92 MR. В. P. GREGORY ON THE 
which have come under my notice. If suitable examples be chosen there 
is no difficulty in diseriminating between these four types, but each type- 
form is connected with the next in the series by a number of intermediate 
forms, which so completely fill in the gaps between the types, that the latter 
ean only be looked upon as the central forms of the 
convenient to arrange the various forms of flower *. 
For the sake of convenience I have applied Darwin’s terms of “ long- 
styled " and * short-styled " to the two type-forms in the males, since these 
terms fairly describe the main point of difference between them. At the 
same time it must be pointed out that, in Valeriana dioica, there is no sharp 
discontinuity between the two forms, such as the use of these terms might, 
by analogy, seem to imply. 
Among the female plants, Г have found large numbers of plants which 
agree with the description of Müller's fourth class :—% Female, with scarcely 
visible rudiments of anthers" ; but with regard to his third group there is 
groups in which it is 
some doubt. Perhaps it corresponds with those plants which it has seemed 
to me to be best to class as hermaphrodites. 
In the following deseriptions, therefore, the four groups are called: 
(1) short-styled male ; (2) long-styled male ; (3) hermaphrodite ; (4) female. 
Of these the last is much the most easily recognized, and perhaps shows 
fewer transitional forms, if only for the reason that, in the production of 
pollen, one has a definite character which marks off the hermaphrodite. For 
this reason the description of this form is given first, and is followed by those 
of the other forms, in the reverse order to that used by Müller. 
Female Plants. 
The great majority of the female plants which I have examined have borne 
flowers which correspond well with the description of Müller's fourth type— 
* Female, with searcely visible rudiments of anthers, with the smallest 
corolla." 
These plants ean be recognized readily by means of the small corolla and 
relatively inconspicuous flowers. The fully open flower measures about 4 mm. 
from the base of the inferior ovary to the tips of the petals. In such a flower 
the ovary will be about 2:5 mm. long, the corolla about 1:5 mm. long, while 
the style projects beyond the corolla-lobes for at least 1 mm. (РІ. 8. fig. 1). 
These measurements are of course only roughly approximate ; the more so as 
the relative sizes of the various parts vary very much as the age of the flower 
increases. In the newly opened flower, the style scarcely projects at all 
beyond the corolla, and the ovary is relatively short (figs. 2, За). As the 
* Müller leaves open the question of the existence of forms intermediate between his 
types. 
