Arthur and Kern : Peridermium 



407 



Ke>- to the assignment of species of Peridermium 



Pycnia subcuticular. 

 Aecia cylindrical. 

 Aecia tongue-shaped. 

 Pycnia subepidermal. 



Aecial peridia one cell thick. 

 On Finns. 

 On other Coniferae. 

 Pycnia subcorticular. 



Aecial peridia more than one cell thick. 



Puccin iastru m , Calyptospora . 

 Melampsorella, M e lamp sori din vi . 



Coleosporium . 

 " Chrysomyxa^ 



Cronartium. 



Our guide for selecting the proper telial genus has been the 

 kind of aecia found to belong to certain species of those genera 

 in the culture work accomplished mainly by European botanists, 

 although the characters given as dominant are of our own selection. 



After this explanation of the method of deriving the data of 

 the table, we may turn to its statistics to see why in some cases 

 more telial species are known than corresponding aecial forms, 

 and why in two cases the reverse should be true. It must be 

 borne in mind that the whole tabular presentation rests upon a 

 large basis of assumption, in which there is plenty of opportunity 

 for error after every precaution is taken. The rust flora of North 

 America is, moreover, very imperfectly known, and especially the 

 part under discussion. Turning to the table we find that in the 

 first line twenty- five species of Coleosporium are recognized and 

 only seven corresponding species of Peridermium, which is easily 

 enough accounted for by the fact that the aecia of this genus are 

 remarkably similar. Fischer * in his recent flora of Switzerland 

 describes with much fullness ten species of Coleosporium, but gives 

 almost no characters to separate the aecia, saying that these " can 

 not be determined with certainty without infection experiments." 

 It is quite likely that when cultures are made, the seven species 

 as now understood will be segregated into two or three times that 

 number, and in the meantime others may be found. In the second 

 line of the table the situation is quite different ; more aecia are 

 known than telial forms. Here the aecia are conspicuous, and 

 the telia comparatively inconspicuous. Probably more species of 

 Cronartium are yet to be discovered, or else closely related genera 

 not yet recognized. In the fourth line of the table the two re- 



* Fischer, En. Die Uredineen der Schweiz. Beitr. Krypt. Schweiz 2: 439-455 



1904. 



