CONTRIBUTIONS TO BOTANY. 
35J) 
52. Spirospermum. 
This genus was founded, in 1806, upon a Madagascar plant, 
by Du Petit-Thouars, who gave a very meagre description of it. 
De Candolle, in 1818, arranged the genus in Menispermacece, in 
his ‘ Systema,^ comprising all the details afforded by Thouars 
within the space of six lines ; and that is all we know of the 
plant since that time. In my prefatory remarks on this order 
{supra, p. 19), not having then seen the plant, I excluded the 
genus from the family, on account of the spiral form of its 
embryo, and upon the following grounds. In every instance 
throughout the Menispermacece I had found the embryo always 
more or less incurved, the degree of its curvature invariably 
corresponding with the extent of excentric growth of the ovary 
and fruit, the cotyledonary end of the embryo being seen inva- 
riably in close proximity to the basal point of attachment of the 
fmit, while the radicular extremity as constantly points to the 
vestige of the deflected style, the latter being generally drawn 
down near to the basal point of attachment : hence, in the most 
extreme cases, the embryo never completes an entire circle ; and 
from the constancy of this feature, it was naturally inferred that 
a spiral embryo could not occur in Menispermacece. A sub- 
sequent examination of the seed convinced me that I was quite 
mistaken in this conclusion, and that Spirospermum offers a very 
anomalous departure from the above-mentioned otherwise uni- 
versal rule. Here, although the radicular end of the embryo 
remains in its normal position, its cotyledonary extremity is not 
directed as usual to the point of attachment of the fruit, but it 
wanders to some uncertain station through a helical channel. 
The putamen contains a single orbicular seed, which is greatly 
flattened and covered by a thin membranaceous integument ; 
from a point on its periphery, just below the persistent style, 
and close to the basal attachment of the putamen, the cell be- 
gins to be intercepted by a thin partition, which curves spirally 
until it terminates in the exact centre of the seed, thus com- 
pleting in its course two and a half gyrations, and the embi’yo 
is found within the spiral cell of the integument, without any 
albumen. This spiral division is, in fact, the condyle, which at 
its commencement is like that seen in Tiliacora, Diploclisia, &c. 
— where, terminating a little beyond the middle of the cell, it 
divides it into a bimarsupial or hippocrepiform pouch ; but in 
Spirospermum this septiform condyle is continued far beyond 
\hat point, in an extremely attenuated state, under the form of 
a spiral coil, which reaches the centre in the manner before de- 
scribed. This septiform line is attached to the two opposite 
inner faces of the putamen, as in the other genera ; and when a 
