148 
CONTRIBUTIONS TO BOTANY. 
line of the raphe to the peripherical margin of the funicular 
plate. It would certainly be very incorrect to call this fleshy 
coating a testa : indeed it was for a similar reason that I pro- 
posed for the analogous coating in Clusia and Magnolia the 
term “arilline/’ in order to distinguish it from another more 
internal testaceous tunic, which is destitute of vessels of any kind. 
In flne, any one who will take the trouble to reflect on the con- 
secutive changes that must take place in the development of the 
seed, and will keep his attention flxed upon the continuity and 
movement of the coats of the ovule during their excentric deve- 
lopment and subsequent growth, must perceive how absolutely 
impossible it is that the external shell, circumstanced as we And 
it in this instance, can have derived its origin from the mere 
increment of the primine. 
If we examine this external shell, we find it to be a thin, 
black, chartaceous or coriaceous entire crust, easily separable 
from the rest of the seed, owing to the deposition between it 
and the proper seminal integuments of a quantity of loose cel- 
lular tissue, mixed with coloured glands, which adhere to one 
as well as the other, and give to both a furfuraceous appearance. 
The coating which immediately invests the entire surface of the 
albumen in all its curvatures is composed of two distinct tunics, 
which, though adhering together, are separable. The outer one 
is soft, opake, and somewhat fleshy, containing numerous di- 
stinct cells, apparently filled with oily matter ; it is much thicker 
around the region of the large hollow raphe : the inner integu- 
ment is thin and membranaeeous, and closely invests the albu- 
men : both are also thickened around the chalazal point, which is 
situated in the lower part of the cavity near the site of the embryo. 
It is said that in some of the Lardizahalaceae {Akebia and 
Hollbcellia) the ovules and seeds are partly imbedded in the fleshy 
walls of the pericarp ; but in Lardizabala the reverse of this takes 
place ; for here there is a protrusion of a part of the placenta, form- 
ing an expanded funicular support, about which the ovule bends 
itself, and becomes peltately convex around it, — the whole, as I 
have above shown, becoming subsequently enveloped by a com- 
plete arillus emanating probably from a growth of the placenta, 
or of its funicular extension. A metamorphosis of an analogous 
kind, that is to say, an excentric replicature of the ovule round 
its funicular support, occurs in the seeds of the tribes Hetero- 
cliniecB and Anomospermece of the Menisperniacece, only that in 
these cases there is no extraneous growth of any enveloping aril- 
Iseform tunic, such as we find in Lardizabala : in the former cases, 
where the.carpels are always unilocular and uniovular, and where 
the growth of the pericarp is constantly upon its dorsal face, 
while that of the ventral face is more or less stationary, it is 
