478 REPORT OF THE CONFERENCE ON GENETICS. 
with very short filaments and large anthers; and three other species with 
five stamens of the first kind, and one very different, much larger, declinate 
stamen of the latter kind, which corresponds exactly with the only fertile 
stamen in monandrous orchids. Miers plainly says “one fertile stamen 
placed anteriorly and five substerile anthers.’’ In the description of 
C. capensis (“ Bot. Mag.” t. 568) the interesting remark is made that upon 
the fertile stamen the style is incumbent. 
By combining the mediano-symmetrical flower of Tecophilea with 
its staminodial posterior stamens and the large stamen of Cyanella, we 
have nearly the diagram of orchids ; but it is also necessary to compare the 
other parts of the flower, and the whole plant, in the tribe of Conantherea, 
with the most simple orchids. 
The ovary is superior, semi-superior, or nearly inferior in Conantheree. 
If we acknowledge that the column of orchids is only a prolongation of 
the axis, in which the carpels of the inferior ovary are included, there is 
no difficulty in seeing that the ovary of orchids may have been derived 
from that of Conanthereg. There are numerous small anatropous ovules 
in this tribe, but the ovary is trilocular, as in Apostasieeg, and in 
Phragmopedilum only among orchids, and the seeds have a horny 
endosperm, which has quite disappeared in orchids. 
The Conantheree have a subterranean tuber, a distinct cylindrical 
stem and vaginate, distichous, conduplicate, narrow or broader convolute 
leaves (Cyanella). The inflorescence is racemose, as in most plants with 
mediano-symmetrical flowers. 
I must also mention here the little group of the Philydracea. They 
have a superior ovary, but the diagram of the racemose flowers is very 
similar to that of monandrous orchids. In Philydrwm the placentie 
do not meet in the middle of the nearly unilocular ovary, and the seeds 
are small and very numerous. These plants are found in the tropical parts 
of Asia and Australia, while the Conantheree grow now in South America 
and South Africa. All these parts of the globe have had no general 
overflowing of the sea since the Early Tertiary period, and may therefore 
still retain ancient types of plants. 
If we are of the opinion that these plants, and the few now existing 
Conantheree and Philydracee, are the remainder of the connecting link 
between Amaryllidacee and Orchidaceae, there arises the question which 
of the latter are the most ancient and original types. 
I think that the articulation of the leaf in monocotyledonous plants is 
a character which points to a recent origin. We find this articulation 
only in the tribes of Bambuse@ among grasses, the highest group of this 
family, and in most tribes of orchids (the Apostasiee, Cypripediee, 
Ophrydee, and Neottiee excepted), and I should think they are the oldest 
orchids. Also most species of these tribes have granular pollen, and are 
terrestrial. We may suppose that the possibility of dejecting the lamina 
of the leaf was developed as the orchids began to grow as epiphytes, and 
were sometimes constrained to diminish their water-evaporating surface 
if they would keep themselves alive in times of drought. 
Of the orchid tribes named above ; the Apostasie@ come very near the 
diagram of Tecophilea, the perianth is only a little symmetrical ; even the 
prolongations on the back of the perianth-leaves, which are present in 
