360 FLORA AMERIC/E BOREALI-ORIENTALIS. 
staminal anthers.* A split stamen, bearing one cell of the anther on 
each branch, may be compared to a bilobed leaf; the double stamen of 
Cruciferae, each bearing a two-celled anther, to a bifoliolate sessile leaf ; 
the treble stamina of Fumariacee to ternately digitate leaves; and all 
other cases of what is termed collateral deduplication may thus be 
referred to the ordinary ramifications of leaves, without the necessity 
of the creation of a new term to explain them 
e so-called transverse deduplication is evidently a very different 
process, and, if really such as it is supposed to be, alone deserves the 
name of deduplication; but, being totally at a loss to find anything 
analogous in the ordinary mise we have endeavoured to explain 
it by other more normal pro 
There is no doubt that in icu ease of many polyandrous flowers, 
such as Hypericum, several Tiliacee, and probably a considerable 
number of Mimosee and Swartziee, a number of stamens occupy the 
place of one, and that in Malvacez, as so clearly explained in the work 
before us, the stamens and petals together occupy that of the petals 
only ; but might not this be better explained by the Heath-like develop- 
ment of a fascicle of leaves in the axille of the petals in Malvaceae, and 
in those of the staminal leaves in other cases, the subtending leaves 
being represented by the sterile stamens in Luhea, and by the large 
outer stamens in Mollia and Swartzia ? 
A greater difficulty may occur in the case of Rhamnee and Bytt- 
neriacee, where Dr. Gray is undoubtedly right in considering the 
stamens as belonging to the corolline verticil; but in this case it 
appears to us quite as conformable to the ordinary course to consider 
the stamen as an axillary production, as to resort to a theory which has 
no analogy in stem-leaves. 
With regard to the inner appendages of the petals of Ranunculacee, 
Caryophyllee, Sapindacee, and so many Gamopetale, upon which the 
theory of deduplication has been chiefly based, it appears to us that 
they are in most cases, if not always, deformed glands; their gradual 
passage into anthers in some flowers, their position on the principal 
veins of the petals, and, in some cases, their real glandular nature, 
tending to confirm the supposition. 
* The correspondence of the anthers to the glands of leaves would receive still 
farther confirmation, if it be true that in — fertilization is effected by a 
viscid fluid exuded from the glands of the bract 
