Pare RTT A 
a IRIE NT EEE IE 7 
* 
1901] SPORANGIA AND GAMETOPHYTES OF SELAGINELLA 125 
sections which may be obtained by ordinary means is very 
limited. I found it necessary to imbed the strobili singly, 
examine each section as its plane approached the median region, 
and constantly alter the angle of the paraffin block in the micro- 
tome, until by the vascular strand of the axis and the leaf traces 
of the older sporophylls I could estimate approximately the 
section sought. Even with these precautions, I was not infre- 
quently doubtful as to whether a cell that appeared in the right 
place theoretically for an archesporium might not be the initial 
cell of a sporophyll from the next rank that a slightly oblique 
section would display. These explanations are necessary, for 
after much painstaking study of many sections of tips of very 
young strobili, I find myself becoming less certain of there being 
a definite rule governing the initial phases of the sporangial 
growth. In many median longitudinal sections cut in the man- 
ner described above, a single epidermal cell projects from the 
surface in a vertical line between the youngest sporophyll and 
the apical cell of the strobilus. Occasionally it takes a deeper 
stain than the neighboring cells. There are usually from ten to 
twelve cells between it and the apical cell, and two or three 
between it and the subtending sporophyll (fg. 5). Quite as 
often this projecting papilla is composed of fwo cells of equal 
size lying in the vertical line (figs. 3, 4). In either case peri- 
clinal walls are formed, cutting off one or two cap cells (figs. 2, 
5,6). From the hypodermal cell or cells, thus formed, origi- 
nates the sporogenous tissue. From the cap cell, which divides 
much more rapidly, the sporangium wall is formed. It is 
possible, I think, to determine in sporangia of various stages, 
nearly up to maturity, whether in a given case it originated from 
a single epidermal cell or from the two superimposed cells. In 
the second case, the complex of cells resulting from each of the 
two primary cells consists of more regular radial rows, and there 
is a quite definite plane of cleavage between the two groups. 
The only other interpretation of appearances like figs. 6 and 8 is 
that the primary single archesporial cell divided first anticlinally, 
thus producing the superimposed cells, each of which then cuts 
