Apparent Apogamy in Pterygodium Newdigatae. 
595 
Although a single type of column is usually constant for any one spike, 
yet individuals were observed in which the number and shape of the rostellum 
arms of the different flowers showed considerable variation. 
The strong perfume of the cleistogamous flower is possibly a relic of the 
time when insect pollination was essential for seed production. The zygo- 
morphy of the normal flower is more or less completely lost in the cleisto- 
gamous form, but resupination of the ovary occurs as usual. 
The small gland-like pollen masses (?) remain permanently embedded in 
the tips of the rostellary arms (Fig. 26). The supposition that the pollen grains 
germinate in situ proved to be incorrect. Freehand sections through the base 
of the column of flowers of various ages showed no trace of pollen tubes, 
though these are readily demonstrated in the case of other Orchid flowers. 
Material, collected at different times, was fixed in chromacetic acid and 
microtomed. The youngest buds available had ovaries from 3-4 mm. in 
length. The tips of the rostellary arms contained compact sporogenous 
tissue (?), composed of somewhat elongated cells with dense protoplasm and 
well-defined nuclei (Fig. 27). No indication of nuclear division was observed. 
The tissue appeared to be in the mother-cell stage. Ovules of the same 
flower showed no trace of integuments. The single hypodermal cell enlarges 
and functions directly as the megaspore-mother-cell (Fig. 28). 
Somewhat older flowers exhibited no further development of sporogenous 
tissue (?) in the rostellary arms (Fig. 29). The megaspore-mother-cells, 
however, appeared for the most part in synapsis (Fig. 30). This condition 
evidently lasts for some time, as not only was it present in material collected 
on several different occasions, but was also found to occur in inverted ovules 
with fairly well developed integuments (Figs. 31-33). In no case was there 
any trace of pollen tubes in the ovary, in the tissues of the column, or in the 
rostellum arms. The statement is often made that Orchid ovules do not 
develop far unless pollination has occurred. It is worthy of note that in the 
species under discussion ovular development is apparently completed in the 
absence of fertile pollen. 
Unfortunately several gaps occur in my material, the most serious being 
from October 19 to November 8 ; thus, further stages in the development of 
the megaspore-mother-cell and embryo sac are missing. Ovules of flowers 
received on November 8 contained young embryos, while much of the 
material received two weeks later had the embryos fully formed. Microtome 
sections through the rostellum arms of flowers with conspicuously swollen 
ovaries showed no appreciable change in the archesporial (?) tissue. Fig. 34 
is a drawing of isolated cells from the disorganising sporogenous (?) tissue 
of a flower, the seeds of which had already been shed. 
Sterility of the anthers has been shown in certain genera to be associated 
with apogamous development of the embryo. It seems reasonable to conclude 
that we are here also dealing with a case of apogamy. 
