As was stated earlier, 4 postasioideae is considered by 
most recent phylogeneticists to represent a distinct fam- 
ily and has consequently been removed to distant alli- 
ances, such as Haemodorales and Liliales by Hutchinson 
and Takthajan respectively. It is obvious that if Apos- 
tasioideae is removed from the Orchid family, this pro- 
cedure automatically demands also the separation of 
Cypripedioideae, as was suggested by Mansfeld, because 
of the absence of pollinia formation, different fertile stam- 
inal circles, ete. If A postasioideae and Cypripedioideae are 
removed from this interrelated complex, the remaining 
eroups, Neottioideae, Ophrydoideae and Kerosphaeroid- 
eae, would also require a new family status, because the 
equilibrium between these five groups is destroyed. On 
the other hand, the relationship between Cypripedium 
(Cypripedioideae) and Cephalanthera (Neottioideae) is so 
close that to regard them as members of distinct families 
would defy our whole evolutionary approach to syste- 
matics. 
When we study the origin and phylogeny of the Or- 
chid family devoid of paleobotanical documentation, our 
analysis is strictly limited to the uncovering of primitive 
features in living species. Since our approach is based 
a priori on such terms as genus and subfamily, these 
higher categories will aid our investigation only if they 
represent expanded dynamic units, although it is almost 
impossible to visualize the occurrence of such units in 
nature. The various evolutionary forces which have 
shaped and brought about the present-day orchids ob- 
viously did not act upon the family or even on a given 
genus, but rather on the species, because a species is the 
only tangible unit in nature with potentialities to mutate 
or evolve. Therefore, it would not be surprising if an 
absolute delimitation of a family becomes impossible. 
The structure of an orchid flower is definitively a de- 
[ 80 ] 
