CONTRIBUTIONS TO BOTANY, 
175 
distinct from the hard shell which has been called the testa, and ^ 
which is probably here developed from the secundine, while the 
i nner tegmen proceeds from the tercine. Prof. Agardh, following 
Martins and Endlicher, considers Aristotelia as the type of a 
distinct family ; but I shall he able to show that it differs little 
from Tricuspidaria, Vall(sa, and Elceocarpus, and that Sloanea 
and Dasynema resemble these in all essential features, if we 
except their want of petals through abortion. I cannot therefore 
accord with his opinion concerning Elceocarpece and Styracece, 
that “^gemmularuni positio et forma, fructus, gemmulse paucse 
evolutse, et forma embryonis, in utroque ordine fere eadem 
sunt.^’ Nor can I agree with him concerning the latter family 
in saying, “certe enim nec cum Meliaceis, nee cum Olacineis 
quaedam affinitas.” 
The same authority (Z. c. 270) says of the ovary of Styrax 
officinale, ” plusquam dimidia parte calyci adnatum vidi.” I find 
on the contrary, in its complete fiower, that the ovary is generally 
wholly superior ; sometimes, however, only a small portion of 
having myself watched the actual mode of growth of the ovule, as explained 
in a subsequent article (Ann. Nat. Hist. 3 ser. vol. i. p. 358) : my reasoning 
then would have been correct if the prevaihng theory of the inversion of 
the nucleus upon its centre, owing to the one-sided growth of the coats of 
an anatropous ovule, as taught in all our elementary books, had not been 
quite fallacious, as I am since convinced it is. I cannot, however, agree with 
the distinguished American Professor in considering that the outer fleshy 
tunic and the hard nut which it covers, in the seed of Magnolia, are both 
derived from one common origin (see ante, note, p. 163). Admitting the 
correctness of the facts, as detailed by that learned botanist, relative to the 
progress of growth of the ovule of Magnolia (Linn. Proc. i. 106), especially 
in regard to the period of the deposition of sclerogen in the tunic of the 
ovule, in the manner he relates, it appears to me far more reasonable to 
conclude that the tunic there described is composed of two integuments 
(primine and secundine) agglutinated together — the latter becoming solidi- 
fied subsequently, as he shows — than that we should infer, as he does, that 
sclerogen is thus copiously deposited upon one half of the cellular tissues 
of the fleshy mesoderm, to constitute the nut, while the other half of the 
same tissue retains its lax cellularity — thus forming two seminal coatings 
of very different nature out of a simple ovular integument (see the former 
note, p. 163). We have evidence that such an assumed action cannot 
have occurred, because, in such case, the vessels of the raphe, instead of 
existing (as they are seen) in the outer fleshy tunic of the seed, ought to 
have been found imbedded in the inner bony shell, deposited, according to 
that hvqjothesis, upon the internal face of the primine, where the vessels of 
the raphe are always seen attached. I am the more confirmed in this 
opinion by the instance of Tricuspidaria, which offers a strong case of 
analogy. I have found there, after the ovary has somewhat advanced in 
growth, subsequent to the fall of the corolla, that the primine becomes 
thick and fleshy, and remains quite loose over the firm integument that 
subsequently becomes the nut, being perfectly free from it at all parts of 
its surface, except at its broad chalazal extremity. 
