ment of anthesis some irreversible surge of power were 
beginning to operate to cause the ovaries to twist. If a 
raceme is inverted before this power begins to manifest 
itself, resupination is prevented. However, as my obser- 
vations on Goodyera have been confined to just a few 
plants, I feel that this matter is in need of further study, 
although Ziegenspeck in Lebensgeschichte der Bliiten- 
pflanzen Mitteleuropas (1936) p. 81, states that accord- 
ing to recent experiments the prevention of resupination 
by means of the klinostat is successfully done with Or- 
chis (Dactylorchis), Habenaria (Platanthera) and Good- 
yera, only if very young buds are used. It has been stated 
that in Stanhopea, a very remarkable genus of orchids oc- 
curring in the American tropics, the flowers of the pen- 
dulous inflorescence if held erect will completely twist in 
twenty-four hours and assume the position they would 
have occupied had the inflorescence been pendulous. In 
the tropics where epiphytic orchids are not infrequently 
forced into unusual situations when they grow, with the 
racemes pendent, on tree trunks or on the sides of rocks 
or cliffs, I have observed on several occasions that the 
flowers become adjusted to whatever situation they are 
in and the labellum assumes the position that is normal 
for the species. This is true also of compound racemes or 
panicles in which some of the branches are at right angles 
to the main flower-shoot. In this case the ovaries of the 
laterally placed flowers twist through 90 degrees. ‘This 
is true also for the lateral flowers of those species which 
have both terminal and lateral racemes. In the preva- 
lently dimorphic genus Catasetum and in the dimorphic 
genus Cyenoches, the male flowers are frequently borne 
in elongated, drooping racemes; those of Catasetum be- 
ing resupinate, those of Cycnoches being non-resupinate. 
Furthermore, in Catasetum the female flowers are usual- 
ly produced on erect peduncles and are strikingly non- 
[ 159 | 
