OF THE RAPHE IN SEEDS. 101 
the nature of the two thin membranaceous plates, or cellular strata, that line the cavity 
of the albumen, as before described, and I think these may be referred, with little doubt, 
as we shall see in other well-authenticated cases, to the existence of a large persistent 
embryo-sac which has not been absorbed, as it usually is in most seeds. In the case in 
question, at the upper extremity of this cavity, the embryo only one quarter of its length 
is found, and here the radicle, pointing upwards, terminates in a long mucronate point, 
which is a kind of suspensor, by which it is pendent in the mouth of the cavity, and where 
it is enveloped in a small quantity of glutinous matter, which lines the internal surface of 
the sac. Instances of the existence of this cavity in the axis of albuminous seeds are not 
unfrequent, but the discovery that it originated in the persistence of the embryo-sac is 
due to Mr. Brown (App. King’s Voy. p. 21). 
The first consideration that arises is the question whether the existence of this persistent 
embryo-sac has any necessary connexion with, or has had any share in producing the phæ- 
nomenon of the retroversion of the raphe, and the inversion of the embryo in regard to it, 
in the manner before stated; whether the two latter unusual circumstances are independent 
of each other, or whether they are necessarily connected. The only method of forming 
any conjecture on this head, is to search for facts that have any collateral bearing on the 
question. In Diospyros we meet with a parallel case of a large cavity in the middle of 
an albumen, which is open at the hilum, and is in like manner lined with an adherent 
embryo-sac; the embryo is found in the outer end of the much larger cavity, the end of 
the radicle almost protruding out of its mouth: the albumen is deeply divided into 
numerous lamelliform folds, as in Amona, and is covered by two thin integuments, of 
which the inner one is plicated within the interstices of its lamellar clefts, while a distinct 
raphe, adhering to the outer integument, proceeds along a groove formed in one of its 
marginal sides, from the basal hilum to the summit, or opposite cotyledonary extremity, 
where it disappears. We have here therefore a parallel instance of the existence of an 
open cavity in the albumen, lined with the embryo-sac, but this circumstance is accom- 
panied by the presence of an ordinary raphe, while both the seminal tunics and the 
embryo have gone through the usual process of anatropal inversion. 
In Anona we meet with a complete peripherical raphe, as in Stemonurus, but the 
embryo-sac has disappeared; there is therefore no cavity, as the embryo lies tightly 
imbedded in the substance of the albumen, and the radicular end of the embryo, with a 
very short mucronate point, or suspensor, is directed to the spot where the raphe, after 
completing an entire circuit round the periphery of the seed, disappears close to the hilum. 
as indicated by 
In this instance we find that a complete gyration of the original tunics, 
the raphe, and only a semi-gyration of the embryo, have taken place ; pam m a 
perfect analogy with the structure of Stemonurus, excepting the presence en em - 
sac and its accompanying cavity. These two instances seem to prove t : e presen 
of this albuminous cavity and the complete gyration of the raphe do no — aá 
accompany each other, and they leave us as much in the dark as ever, regarding the ca 
of the phænomena observed in the seed of Stemonurus. 
In Opuntia we have another example of a completely annular raphe, på — 
imbedded in a small quantity of albumen, is quite campulitropal : the raphe 
