Arthur: Uredinales on Carex in North America 243 
As this paper is intended to deal chiefly with the Carex hosts, 
nothing has been said about aecial hosts of the 24 species involved, 
but incidentally it may be stated that aecia belonging to il of 
these species are known, and are represented by 676 collections 
at the present time, for which it is hoped to secure ecpial accuracy 
in determinations by enlisting the aid of specialists. 
So much as to the second test used in the rust work for the 
North American Flora, the one regarding hosts, and we now 
turn to the third test, the one relating to microscopic characters 
of the fungus. It being difficult to carry in mind the minute 
appearance of the spores for purposes of critical comparison, as 
many of the collections as possible have been illustrated by 
camera lucida drawings made with a uniform amplification of 
625 diameters. There have been 693 of the Dicaeoma collec- 
tions on Carex illustrated in this way, giving >1029 individual 
teliospores, 937 urediniospores and 69 amphispores ; while of their 
aecial connections 174 collections have been similarly illustrated, 
showing 135 individual aeciospores and 86 peridial cells, with 17 
sections of pycnia more or less completely drawn. 
Many of the collections have had sets of data taken regarding 
length and breadth of a number of spores, usually six to eight, 
thickness of apex, side wall and pedicel, the color, sculpturing of 
surface, and number and position of pores. With the Carex 
collections for Dicaeoma are 355 such sets of data, and with the 
associated aecial collections 90 similar sets of data, extended to 
include peridial cells and pycnia. 
Regarding the fourth test applied to each collection, that of the 
limital characters for the species, nothing need be given in this 
connection, as the purpose of this article is especially to call 
attention to the material being used for study. It may be inter- 
jected, however, that the term species in the manuscript for the 
final work is generally made to include such varieties, races or 
physiological species as many authors prefer to distinguish by 
separate names. 
The presentation of the above statistics will illustrate some 
features of the method employed in preparing the manuscript 
for the rust part of the North American Flora, as other codec- 
