THE UREDINALES FOUND UPON THE ONAGRiVCEAE 529 
larity may be expressed in different ways. One of the earliest noted 
resemblances between species was that between certain Uromyces and 
Puccinia forms with the same life cycle, the only difference being in 
the possession, by the one species, of one-celled teliospores, by the 
other of two-celled teliospores. Fries^ noted the fact that such an 
analogy sometimes existed. Orton^ has reviewed the literature per- 
taining to this type of correlation, and shown some specific instances. 
This kind of correlation is shown in the Onagraceae rusts by the re- 
semblance of Uromyces plumharius to Puccinia Epilobii-tetragoni. 
A second type of correlation is that illustrated by a similarity 
in the characters of a short-cycled species and a long-cycled heteroe- 
cious species bearing aecia upon the telial host of the short-cycled rust. 
Dietel,^ and about the same time, Fischer,^ pointed out this type of 
resemblance. Travelbee,^ from this laboratory, has briefly sum- 
marized the literature regarding this type of correlation, and listed 
some proven examples. This indicated relation has been successfully 
used to predict alternate hosts: indeed, in the Onagraceae rusts, through 
the similarity of the telial stage of Puccinia Epilobii to that of Puccinia- 
Veratri, the alternate host of the latter was forecasted, as is noted later 
in this article. 
A third type of correlation is that between two species of rust, 
with life cycles of different lengths, occurring upon the same or similar 
hosts. Fischer^ has indicated such a relation between two species 
of rust upon Epilobium ; other possibilities are suggested in this paper. 
It is obvious that a similarity of morphological characters does 
not of necessity reflect a phylogenetic relationship. Certainly, how- 
ever, when such similarities are found between rusts upon related 
hosts, it is a noteworthy suggestion of a definite relationship between 
the rusts, and it seems to the writer that relationships may be inferred 
in certain cases even when there is some slight variation between the 
parallel characters of two such species of rust. 
Races are designated in this paper as occurring within the long- 
2 Fries, E. M., Summ. Veg. Scand. i: 514. 1846. 
^ Orton, C. R., Correlation l3etween certain species of Puccinia and Uromyces. 
Mycol. 4: 194-204. 1912. 
^ Dietel, P., Uredinales, in Engler and Prantl, Pflanzenfam. 69. 1897. 
5 Fischer, Ed., Beitr. Krypt. Schweiz i^: 109. 1898. 
^ Travelbee, H. C, Correlation of certain long-cycled and short-cycled rusts. 
Proc. Ind. Acad. Sci. 1914: 231-234. 1915. 
^ Fischer, Ed., Beitr. Krypt. Schweiz 2^: 154-155. 1904. 
