FORMS OF FLOWERS IN VALERIANA DIOICA, 95 
whereas the staminal rudiments of the females are colourless. Further, these 
two plants, together with the two hermaphrodites obtained from the same 
cross, were in flower оп May 11, 1906, at a time when as yet the only other 
plants in flower were males. 
Both the two hermaphrodite plants produced small quantities of pollen, 
which appeared to be good. The death of one of these plants, during the 
winter of 1905-6, prevented further observations on it. The other plant is 
shown in figs. 13, 12a. Although not very different from its sister plants, it 
had somewhat larger anthers, which showed the reddish tinge like that of the 
young anthers in male plants. Many of these anthers contained a small 
amount of pollen, but, on the other hand, those of other flowers of the same 
plant were devoid of pollen. These two plants therefore approach the female 
type very nearly. 
One other hermaphrodite plant, raised by breeding from another cross, 
must be placed at the other end of the series of. forms, approaching the type 
of structure shown in the long-styled males (see p. 96). In this plant 
(figs. 15-18), the corolla was as large as that of the ordinary male plants, the 
stamens were larger than in the typical hermaphrodite (and therefore very 
much larger than those of the plants just described), and the stigmatic 
lobes at the apex of the style were smaller. In 19006 this plant showed its 
hermaphrodite nature by producing some fruits. 
One of the plants found growing wild was of a type almost exactly similar 
to this. It was used successfully as a male parent in some breeding 
experiments. These * Hermaphrodite” plants, then, constitute a series of 
forms, which may be grouped round such a form as that shown in fig. 14 as 
a central type, but on either side approach the neighbouring classes very 
closely. 
Male Plants. 
Müller divides the males into two classes: (1) those without the rudiment 
of a pistil, with the largest corolla; and (2) those with a rudimentary pistil, 
with a somewhat smaller corolla. 
It may be said at once that, in all the males which have come under my 
notiee, some rudiment of a pistil is present, though it may be scarcely half a 
millimetre in length, requiring careful dissection of the flower in order to 
expose it. 
There is not, in my experience, any constant difference in the size of the 
corolla in the different types ; and although the usually shorter ovary of 
the short-styled forms gives a higher value to the ratio (length of the 
corolla) : (length of the ovary), I have not found the size of the corolla any 
guide as to the nature of the flower. The point at which the stamens are 
inserted on the corolla-tube is so nearly the same in the different types, that 
I have not been able to diseriminate between them in this respect. Оп the 
