STEVENS: DIOECISM IN THE TRAILING ARBUTUS 535 
corresponding in length to the shortest pistils has been found by 
the writer. 
The pollen germinated readily, and sections made twenty-four 
hours after pollination showed the stigmas crowded with pollen 
tubes. The tubes, however, developed rather slowly, and it was 
only at the end of five days, 120 hours, that pollen tubes were 
found extending to the ovules, and in a few cases the polar nuclei 
appeared about to fuse with the male nucleus. It is noteworthy 
that the pollen from both lengths of stamens germinated with 
equal readiness on all lengths of pistil. Such would not be the 
case, of course, in a truly heterostylous plant. 
DIOECISM 
The evidence as to dioecism is much more satisfactory. Un- 
questionably the large stigma form is uniformly pistillate, for 
usually only mere rudiments of filaments appear and often none 
at all. Gray (loc. cit. 74), however, found in some anthers in a 
flower of this form a few ‘‘ perhaps well formed pollen grains,” 
but the anthers never opened. In the one thousand plants 
examined by him, Wilson (loc. cit. 59) found only one large stigma 
plant that showed even aborted anthers. The writer examined 
over two hundred plants of the large stigma form, and only three 
bore flowers with any rudiments of anthers. Microtome sections 
of these flowers were prepared, and while the four locules could in 
some cases be distinguished they contained merely masses of 
broken down cells. 
The form with well developed stamens, on the other hand, has 
never been observed to set fruit; and as Gray (loc. cit. 76) points 
out, it is easy to ascertain in any case which kind of flower has 
matured fruit, because the style and stigma persist until the fruit 
is fully mature. Indeed, in the flowers in which the ovary has 
not developed, the style and stigma often persist for over a year, 
somewhat shrunken, to be sure, but still sufficiently well preserved 
to make it certain to which form the flower belongs. This unusual 
persistence of the style and stigma seems to be due to the fact 
that the carpellary bundles contain strongly lignified tracheids. 
The pistil of this form, however, is, as Gray (loc. cit. 75) 
observed, apparently normal except for its smaller and rather 
