530 MACKENZIE: NOTES ON CAREX 
natural method of telling C. déflexa and its allies from C. umbellata 
and its allies is that there are always pistillate spikes at the base 
of the staminate in the former, and the lowermost of these has a 
leaflet-like bract exceeding the inflorescence. This bract is rarely 
auriculate, and if colored at all at base is purplish-brown tinged. 
In the other group all the pistillate spikes are frequently subradical, 
but when there is a pistillate spike at the base of the staminate 
its bract is squamiform and shorter than the inflorescence, or else, 
if rarely longer, it is auriculate and strongly reddish-tinged at base. 
Carex deflexa has a close ally in C. Rossii, the two species having 
been often confused. However, in C. deflexa itself the rootstock 
is very slender, horizontally creeping and short-stoloniferous; the 
perigynium is very small (about 2.5 mm. long) and strongly exceeds 
the scales; and the beak is short and but inconspicuously bidentate. 
In C. Rossi, as stated by Dr. Holm (Am. Journ. Sci. IV. 16: 37. 
1903), the rootstock is ‘much more robust, relatively shorter and 
ascending, not horizontal and not stoloniferous in the stricter sense 
of the word.” This last species, too, has larger, longer-beaked and 
conspicuously bidentate perigynia, with relatively longer scales. 
Carex brevipes W. Boott, a little-known species of the far 
western mountains, seems to be valid. It has the habit of growth 
of C. Rossti, but the short-beaked, narrow perigynia are about as 
small as those of genuine C. deflexa. 
A very interesting collection made by Mr. W. W. Eggleston 
in the northern part of New Mexico in IQII, contains two new 
species belonging to the group of species now under discussion. 
One of these is related to Carex deflexa and its allies. It differs, 
however, in having few-flowered pistillate spikes even on fully 
developed plants. The beak of the perigynium is obliquely cut 
at apex and becomes slightly bidentate in age, and the margins are 
scarcely serrulate. 
Another plant collected b? him, now named Carex geophila, 
differs from the real Carex umbellata in having a globose perigynium 
body fully 1.75 mm. or more wide; the beak is not strongly 
2-edged. The eastern species grouped with Carex umbellata 
have the perigynium body triangular-globose and but I-1.5 mm. 
wide, while the beak is strongly 2-edged 
Another species most closely related to Carex geophila occurs 
