I9Io] * KERN—PERIDIAL CELLS IN ROESTELIAE 449 
case the cells lie straight, and there is usually not much disparity 
between the breadth and thickness (figs. 1, 10, 13, 14). There are 
a number of species, however, in which the cells tend to lie only on 
their sides when mounted in water, and it is only with difficulty 
that a face view can be observed. The cause for the cells taking this 
position is that they become much curved in water (jigs. 2, 5, 6, 7), 
and their equilibrium, therefore, is much more stable when they 
lie on their sides. Such cells are hygroscopic and will straighten 
out when they become dry. It has been found that the curvature 
<p 
a b ¢ 
a o. 
a 
_ Fic. 2,—Cross-sections of peridial cells of several species: inner walls are above 
in the drawings; a, G. clavariaejorme (redrawn from FISCHER), inner and side walls 
of almost equal thickness; b, R. cornutum, inner and side walls of about same thickness; 
¢, G. inconspicuum, inner wall excessively thickened, side walls not so thick, becoming 
thinner toward outside. 
is always inward, the outer wall being on the outer side of the curve. 
It is interesting to note the effect which'this incurving of the individual 
cells has upon the peridium as a whole. R. pyrata (Schw.) Thaxt., 
the aecial stage of G. Juni peri-virginianae Schw., is one of the species 
which has very marked hygroscopic cells. In the ordinary dried 
condition of herbarium specimens, the peridium of this species is 
finely fimbriate and strongly recurved, giving the appearance of 
having been combed outward. In a saturated atmosphere or in any 
way under the influence of moisture, the revolute chains of cells may 
be seen to unroll themselves and take a more or less erect position, 
or even become somewhat bent inward. They assume their recurved 
position again upon drying. 
With regard to the surfaces of the peridial cells, it has been found 
that the species so far recognized divide at first into two classes, one 
