134 Rhodora [AvausT 
two plants could be separated with a marked degree of success. With 
the present interest in E. peregrina, a comparative study of the three 
plants was undertaken — not only because of the previous conviction 
of the distinctness of E. Purshii but particularly because of a number 
of exceptions noted by Professor Wiegand in his discussion of the 
distinguishing character of E. peregrina. Hackel had compared his 
plant chiefly with E. pilosa but he also distinguished it from Æ. 
Purshii. Professor Wiegand on the other hand, including E. Purshii 
in E. pilosa, compared his E. peregrina with a much more complex 
species-group. It was suspected that the breaking down of certain 
characters through exceptions was alone due to the failure to dis- 
tinguish E. Purshii. With a further study of the group these excep- 
tions were found to clear away and E. peregrina, as well as E. Purshii, 
to stand free from E. pilosa. 
"The characters of Eragrostis peregrina compared with those of E. 
pilosa and E. Purshii, critical comments, and certain observations 
apparently new may be taken up in the general sequence of a detailed 
description. 
Of the most constant differences, the greatest stress is rightly laid 
by Professor Wiegand upon the absence of the long hairs on the 
auricles of the sheath. A slight delimitation of this character, ap- 
parently, should be made. The condition would seem to be more 
accurately described, if the absence of the hairs be noted on, say, the 
upper sheaths — the lower sheaths and those on short sterile branches 
are very often supplied with well developed auricular hairs. This 
possibly somewhat technical observation in no way depreciates the 
value of this important character brought to light by Professor 
Wiegand. l 
The distinguishing character of the solitary branches of the panicle, 
maintained by Hackel, is to be given critical consideration. That 
the panicle-branches of E. peregrina are solitary and those of E. pilosa 
in 2's or 4's in the strict sense does not seem to be borne out by a 
series of specimens, but there is here the germ of a very excellent 
diagnostic character. In the examination of several hundreds of 
specimens from many different stations it has been noted that the 
base of the panicle in E. peregrina is consistently composed of a single 
branch, its point of origin well differentiated from the branch next 
above, while in E. pilosa it is characteristically a pair or a whorl of 
branches. In E. Purshii a somewhat intermediate condition seems to 
