536 STEVENS: DIOECISM IN THE TRAILING ARBUTUS 
dry stigma. The writer experimented with the artificial polli- 
nation of this small stigma form both with plants cultivated in 
pots and with flowers kept in water. The pollen adheres to these 
stigmas to a considerable extent, though this is probably due to 
the fact that the wall of the pollen grain is somewhat mucilaginous, 
for the pollen adheres also to the style below the stigma and even 
to smooth objects, such as a glass rod. In no case, however, was 
a pollen grain observed to develop a tube when placed on a stigma 
of this kind. The pollen was retained without degeneration for 
some time, frequently for over a week, but it never established any 
organic connection with the pistil and could easily be brushed off. 
As these cultivated plants appeared perfectly healthy and after- 
ward developed apparently normal new shoots, the failure of the 
pollen to germinate cannot be attributed to any unusual weakened 
condition of the pistils. Moreover, the same condition was found 
to occur in long-styled plants of the small stigma form growing 
under natural conditions. Although the stigma in these plants 
projected well above the stamens it was covered with pollen, 
probably as the result of the visits of insects. Not a single gra‘n, 
however, had emitted a tube. 
This evident sterility of the pistils in the staminate form seemed 
still more remarkable when microtome sections of the ovules were 
examined. At the time of fertilization the embryo sac of the large 
stigma form shows uniformly typical egg and synergids, well 
defined antipodal cells, and the two polar nuclei lying close to- 
gether near the middle of the sac. In all probability it remains in 
this condition over winter (Coulter and Chamberlain,? p. 53). 
The ovules of the sterile small stigma form present an identical 
appearance both in size and in the condition of the embryo sac. 
It is of course possible that the nuclei of this functionless embryo 
sac have not undergone a reduction division. There appears to be 
no evidence, however, of a tendency toward apogamy, and these 
unfertilized ovules gradually degenerate. 
We have then, in the small stigma form of Epigaea, flowers 
that are apparently perfect but functionally male. The mor- 
phological differences between the functional and non-functional 
pistils in this species are, as pointed out above, very slight and are 
confined to the stigmas. Their physiological differences are so 
