340 
CONTRIBUTIONS TO BOTANY. 
The leaves are 2|-3|^ inches long, Ij-lf inch broad, on a 
petiole 4-5 lines long. 
48, SciADOTENIA. 
This very interesting genus was proposed by me in 1851, the 
typical species being a Cayenne plant, remarkable for the pecu- 
liar development of its fruit. The plant has much the habit of 
Elissarrhena or Anelasma, having large polished leaves with five 
elongated nerves springing from the base, which form only a 
small angle with the midrib and are therefore somewhat parallel, 
all arcuately anastomosing with two superior pairs of shorter 
lateral nerves, and all connected by many straight transversely 
parallel veins. The petiole is slender and not exceeding one-fourth 
or one-fifth of the length of the leaf. The inflorescence is axil- 
lary and borne upon a single very elongated peduncle, as long as, 
or longer than, the leaf. At the period above mentioned I had 
seen only a solitary specimen, in which all the floral envelopes 
had fallen away, leaving no indication of their position ; the 
summit of the peduncle bore nine immature drupes, each sup- 
ported by its own long apparent pedicel ; so that I then natu- 
rally concluded that the inflorescence was umbellate, each pedicel 
bearing its fruit. Shortly after the printing of my paper, I met 
with another specimen with more mature seeds ; and in one of 
its axils I observed, to my surprise, eight ovaria sessile on the 
summit of the peduncle, and a single seed borne upon a length- 
ened pedicel-like support, as in the case before mentioned : this 
at once afforded a key to the real structure, making it .evident 
that the supposed umbel is a development proceeding from a 
single flower. Ten years subsequently Mr. Bentham noticed 
the same fact (Proc. Linn. Soc. v. Suppl. 51), and explained this 
development as a separate growth of each ovary, considering 
each support to be a '^podocarp” — that is to say, an expansion 
of the fruit or ovary. But this is not the case : its true nature 
was shown in my explanation of the analogous growth in 
Tiliacora {supra, p. 74), where the clavate torus spreads into 
many short forks, which vary from three to twelve, according to 
the number of drupes perfected. In that genus these forks do 
not exceed the length of the torus (2 lines) ; and they are not 
united at the base, as in the very elongated carpophora of Sciado- 
tenia, which are nearly an inch long, in shape like a 4-angular, 
8-grooved puherulous pedicel, and united at the base inside by 
a membranaceous web, leaving a hollow space in the centre. 
The number of these supports corresponds with that of the 
ovaries perfected ; they are prolonged in the ratio of the growth 
of the fruit, and are wholly suppressed whenever the ovary is 
