338 
Their flowers, however, will all be found to consist of three outer pieces belonging 
to the calyx, and three inner belonging to the corolla ; and all departure from 
this number, six, depends upon the cohesion of contiguous parts : with the 
solitary exception of Monomeria, in which the lateral petals are entirely abor- 
tive. Sometimes two of the sepals cohere into one, as in certain species of 
Oncidium, and then the calyx has the appearance of consisting of but two se- 
pals ; sometimes the lateral petals are connate with the column, as in Gongora 
and Lepanthes, and then the column appears fm'nished with two wings. In 
nearly the whole order the odd petal, called the lip, arises from the base of the 
column, and is opposite it ; but in the Cape genus Pterygodium, the lip 
sometimes grows from the apex of the column, and sometimes is stalked 
and turned completely over between the fork of the inverted anther, and 
thus seems to belong to the back of the column. Nor is the anther less 
subject to modification, although constant to its place : sometimes it stands 
erect, the line of dehiscence of its lobes being turned towards the lip ; some- 
times it is turned upside down, so that its back regards the lip ; often it is 
prone upon the apex of the column, where a niche is excavated for its recep- 
tion. The poUen is not less curious : now we have it in separate grains, as 
in other plants, but cohering to a meshwork of cellular tissue, which is col- 
lected into a sort of central elastic strap ; now the granules cohere in small 
angular indefinite masses, and the central elastic strap becomes more apparent, 
becomes attached to a glandular process of the stigma which is often en- 
closed in a peculiar pouch especially destined for its protection ; again, the 
poUen combines into larger masses, which are definite in number, and attached 
to another modification of the elastic strap ; and finally a complete union of 
the pollen takes place, in solid waxy masses, without any distinct trace of this 
central elastic tissue. Such is a part of the singularities of Orchideous plants, 
and upon these the distinctions of their tribes and genera are naturally founded. 
Whoever studies them must bear in mind that their fi'uctification is always re- 
ducible to 3 sepals, 3 petals, a column consisting of 3 stamens grown firmly 
to one another, and to a single style and stigma ; and, with this view, he 'will 
have no difficulty in understanding the organisation of even the most anoma- 
lous Cape species. For many curious, and, for systematic purposes most 
important, remarks upon this order, see Brown’s observations as above 
quoted. 
If the following diagram be compared with those employed to illustrate 
the distinctions of Marantacese and Zingiberaceae, p. 326, the relation borne to 
those orders by Orchidaceae will be distinctly seen. In the diagram the parts 
are arranged as they are in nature before the ovary twists ; that is, with the 
lip next the axis, or uppermost, and the stamen undermost. Let C, C, C re- 
present the outer series of fioral envelopes or calyx, and PP, P, P the inner, or 
corolla, of which PP is the labellum : then the position of the single fertile 
stamen wiU be at S, and the sterile ones at s, s ; that is to say, in the situation 
of the supernumerary petaloid stamens of Zingiberacese and Marantacese, while 
the second series of stamens, to which the fertile stamen of these orders be- 
longs, is not developed in Orchidacese. 
