On the Species of the Genus Primula in Essex. 155
were equal-styled through deformity of the style. On one oc-
casion, when over five per cent., of all the Primroses gathered
were equal-styled, a very severe frost had occurred about a
fortnight before, which probably had checked the development
of the pistil. The stigmas were unusually small. It was
noticeable that the stigmas of other flowers on those plants
of P. elatior which bore any equal-styled flowers were variable
in length. All the foregoing instances may, I think, be
regarded as long-styled flowers with imperfectly developed
styles; and as my first observation on the Cowslip, which
gave 2.2 per cent. as equal-styled, was made as early in the
year as March 29th, 1882, when very few were out, it is
possible that some of these might ultimately have become
normal. I have heard of cases, but never saw one, in which
cultivated Primroses, properly short-styled, have become
equal-styled from another cause, namely, imperfect develop-
ment of the tube of the corolla. Thus it may be seen how
constant this heterostylism is; neither Mr. Darwin nor
myself, after an examination of an enormous number of wild
Primula flowers, being conscious of ever having met with a
truly equal-styled one.12 Nevertheless the length of the style
is variable to a certain extent, especially, I believe, in the
flowers of the long-styled form, and even in the same umbel.
Thus I have seen the stigma in three flowers on one umbel to
project as much as one-fifth of an inch, which is not unusual,
and often the stigma is as far within the tube. I have also
not unfrequently seen the stigma of unopened buds projecting
one-eighth of an inch beyond the tips of the teeth of the
calyx.
In the spring of 1882 the idea occurred to me to enquire
how early in the development of the bud heterostylism exists.
I therefore procured some very young buds of P. elatior and
P. vulgaris, and, dividing them longitudinally, I found that
in the very youngest, in which not a vestige of the petals
appeared externally, the two forms were totally indistinguish-
able unless it were by some microscopic difference. The
anthers seemed to reach their full size and to become of a
bright orange-colour long before the bud opened ; they were
attached to the petals slightly above the ovary. The petals
were quite rudimentary and almost devoid of colour, much
12 If I may trust my memory, some specimens of P. farinosa which I
gathered in the Alps were truly equal-styled, the stigma appearing with
the anthers at the top of the tubes. If so, they were of singular interest,
as I never met with such a case in any other species. Mr. Scott says
(Proc. Linn. Soc. (1864), vol. viri., p. 115) that it is not uncommon for
this species to be equal-styled.