254 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 
oil bodies, but rhizoid-initials are apparently not present. The ap- 
pendages vary in number from one to four, when the genus as a 
whole is considered, and a good deal of variation is to be expected in 
most of the species. In certain cases, to be sure, a single appendage 
is the rule, but even here prolonged search will usually bring to light 
scales with two appendages. Sometimes, under these circumstances, 
one appendage is smaller than the other or appears in the form of a 
basal lobe. The appendages vary greatly also in size, in form, in the 
character of the margin, and in the apex, many of the variations be- 
ing associated with environmental differences. In certain cases 
specific differences have been based upon these inconstant features. 
Stephani, for example, described the appendages of A. wrightii as 
being sometimes armed with a single large spine, and on the basis 
of this character placed the species in a group with dentate or 
lacerate appendages, instead of in the same group as A. elegans, of 
which it is actually a synonym. In this case the marginal teeth 
represent a somewhat unusual feature, possibly associated with a 
more xerophytic environment. He likewise assigns acuminate ap- 
pendages to A. tenella, in which, as a matter of fact, rounded ap- 
pendages are not infrequent. 
The inflorescence in Asterella may be paroicous, autoicous, or dioi- 
cous. As a rule, each species shows a definite type of inflorescence, 
but certain autoicous species sometimes exhibit a tendency toward 
a dioicous condition. When the inflorescence js paroicous, as in 
A. tenella, the antheridia form a vaguely defined cluster close to the 
base of the stalk of the female receptacle. Such an androecium is 
very slightly elevated, the ostioles are low, and there is no surround- 
ing fringe of narrow scales. In the autoicous species the androecium 
is more clearly defined. In some cases, as in A. pringlei, it forms an 
elongated and slightly elevated median patch which may be forked; 
the ostioles are more pronounced than in the paroicous species, and 
a scanty fringe of paleae may be present. Such an androecium ap- 
parently never limits the growth of the branch and makes its ap- 
pearance at some distance behind the apex. In other cases, as in A. 
elegans, the androecium forms an oval or circular disk, distinctly 
elevated, and usually with a well-developed fringe of paleae. Such 
an androecium limits the growth of the branch and thus appears 
terminal in position. In A. californica, the only definitely dioicous 
species at present known in America, the androecium is of the elon- 
gated type, which does not limit the growth of the branch. So far as 
observed, the epidermal pores of the androecia are simple. 
The female receptacle is borne on an elongated peduncle com- 
posed of compact parenchyma throughout and showing on its mor- 
phologically ventral side a single furrow with tuberculate rhizoids. 
