AND AFFINITIES OF ALANGLE.E. 181 
of the flower in these instances are quite equivalent to those of Marlea 
begonifolia, where the numbers are 5, 6, and 7, the usual number 
being 6. 
That a tendency also exists to an increase of carpels in Marea, is 
shown by an occasional increase taking place even when growing in 
confined circumstances : in one instance there were two additional 
imperfect cells separating the original eells, and in another, three fully- 
formed fertile carpels (Vide Pl. V. A. f. 9). And it is not improbable 
that Alangium has the same tendency ; Rheede, in ‘ Hort. Malabar.,’ 
having described an Alangium with a fruit containing 2 or 3 nuclei, the 
same species being figured as having a stigma of a triangular form. 
The comparison of Marlea to Cornus may be extended also to the 
ovules, which are placed in the same manner in the cells of the ovary, 
viz., the raphe is not next the placenta or axis of the ovary as in pen- 
dulous anatropal ovules generally, but is always situated laterally 
(Vide Pl. V. A. f. 8a, and C. f. 3a). In Cornus, this gives a peculiar 
form to the cell, causing it to be rounded on one side and somewhat 
angular on the other, and this becomes more apparent as growth 
advances. 
In diseussing the afünities of dicarpellary Orders, there remains 
another character which should not be overlooked, and that is, the 
position of the carpels relatively to the axis; and in this also Marlea 
corresponds more nearly with Cornacee than with any of the Orders 
which are regarded as nearly allied to Alangiee. Thus, of 51 ovaries 
of M. begonifolia, the carpels were anterior and posterior in 29, and 
right and left in 22; and of the same number of C. sanguinea the —— — 
carpels were anterior and posterior in 37, and right and left in 14, — — 
carpels having an oblique relation to the axis occurring also in both; — — 
while in two genera of Hamamelidee, three of Myríacez, and one of | 
Onagrariz, all the carpels are anterior and posterior, and in a Brunia, — 
having but one carpel, all anterior. 
In Combretum the single carpel varies in its relation to the axis, 
being indifferently anterior, posterior, or lateral, which seems to indi- 
cate that, if dicarpellary, it would correspond with Marlea and Cornus, 
which is rendered in some degree probable by the single carpel of 
Aucuba, among Cornacee, varying in its relation to the axis nearly in | 
the same manner as that of Combretum; but the valvate corolla and — 
albuminous seeds of Marlea clearly separate it. Begonia ao siis 
Los i 
