556 
G. R. BISBY 
The material is so scanty and the uncertainty so great, however, 
that one can only direct attention to this point. 
As is to be expected, several genera of the Onagraceae are not 
known to be attacked by rusts. No other species of rust, than those 
herein listed, appear at present to be known to occur upon the Ona- 
graceae. 
As is stated in the beginning of this paper, the Sydows list 27 
species of rust upon the Onagraceae, and 4 additional species occur in 
North America. Furthermore, the Sydows have not yet published 
the species Aecidium Jussiaeae, Aecidium Circaeae, and Uredo oeno- 
thericola. These 34 species are accounted for in this article under 17 
titles. The arguments for this halving of the number of species are 
presented wherever a union is involved. Furthermore, a few other 
species are suggested as being of doubtful validity, notably numbers 
16 and 17. It is suggested that the two races exist under the title 
Pucciniastrum pustulatum as herein treated. 
Cultural data is necessarily of importance in limiting species and 
races; such data is at hand for but few of these rusts. The writer 
submits the foregoing arrangement of species, based upon a consider- 
ation of morphological characters and life histories, as well as hosts, 
distribution, and such limited cultural data as is at hand, in the hope 
that a workable arrangement may be presented. Finality of place- 
ment, is, of course, at the present time impossible. 
The Onagraceae rusts, as far as the evidence in hand can be analyzed 
by the writer, appear, as partially indicated heretofore, to fall into 
three fairly definite groups, with an uncertain fourth group. (See 
the diagram.) The first is that of the Pucciniastrums, with the 
alternate stage, insofar as cultures have been successful., upon Abies. 
The morphological characters of the uredinial and telial stages upon 
Onagraceous hosts, are very similar within the different species of 
Pucciniastrum. They, however, can scarcely be construed to hint 
at any relation with the uredinial or telial stages of others of these 
rusts upon the same or similar hosts. 
Following the suggestion of DieteP^ that the Uredinales have 
developed during geologic times with their hosts, the Pucciniastrums 
would be the oldest of these rusts, since their aecial stage, so far as 
Dietel, P., Centr. Bakt., etc. 12'-: 218-234. 1904, and Hedwigia 48: 1 18-125. 
1908. 
