CONTRIBUTIONS TO BOTANY. 
5 
disposition of floral envelopes as in the male; and there is some- 
times a similar number of sterile stamens around the ovaria, 
but in most instances they are altogether wanting. In the 
centre of the flower the torus rises more or less in a cylindrical 
form, to the sides of which the sterile stamens, when present, 
are attached; they are generally free froip one another, but 
are more rarely attached at their base by a short ring that 
surrounds the more elevated carpophorum. This latter, in some 
few cases, bears on its summit only a single ovary ; but most 
generally it carries three distinct ovaries, occasionally four, five, 
or six, or rarely as many as twelve, arranged in a single whorl. 
These ovaries are generally sessile, but are sometimes borne each 
upon a stipitate support, that lengthens considerably with the 
growth of the fruit. The ovary is unilocular in every instance 
that has fallen under my observation, and never contains more 
than a single ovule — a character which forms a valid line of 
distinction between this order and the Lardizabalacea, Schizan- 
dracece, and Winteracea. 
The growth of the ovary and the development of the ovule, 
together with the changes produced in the structure and form 
of the fruit, present excellent and constant characters, that have 
not been sufficiently attended to. St. Hilaire was the first 
botanist who devoted any consideration to the subject, when, in 
describing a species of Cissampelos (PI. Us. tab. 35), he gave a 
detailed account of this growth, from the period of the impreg- 
nation of the ovule to the final perfection of the fruit. Accord- 
ing to his view, the ovary, by its excentric growth, gradually 
curves itself round in the form of a horseshoe, until the two 
sides thus bent round touch one another, when they become 
agglutinated together {se soudent) : it thus assumes an ovoid or 
subglobular form, and the original apex, indicated by the style, 
is thus approximated to the base, the two being separated by 
the septum thus formed, which extends far into the cell, and 
which is generated by the “ deux portions rapprochees et soudees 
du pericarpe.” The cell, and consequently the seed, thus assume 
a corresponding hippocrepical shape. 
This view, not altogether correct as far as regards Cissampelos, 
wholly fails to explain the changes attendant on the development 
of the fruit in other cases. Although the ovule, in an early 
stage, is simply anatropal and attached to the ventral face, at a 
point somewhat above its middle from the summit, of a linear 
placenta on the inner angle of the cell, there is always seen 
upon the corresponding concave margin of the ovule, below the 
point of its suspension, a thickened and somewhat curved rib, 
which is probably the indication of the raphe and chalaza : the 
ovule is now partly free from the placentiferous angle of the 
