4 
in the Stock (Matthiola iucanci). 
475 
8. A cursory survey of the Ericales .certain Malvaceae, and some isolated genera 
which are held to have commissural stigmas , leads to the same conclu¬ 
sion. The commissural stigma also not a reality in the Orchidaceae , the 
gynoecium being composed of six carpels (G 3 + 3). 
The elimination from the Rhoeadales of these anatomical anomalies, 
and the acceptance in their place of the differentiation of two kinds of carpels 
and of a certain interplay between them, necessitates at least a brief con¬ 
sideration of certain other unrelated families in which a similar disposition 
of the stigmas or mode of origin of the dissepiments has been held to 
prevail. Chief among these are the large group of families included in the 
series Ericales, the Orchidaceae and certain of the Malvaceae. To deal 
adequately with these cases within the limits of a single paper is impossible. 
It can, however, be stated at once that an examination of even a few types 
has already shown clearly that the same processes of reduction and con¬ 
solidation have been in progress in these diverse groups as in the Rhoeadales. 
It is therefore proposed to conclude the present account with a brief refer¬ 
ence to one Monocotyledon family, and to present the evidence in the case 
of the other families in a later communication. We will, therefore, very 
briefly consider the Orchidaceae from the present standpoint. 
A perusal of Eichler’s statement of the facts regarding the Orchid 
gynoecium ( £ Bliithendiagramme ’, ii, p. 182) and a consideration of the figure 
there given (Fig. 108) of a cross-section of the ovary of Cypripediiun 
Calceolus , L., make it difficult to understand how any one not already com¬ 
mitted to the commissural-stigma view and hence concerned to bring all 
facts into line with it, can fail to realize that the six structures composing 
the Cypripedium gynoecium, structures fundamentally equivalent in every¬ 
thing but size, represent six individual carpels, three of a pseudo-valve shape 
alternating with three which have become solid. Of the fact that in some 
Orchidaceous genera three of the six structures—‘ Zwischenstiicke \ as Eichler 
terms them—are sometimes wanting, he offers no explanation. In the light 
of our conclusions in regard to the Rhoeadales we have a clue to this vary¬ 
ing conformation within the Orchid family. The full Orchidaceous ground 
plan is undoubtedly G3 + 3, but in existing forms where all six are present 
one whorl of carpels has become solid and sterile, as can be seen from 
Figs. 54, 55, showing the appearance of the ovary of Calanthe vestita , L., 
and Cattleya labiata, L., in cross-section. A similar condition is depicted in 
the figures of the cross-section of the gynoecium of Selenipedilum , Phragmo- 
pedilum , Paphiopedilum , Cypripedium , and Trichopilia fragrans , Reichb., in 
‘ Pflanzenreich ’, iv, 50, p. 22, Fig. 12 A, B, c ; also ‘ Pflanzenfamilien’, ii, 6, 
p. 82, Fig. 83 E, F, G, and p. 194, Fig. 208. The ripe fruit, generally 
a capsule, splits more or less completely into three broad strips (Eichler’s 
‘ Klappen ’) and three narrower ones (his * Zwischenstiicke ’) (as shown in Fig. 
