354 NOTES ON THE ROOGEE OF KUMAON. 
to that of the petals and sepals; and Colonel Madden, in his description 
of the plant (Proceedings Bot. Soc. Edinb. 1855, p. 43), says that the 
stamens are “disposed in two or four sets." But upon a careful re- 
examination of a number of flowers, I cannot discover any such ar- 
rangement. The stamens, especially when numerous (never however 
more than sixteen in any flower I have opened), are crowded into a 
tuft surrounding the ovary, so that some three or four appear to be ex- 
ternal, sometimes one opposite a petal, sometimes two side by side, but 
they are so dense that one can never say that two are nearer together 
than to the adjoining ones, and no one is really withinside another a£ 
ihe base. When the stamens are detached (and they fall off with the 
greatest facility when fresh), their scars form a single, irregularly waved 
line, at some distance from the ovary, and surrounded by a slightly 
glandular ring, waved and indented by the cavity left by each filament. 
This arrangement is particularly evident after the flower is fully ex- 
panded, and the filaments have more room to assume their natural 
position. To me therefore it is clear that the whole of the stamens 
belong, in this as in other Crucifere, to a single verticil. 
-This view of the case would tend to confirm the most plausible of 
the modern theories of the morphology of Crucifere—that one so clearly 
expounded by Messrs. Webb and Moquin-Tandon in the seventh volume 
of Hooker's ‘London Journal of Botany, and almost simultaneously 
by Dr. Asa Gray, in the first volume of his beautiful * Illustrations of 
the Genera of North American Plants.’ In a review of the latter 
: work, inserted in the first volume of the * Kew Journal of Botany; 
. p. 359, I did indeed object to the word deduplication, as including what 
. is called £ransverse duplication, a principle which, as I then thought and 
Still believe, it has been attempted to carry too far; yet I cannot but 
most cordially agree in the theory of collateral multiplication, as in- 
stanced in the two double stamens of Crucifere, and in the much more 
divided ones of Malvacee.* So also in the Roogee, all the stamens 
appear to me divided collaterally, so that two, three, or more occupy 
the place of a single one. They all would thus have their origin in a 
single verticil, want of space for their development forcing some of them 
* [cannot however go so far as Dr. Gray in 
Maloacee are always opposite to the staminal 
the European Lavateras at least, where 
ed, they surely alternate with the petals, 
the supposition that the petals of 
l leaves, and belong to the same verticil. 
the five staminal leaves may be easily 
