80 AMERICAN FERN JOURNAL 
lying immediately below such ledges. It prefers full 
exposure to the sunlight, and sends its roots back deep 
into the joints of the rock. On prying off a piece of the 
ledge, I have exposed a mass of fern roots which ex- 
tended back nearly a meter into the crack. It is w- 
doubtedly this habit of deep rooting which enables the 
plant to live on dry and sun-baked ledges. In winter 
it can get almost no protection from the snow, but I 
have never seen it injured by the very low temperatures 
(—35° C and even —40° C) which are not infrequent. 
have never seen P. atropurpurea in Minnesota or west- 
ern Wisconsin. All of the cliff-brakes there are um- 
formly of the other species and they uniformly show the 
blue-green color mentioned by Pickett in his first article 
on these ferns.‘ 
It is thus seen that Pellaea atropurpurea and P. gla- 
bella occupy almost entirely distinct ranges, which 
barely overlap along a line running from Vermont 
ansas. Each species is remarkably uniform through- 
out a very extended range and the differences betwee? 
them, so well brought out by Pickett, are equally wel 
marked whether specimens are compared which come 
from the extreme limits of their range, or are taken from 
plants growing together on the same cliff, and collected 
at the same time. In view of these facts, there can be, 
I think, no question that they are perfectly distinct 
species. 
In regard to the differences between them, I may 
cite a few points not noted by Pickett in his article, : 
and one point in which my observations are not fully 
m accord with his. The scales of both species are COM 
posed, except at the very base, of elongated cells, thom 4 
m4 P. atropurpurea being 10-20 times as long a the 
are wide, those of P. glabella about six times as loné * 
they are wide. The scales of P. atropurpurea are a 
* Amer. Fern Jour. 4:97. 1914. 
