8 ON THE MORPHOLOGY OF CRUCIFERiE. 
dferm^ and the increased number of stamens in Megacarpceay might 
thus be accounted for, and it seems not unlikely to he the case in that 
genus, as the glandular receptacle on which the stamens are inserted 
(p. 272) probably indicates more stamens, as it undoubtedly does in 
Reseda. 
But why two stamens longer than the single ones on the sides of the 
flower should be produced so regularly in the place of one remains to 
be accounted for, and I think this may be done by a comparison with 
Resedace<je and Capparidetje . On examining the flowers of a species of 
Cleome having four sepals and four petals with six stamens, I found one 
with five sepals and petals all fully developed, but yet it had only six 
stamens. Supposing, then, such a flower to have six sepals and petals, 
it might be assumed that the production of two stamens in place of one 
was the consequence of a tendency in a tetramerous flower to become 
hexamerous ; and as Reseda often has six sepals and petals or more, 
this explanation seems so far practicable. If, then, we suppose there 
exists in Cruciferce the same tendency to become hexamerous that we 
w 
find in Reseda^ the cause of the production of the two long stamens 
in place of one would become apparent. The doubling of the short 
stamens would then be accounted for by a tendency to the production 
of more than six sepals and petals, which occurs in Reseda, or it might 
be the commencement of an increase of stamens like that oi Megacarj)(say 
more probably the latter. 
I cannot conclude these remarks without especially acknowledging 
that whatever truth there may be in the hypothesis now advanced, it 
is entirely owing to the new observations and facts disclosed to us in 
Mr. Smith's very original and valuable paper. 
rently quite opposite, the intemodes being sometimes rather long. Each whorl 
is, however, found to be subtended by a sheatMug leaf, the membranous exten- 
sion of which inckides all the others when very young, while they are quite 
destitute of sheaths, but are otherwise of the same linear form. These sheath- 
ing leaves, if traced up on the stem, prove to be alternate like those of any ordi- 
nary Endogen. In this plant the calyx consists of a very short cup with six 
linear segments, the anterior one being the larger, and having the only stamen 
attached at its baswB ; the style is rather short with three stigmas, the anterior 
one being the longer ; so that if in Monockoria the cells at all communicate, it 
might be referred to that genus. 
