rived one, and several attempts have been made by sun- 
dry workers to visualize its prototype as having three 
free segments in each of the five whorls. This arrangement 
by itself suggests but one course of evolution: progressive 
differentiation in a linear sequence. The evolutionary 
makeup of the Orchid family, on the contrary, is indica- 
tive of a rather complicated origin, because in certain 
species characters of a primitive and advanced nature 
occur simultaneously. <A striking example is Vanilla 
(Neottioideae). The plant itself starts out as a terrestrial 
(primitive), but soon climbs up to the tree-tops where it 
leads an epiphytic mode of life (advanced); its ovary has 
a parietal placentation (advanced) yet the seeds possess 
a heavy sclerotic seed coat (primitive) and the embryo is 
sometimes known to develop an endosperm (primitive). 
Vanilla itself may be considered an advanced type with 
respect to other members of the Neottioideae. 
The association of primitive with advanced characters, 
which is to be found in each of the five groups, speaks 
rather clearly in favor of considering the Orchid family 
as one of those groups established relatively early during 
the evolution of the Angiosperms. Stebbins places its 
origin in the early Cretaceous Epoch and, on the basis of 
other evidence, I am inclined to agree with him. My 
studies of distribution, for example, based on deductive 
reasoning, lead to the same conclusion. The presence of 
the phenomenon termed pseudocopulation, noted in the 
Ophrydoideae and Neottioideae, may be mentioned here 
as further supporting evidence of age. If the Orchid 
family has evolved relatively recently, as some scientists 
postulate, such a precarious adjustment between the flow- 
ering time and the emergence of the fertilizing agents, as 
we find in the case of pseudocopulation, could hardly have 
come about. Nevertheless, I consider that the Orchid 
family is still in a state of evolutionary flux, and am in- 
[ 81 | 
