OF THE RAPHE IN SEEDS. 109 
at the point where, according to the laws of structure, the chalaza ought to be found in 
all cases where a raphe exists: it is precisely in the situation ascribed to the chalaza by 
Mirbel, in his figure before quoted of the ovule, after the formation of the embryo. It is 
however only a false chalaza, and is most probably a thickening of the mouth of the 
micropyle (exostome of Mirbel) of the ovular integuments, a circumstance of frequent 
occurrence in certain seeds; or it may possibly be ascribed to a scar of the embryo-sac, at 
the point where it has broken away from the gangylode, as figured by Mirbel (loc. cit. 
fig. 10. letter f). That it cannot possibly be the true chalaza, is most evident, because 
it has had no apparent connexion with the outer crustaceous tunie, which exhibits no 
corresponding scar or trace of any former connexion at this point; and still more obvi- 
ously, because the raphe passes straight across it, without holding the slightest commu- 
nication with it. The embryo, which has no albumen, fills the whole cavity of the inte- 
guments in the ripe seed, the radicle being directed towards the hilum and to the 
terminating point of the raphe, and its extremity being covered by the true chalaza; its 
cotyledonary extremity is in the opposite direction, terminating at the areolar micropyle 
or false chalaza. The fact of the existence of the free cord of spiral vessels as above- 
mentioned (the raphe), running between the outer crustaceous tunic and the inner 
integuments of the seed, was evidently known by Mr. Griffith, who notices it (Posth. 
Notes, p. 182); but he omitted to observe the passage of this thread over the areole, 
which has always been considered as the chalaza, and also its course round the entire 
periphery, as well as its termination at the radicular extremity of the membranaceous 
integuments : the presence of the true chalaza at this point also escaped his observation. 
I do not find any mention of this curious structure by any other botanist. 
We thus observe the same phænomena existing in all the genera of the Queurbitacee 
that I have described as being found in Sfemonurus, Anona, and Qitrosma, that is to say, 
where the cord of the raphe makes a complete circuit of the seed, terminating at the same 
time at the radicular point of the integuments, instead of the cotyledonary extremity, as 
it ought infallibly to do, according to the acknowledged laws of structural development. 
What has been the nature of the metamorphoses within the ovule, that have produced so 
manifest a deviation from the ordinary course of structure? We have it demonstrated 
on the high authority of Mirbel, that up to the period of the impregnation of the ovule, 
the raphe, the chalaza, and the primine and secundine (then agglutinated into one coat), 
had only performed a simple anatropal inversion from their normal position; but we find 
that in the interval between this period and the ripening of the seed, all these parts have 
experienced a farther circumversion, so as to complete an entire circle, while the embryo- 
sac, or at least the embryo, has retained the same position which it held at the period of the 
simple anatropal inversion of the ovule. Mirbel notices, at this last-mentioned period, the 
first indication of the growth of the arillæform coating over the BEE Seton) at 
this epoch as two layers of cellular tissue coating that integument, and as pit i 
(loc. cit. fig. 10. letter 4): these layers of tissue are unquestionably the ru ; p d 
secretions that subsequently form the arillus, under the crustaceous cip a k , 
for which it has always hitherto been mistaken : it is doubtlessly owing a t a Lars am 
of this coating that the subsequent movement of the coats of the ovule has been hi 
