, the validity of the species of Gymnosporangium is not 

 likely to be much questioned. Turning to the Roesteliae, we have R. lacerata and R. 

 penicillata enumerated, which are by many writers united on anatomical grounds, apart 

 from any developmental considerations, and R. cancellata, a species whose presence, or, 

 perhaps better, whose distribution in the United States is not sufficiently well known in 

 my opinion. In speaking of R. lacerata, also, one must not forget that, even in the lim- 

 ited sense in which I have adopted it, it appears under a good many different forms which 

 some botanists on anatomical grounds alone might consider distinct. 



If one is disposed to admit the eight species of each genus with the limitations I have 

 given, he might suppose that the task of tracing the connection between them would be 

 comparatively simple and interesting. A very slight experience, however, would con- 

 vince him of the contrary. In the first place, if we accept the conclusions of Oersted as 

 correct with regard to the Danish species, knowing that two and perhaps all three of the 

 Danish Gymnosporangia are found in the United States, we are struck with the fact that, 

 although G. fuscum, regarding var. globosum as distinct, and the true R. cancellata, its 

 supposed aecidial form, are about equally common, or rather equally rare, with us, when we 

 come to G. clavariaeforme, the case is different, for the Gymnosporangium is not at all 

 common, while its presumed aecidium, R. lacerata, is very common, indeed being found 

 hundreds and even a thousand miles from localities where G. clavariaeforme is known. 

 About the comparative distribution of G. conicum, and its corresponding R. cornuta, 

 little can at present fee said, since the localities of G. conicum are not well known. Yet, 

 in general, what is supposed to be G. conicum is best known towards the South, while 

 R. cornuta is northern in its range, unless, indeed, that species properly includes some of 

 the forms now included in R. lacerata. If, on the other hand, with some writers we 

 regard the var. globosum as identical with G. fuscum of Europe we are, in looking at the 

 distribution, met with the difficulty that G. fuscum var. globosum is very common in Mas- 

 sachusetts, for instance, whereas its supposed aecidium, R. cancellata, is not known with 



