238 
CONTRIBUTIONS TO BOTANY. 
ing that the raphe has no connexion whatever with the outer 
crustaceous coating. Beneath the second envelope is a third 
integument, which immediately invests the albumen, and which 
is thin and membranaceous; these two integuments, though 
free, are in immediate contiguity at all parts of their surface, 
except along the middle of the dorsal face, where they become 
separated to form the square hollow channel in the axis above- 
mentioned, one side of which square, with the imbedded raphe, 
is formed by the induplicature of the intermediate integument. 
The embryo is broad and foliaceous, and is covered by a layer 
of fleshy albumen, equal in thickness to the cotyledons, — the 
margins of the whole, together with the enveloping tunics, being 
conduplicated into the hippocrepical form at first described. 
We see here, therefore, the existence of three seminal envelopes, 
very similar in their nature to those described in Colletia, with 
this remarkable diff'erence, that in the latter the crustaceous 
coating is entire, while in Rhamnus it is always incomplete upon 
one side. In Rhamnus catharticus the groove formed by the 
conduplication of the seed is generally in the middle of the 
dorsal face, though sometimes nearly lateral, caused by unequal 
growth, in consequence of the partial abortion of one of the 
carpels ; but in all these cases, the two lines of raphe always 
correspond with the middle of the cotyledons, while in the Col- 
letiecs they constantly run along their margins*. 
The same structure exists in Rhamnus Alaternus ; but on ac- 
count of the small size of its seeds, and the greater tenuity of 
the several integuments, it requires much care to dissect the 
parts. The crustaceous covering is very thin, and adheres more 
firmly to the extremely delicate intermediate integument, and it 
is therefore difficult to separate them without lacerating the 
latter ; but when the separation is made with care, we fin4 the 
ventral portion of the cord imbedded like a white line in that 
distinct integument, and running from the apex to the base : 
the dorsal groove is more shallow, and embraces the other cor- 
responding portion of the raphe and integument, so that, on 
detaching the external crust, the raphe comes away with it, 
tearing itself from the rest of the integument f. 
This structure is repeated, under somewhat varied circum- 
stances, in Rhamnus chlorophoi'us. Dene. : here, however, the 
raphe is lateral, not dorsal ; the cotyledons are not involutely 
inflected, as in the former case, but are simply complicated or 
folded upon one another, so that their midribs stand opposite 
* Analytical details of this structure in Rhamnus catharticus are shown 
in Plate 33 a of this volume, 
t This structure is also given in detail in Plate 33 c. 
