clined to believe that it has also undergone a rather sud- 
den expansion during post Pleistocene time. 
The manner through which the five main groups ac- 
quired their characteristics is impossible to determine 
without fossil evidence. The presence of primitive and 
advanced characters in the same species may, however, 
be an indication of fusion of independent evolutionary 
trends which very probably date back to a remote past, 
since cytological investigations and genetical experi- 
ments have demonstrated an absolute incompatibility 
of the groups. Whether or not this statement can be 
upheld without alteration in the future is hard to predict, 
inasmuch as the presently available information is very 
meager and unsatisfactory. Unfortunately, with respect 
to experimental genetics, only successful crosses are re- 
ported, while those which failed to “‘take’’ are carefully 
guarded from public possession and scientific scrutiny. 
On Plate XIV, I have shown the results of a few 
crosses between Cypripedioideae and Kerosphaeroideae. 
Also I have on file certain data about attempted crosses 
between Disa uniflora (Neottioideae) and members of 
other groups, but since | am unable to ascertain the 
source of this data, I have omitted them in the diagram. 
The fact that the pollen is apparently unable to induce 
even parthenogenesis in the reported crosses between 
Cypripedioideae and Kerosphaeroideae may be a further 
indication of a remote origin of the respective groups. 
The cytological coverage of the family is exceedingly 
poor. The known percentage of the chromosome counts 
of the species in each group is shown on Plate XIV. 
Cypripedioideae has a 50% coverage, due to the exten- 
sive studies in the genus Paphiopedilum by Duncan. The 
next highest is Ophrydoideae, a group almost exclusively 
of temperate regions, with 8% coverage based for the 
most part on European and Japanese species. Informa- 
[ 82] 
