405 
Embryo of Sciadopitys verticillata. 
Sciadopitys there is a specialized pollen-receiving device which is quite 
unlike any of these. At the time of pollination the integument extends 
but very slightly above the surface of the nucellus, in such a fashion as to 
form a wide gaping but shallow micropyle. Shortly before the pollen is 
shed, the upper part of the nucellus becomes sharply differentiated into an 
organ which, for convenience, I will call the pollen-cushion. This cushion 
consists of comparatively large cells with intercellular spaces, and with thin 
cell-walls forming a loose soft tissue which stands out in sharp contrast 
against the small-celled firmer tissue of the nucellus below. These cushion- 
cells secrete at the surface a transparent liquid substance, which, from an ex¬ 
amination of the fresh ovules, appeared to be more refractive and more dense 
than water. This pollen-cushion, therefore, not only serves as a stigmatic 
surface for receiving the pollen, but also furnishes a soft loose tissue into 
which the young pollen-tubes may penetrate with very little resistance. 
Now, while such special adaptations as these pollen-receiving devices 
may not be of direct phylogenetic importance, they are nevertheless very 
interesting. They show a series of modifications of structure which 
probably followed close upon the adoption of the pollen-tube as a means 
of conveying the sperm-nuclei into the egg. 
In Sciadopitys the pollen-cushion is ready to receive the pollen early 
in April, and, as stated above, the integument at this time projects very 
slightly above the surface, forming a wide micropyle into which the pollen 
is blown by the wind. During the latter half of April numbers of pollen- 
grains were found in this position. In some ovules but one or two grains 
were found, while in others the micropyle was crowded full of them. A few 
of them are shown in Fig. 6 . 
The next step in the germination of the microspore was its rapid 
enlargement and the casting off of the exospore. This takes place 
immediately after the pollen-cushion has been reached. These changes are 
shown in Fig. 5. Meantime, the integument continues its growth forward, 
as may be seen in Fig. 6, and the micropyle is gradually closed. The 
complete closure of the micropyle occupies about three weeks. During 
this period the pollen-tubes have made their appearance, and have penetrated 
for some considerable distance into the tissue of the pollen-cushion. 
Although the pollen-tube has attained a considerable length, no nuclear 
activities occur until the middle of June. About this time the generative 
nucleus divides, giving rise to a large, more or less, spherical cell with 
a distinct cell membrane and a free nucleus. These structures, by their 
later behaviour, were readily identified as the body-cell and the stalk- 
nucleus. Their position and relative size may be easily made out in Fig. 8, 
which represents a longitudinal section of the pollen-tube immediately 
after the division of the generative cell. It will also be seen from this 
figure that these two structures remain for some time in the spore end of 
