32 BOTANICAL GAZETTE {JULY 
megasporophyll is traversed by a large number of vascular strands 
and resin canals; near the top of the sporophyll the latter are evenly 
disposed all round the periphery, and inside each canal is a vascular 
bundle with the phloem on the side nearest the canal. On the lower 
(outer) surface of each scale is a small umbo. 
The structure of the ovule before and shortly after pollination has 
already been described and figured? (10, figs. 1 and 3). The ingrowth 
of the micropyle-closing cells is not followed by septation in them, 
as it is for example in Pinus. A transverse section of the micropyle 
is shown in fig. 2, in which it may be seen that the opening is often 
not entirely closed by the ingrowth of these cells. This figure shows 
a most remarkable resemblance to a similar section of the micropyle 
of Bennettites, figured by LIGNIER (6, fig. 31), as pointed out to me 
by Professor SEWARD. 
Some time after pollination the sporogenous tissue becomes evident. 
Seen in median section it appears as about 18 or 19 large cells with 
dense cytoplasm and large lightly staining nuclei, sharply marked off 
from the surrounding cells, which have vacuolate cytoplasm and 
smaller deeply staining nuclei. The position and structure of these 
cells are indicated in figs. 3 and 4. Their number‘ is about 64, 
possibly exactly that number, in which case they are not improbably 
formed by repeated simultaneous divisions of a single archesporial 
cell. It is however quite as likely that the number is somewhat less 
than 64, and that they have not a common origin from a single 
archesporial cell. They are all exactly alike when first differentiated, 
and their subsequent behavior indicates that they are spore mother 
cells, only one of which, however, is functional. Sometimes a single 
layer of tapetal cells may be seen a little later. 
At a somewhat later stage, one of the central mother cells becomes 
a very little larger than the remainder, the nucleus also enlarges very 
slightly, and the wall becomes a trifle thicker than that of its neighbors. 
This is the functional megaspore mother cell. At about the same time — 
the non-functional cells elongate somewhat, and certain peculiar — 
3 The wings of the ovule were accidentally patient = 166) as being only one 
cell thick. This is never the case, even close to the 
4 Calculated on the assumption that the actual ee is to the number seen in 
median section as $773 to mr? 
