Vll 
exterior of the floral envelopes, not only because in Epistephium there is 
an obscurely three-lobed cup still more external, but especially because 
in Bolbophyllum braeteolatuin there are two scales or rudimentary sepals 
alternate with the so called sepals on the outside. That the supposed 
petals are sterile stamens is rendered probable, 1°. by the general ten¬ 
dency of one of them to become three-lobed, in which case the middle 
lobe may be regarded as equivalent to the connective, and the lateral 
lobes to the cells of the anther; 2°. by the glandular termination of 
these parts in Megaclinium falcatum and elsewhere, a common sign 
among other plants of the suppression of an anther, and thirdly because 
they have all been seen transformed into stamens by His and Brown. It 
is however remarkable, that the last-mentioned botanist does not con¬ 
sider such monstrous cases as good evidence of the so called petals being 
modified stamens, and that he regards the additional part in Epistephium 
to be analogous to what is termed a calyculus in some Santalacese, a few 
Proteacese, and perhaps Loranthace<e. It might however be argued 
that the so called calyx of those orders is really corolla. 
The centre of the flower is occupied by a body called the column, 
which is formed by the consolidation of the style and true stamens. In 
the greater part of the order there is but one stamen present, and it is 
in that case placed opposite the intermediate sepal, and consequently 
alternate with the lateral petals; when, as in Cypripedium there are two 
stamens present, then the usual stamen remains in its customary posi¬ 
tion, in a sterile state, and the two perfect stamens are additional and 
placed right and left of it. It is supposed that in those species which 
have but one anther there are two other stamens present in an incom¬ 
plete condition, and consolidated with the other ; and from the evidence 
offered by monstrous formations, it is thought that such sterile stamens 
are represented in Orchis and its allies by two tubercles, one on each 
side of the column, in Burlingtonia by two auricles near the apex of the 
column, and by other signs in other cases. This theory is in some 
degree confirmed by the genus Zygostates, in which what appear to be 
the sterile stamens are present in a perfectly distinct state, while there 
are no lateral appendages of any kind upon the column. But on the 
other hand it is invalidated by a monster of Platanthera bifolia, 
described by l)r. Brown, in which three stamens were developed, each 
