1919] Long,— Eragrostis peregrina 137 
pilosa. It appears that E. pilosa has essentially smooth empty 
glumes while a scabrous keel seems to be a very excellent index of E. 
Purshii. Good characters are also to be found in the actual and 
relative dimensions of the empty glumes. In E. peregrina the lower 
empty glume is very much reduced, measuring only about 0.5 mm. in 
length; the upper, about 1.0 mm. About the same condition occurs 
in E. pilosa but some variability is present. In E. Purshi the lower 
measures about 1.0 mm. and the upper about 1.5 mm. The minute 
lower empty glume will constantly differentiate E. peregrina (and 
usually E. pilosa) from E. Purshiv. 
That “the florets are. . .somewhat smaller," as Professor Wiegand 
observes, would appear to belong to the comparison with E. Purshit. 
There appears to be little tangible difference between those of E. 
peregrina and E. pilosa but the rather larger florets of E. Purshit will 
be appreciated with a little experience. Dimensionally the former two 
species may be said to have florets usually under 1.5 mm. in length; 
the latter, commonly over 1.5 mm. 
Professor Wiegand, in commenting upon the statement of Hackel 
that “The plant is closely related to the E. Purshii (caroliniana) but 
. js distinguished from it by the absence of conspicuous lateral nerves 
on the flowering glume,” says, “The lateral nerves of the flowering 
glumes are always inconspicuous but there are occasional specimens 
of E. pilosa in which they are equally indistinct." My own observa- 
tions indicate that, as in other characters, E. peregrina shows itself 
to be very closely related technically to E. pilosa, and in the matter 
of distinctness of the lateral nerves very little if any differentiation 
can be drawn between these two species, but E. Purshit may be 
separated from them both, with a fair degree of satisfaction, by its 
quite conspicuous nerves. Professor Wiegand's observation, one 
may believe, is due to the abundant material of E. Purshii which he 
had for comparison and the relatively small series of real E. pilosa — 
if the material at Cambridge and New York runs similar to that at 
Philadelphia. 
The use of*texture and color of the flowering glumes by some authors 
as key characters to separate E. pilosa and E. Purshi induced a 
comparative examination of these features in the three species. At 
best, characters of this kind do not seem overly distinetive. In the 
present case very little satisfaction is obtained in endeavoring to 
distinguish between “thin” and “firm,” and “purplish,” whether 
