STEVENS: DIOECISM IN THE TRAILING ARBUTUS 537 
great, however, that pollen which germinates readily on the 
stigmas of one form is absolutely inert on those of the other, 
and Epigaea repens is functionally a dioecious plant. 
A few other cases of plants that are apparently polygamo- 
dioecious but actually dioecious have been observed. Darwin 
(loc. cit. 288) noted that the “‘ hermaphrodite” flowers of Euony- 
mus europaeus are practically male, so this species is really dioe- 
cious. Mottier'’ (p. 377) and von Kirchner® (p. 116) have pointed 
out that the apparently polygamodioecious maples are functionally 
dioecious, because the anthers of the “‘ perfect ”’ flowers, although 
they contain some apparently normal pollen grains, never open. 
Celastrus scandens is an apparently polygamodioecious plant, 
having pistillate and “ perfect’ flowers. Like the related Euon- 
ymus europaeus, however, it is dioecious in function, for the 
pistils of the “ perfect’ flowers are smaller than those of the 
pistillate and appear upon investigation to be non-functional. 
Celastrus is abundant in the vicinity of New Haven, Connecticut, 
and during the spring of 1911 the writer examined a large number 
of plants in this region. Particular care was taken to observe 
the plants bearing “ perfect’’ flowers. Not only was there no 
indication of any development of the ovary but the flowers 
shriveled and dropped from the plant almost as soon as the pollen 
was shed. Microtome sections showed that the ovules in the 
pistillate flowers are more than ten times as large as those of 
the ‘‘ perfect’ flowers. Many of these non-functioning ovules 
contain, however, an apparently typical embryo sac. 
The condition found in the maples is exactly the reverse of 
that found in Epigaea and Celastrus. In the latter species there 
are two kinds of flowers, one evidently pistillate, the other de- 
veloping both ovules and pollen grains. In some species of 
maples, on the other hand, one kind of flower is evidently stam- 
inate and the other apparently perfect. In Epigaea and Celastrus, 
however, the ovules of the ‘‘ perfect ’’ flowers are never fertilized, 
and in the maples the anthers of the “ perfect ’’ flowers do not 
open. Notwithstanding their morphological differences, then, 
both types are functionally dioecious. 
As Wilson (loc. cit. 59) states, there is in Epigaea repens an 
apparently constant difference in size between the large-stigma 
