re ORC REVEE W. 
VoL. XVIII.] APRIL, tgto. [No. 208. 
THE EVOLUTION OF THE ORCHIDACESZ. 
(Continued from page 36.) 
‘WE now come to the great division of the Order which is largely character- 
ised by its epiphytic habit and the waxy nature of the pollen masses, two 
entirely new characters, and in this we have to retrace our steps practically 
‘to the point of contact with the Arethusez, for there are certain genera in 
‘which it is not always easy to say whether the pollen is powdery or waxy, 
and the epiphytic habit is not by any means universal. In fact we find a 
‘steady progression from a lower to a more highly specialised type, as in the 
‘divisions previously considered. Orchids possessing waxy pollen masses are 
for the most part confined to tropical and subtropical regions, and they are 
‘divided by Bentham into two great tribes, Epidendrez and Vande, the 
latter characterised by having the pollinia attached to a distinct stipes and 
‘gland, which is not present in the former. The difference is generally 
obvious, though in a few cases the distinction is not easy to make out with- 
- out careful examination. 
The rostellum of the Vandee is a highly specialised structure. We have 
‘already seen how the rostellum originally arose by a modification of the 
median stigma into a secreting organ, for the purpose of attaching the 
pollinia to the bodies of the fertilising insects. In the Vande the specializa- 
‘tion is carried further, and Darwin has shown that during an early period 
of growth a line of separation is gradually formed, at first appearing as a 
zone of hyaline tissue, which sets free a portion of the rostellum, together 
‘with the viscid disc and the pollen masses, when an insect visits the flower. 
This separable portion of the rostellum, which connects the pollen masses 
‘with the viscus or gland, is called the stipes. It forms a new character, and 
its presence constitutes the essential character of the Vandex. The 
difference is fundamental, yet the difficulties of observation have left the true 
position of a few genera in doubt. Calanthe, for example, is placed by 
‘Darwin in the Vandez, but by Bentham in Epidendrez, the latter pointing 
out that the viscus, which after dehiscence connects the points or caudicles 
of the pollen masses and hardens into.a disc-like gland, is not a scale or 
Jamina detaching itself from the upper surface of the rostellum, as in true 
Wandee, and its affinities in other respects are with the Epidendrez. 
97 
