532 STEVENS: DIOECISM IN THE TRAILING ARBUTUS 
adapted for retaining pollen. This first group he subdivides on 
the basis of whether the stigma projects above the corolla or is 
included within the tube. In the ‘‘ shorter styled form,’’ how- 
ever, the height of the stigma varies considerably, being ‘ some- 
times as low as the middle of the tube of the corolla, sometimes 
nearly up to the throat.”” This main group is further character- 
ized by having aborted stamens. The stamens of this form show 
varying stages in abortion, from mere rudiments of filaments to a 
condition of fully developed filaments bearing anthers with occa- 
sional pollen grains. They are, however, always functionless and 
the form is evidently pistillate. 
The second group is characterized by having small, erect, and 
comparatively smooth and dry stigmas which do not retain pollen 
readily. The ovaries in this form are apparently as well devel- 
oped as those in flowers of the first group, and the styles show the 
same variations in length. This small stigma form has well 
developed stamens with anthers that abound in pollen. The 
stamens vary much less in length than do the pistils but there 
is ‘‘ a tendency to having lower instead of higher anthers in con- 
nection with the short style.’ In conclusion Gray raises the 
question as to whether Epigaea is really dioecious or the small 
stigma form sometimes sets fruit; and suggests that the modifi- 
cations in the length of the style point toward a heterostylous 
condition like that of the primroses. 
Since Gray’s time several papers dealing with the flowers of 
Epigaea have appeared. All of them agree in most particulars. 
with his observations. Halsted® found all the kinds of flowers 
described by Gray, and notes that about one third of the flowers 
of the small stigma form have the stigmas even with the anthers. 
He measured the pollen grains from the longer and shorter stamens 
and found no difference in their size. As a difference in the 
size of the pollen of the different forms is common among hetero- 
stylous plants Halsted concludes that Epigaea shows no real 
evidence of heterostyly. 
Wilson” 
examined a thousand plants, mostly from North 
Carolina, 
He found that the kind with perfect stigmas showed 
all gradations in length of style from the longest to the shortest. 
In only one of these large stigma plants did he find even aborted 
