CONTRIBUTIONS TO BOTANY. 
7 
apparently confounded, they are distinct from them, as shown 
when the “ tubillus ” is drawn away from the corolla. 
There is a peculiarity of structure in Nastanthus and Ammo- 
carpus, and, in a less degree, in most genera of the family, 
which is most distinctly visible in the polliniferous flowers of 
the two genera just mentioned : — the tube of the corolla and the 
segments of its border appear to consist of two distinct parallel 
laminse, with a vacant space between them, as if the entire sub- 
stance of a very thick mesoderm had disappeared by absorption 
or desiccation, leaving only a small quantity of cellular tissue 
consolidated in the apex of the segments, under the form of a 
glandular callus, which apical callus is a constant feature through- 
out the family. This separation of the two surfaces is greater 
in the segments than in the tube ; the outer face is very convex, 
forming a prominent gibbous hood about the inner face, which 
is nearly flat : these two laminse are perfectly united at the edges 
of the segments ; the outer one is very thin, hyaline, reticulated, 
and contains no nervures ; the inner one is more opake, coloured, 
and is furnished with the nervures above described. This struc- 
ture is confii’med by the observation of Dr. Philippi (Linnsea, 
xxviii. 709) in his diagnosis of Boopis {Nastanthus) Guyana and 
of Boopis (Nastanthus) compact a, wdierein he describes the corolla 
as being “ breviter subcylindrica, membrana externa hyalina ab 
interiori viiidi remota.’’ 
Another character of the Calyceracea, which serves to distin- 
guish this family from the Composite, is deserving of some no- 
tice. Throughout the latter order, the style is bitid at its apex, 
and each branch is furnished towards its extremity wuth a stig- 
matic surface, and frequently also with collecting hairs, that 
assist in the transmission of the pollinic influence. On the con- 
trary, in Calyceracea the style is undivided, clavate, and solid at 
the extremity, and, though here covered with a rugose surface, 
is quite deficient of any collecting hairs. Although the ovary 
in both cases is 1 -locular, the inference may be drawm from the 
above circumstances, that the normal condition of the ovary in 
the one case is to be 2-ovular, and in the other 1-ovular; and 
though we have no positive proof of this conclusion, many cir- 
cumstances tend to favour the opinion of the biovular tendency 
of the ovary in Composite. The placentary point of attachment 
of the solitary erect ovule is always upon one side of the base of 
the cell ; and hence it may be assumed that, as there are two 
stigmata, another placentary point normally existed, which has 
been suppressed * ; this idea is again confirmed by the fact that 
in many of the achsenia of Composites two parallel grooves or 
* A similar view has been advocated by Mr. B, Clarke (Ann. Nat. Hist. 
2 ser. xi. p. 45(1). 
