24 
CONTRIBUTIONS TO BOTANY. 
raced is stated by most authors to be half inferior, but I have 
observed that at an early stage, and even after the fall of the 
flower, it is quite free, although partly surrounded by the tube 
of the calyx ; and if it become subsequently agglutinated to the 
latter, it is probably only at a late period, as we find to occur in 
Liriosma. 
The Ebenacece, by most botanists, have been held to be closely 
allied to the Styracece, but this does not appear to me quite 
evident. Though placed among Corollijiorce, it appears to me 
that they should rather be arranged among the polypetalous 
groups, for their petals are often quite distinct, or when united, 
cohere so slightly as to be separated by a little force. The sta- 
mens, although sometimes adnate to the corolla, are most gene- 
rally free, or at least originate in a fleshy disk, which sometimes 
assumes the form of a very short h 3 fpogynous tube. In one 
Brazihan species of Diospyros, I have found the albumen in the 
seed to be distinctly ruminated, as in the Anonacece, the embiyo 
having a terete radicle and broad foliaceous cotyledons, much 
resembling in structure that of Monodora. Cargillia, according 
to j\Ir. Brown, a genus ot this family, so nearly approaches the 
AnonacecE, that the typical species was described by Jacquin as 
the Anona microcarpa (Fragm. xl. tab. 44. fig. 7), and by Dunal 
as the Monodora mia'ocarpa. In the Brazilian species of Dios- 
pyros above alluded to, the seeds are imbedded in pulp, and 
covered by a mucilaginous arillus : they are also compressed, 
with a linear, basal, and somewhat lateral umbilicus, forming a 
deep marginal fm’row, into the bottom of which cavity the ex- 
tremity of the radicle subtends, as in several genera of the Ano- 
nacece*. Monotheca and Reptonia, placed in Theophrastece, appeal-, 
from the descriptions given of them, to have little in common 
with that family, and to belong rather to Styracece, if we consider 
the basal placentations, which I have shown to exist in this last- 
mentioned order, as in the Olacacece ; and the approximation of 
these genera to the Anonacece is again confirmed by the rumi- 
nated albumen of the seed of Reptonia. The relation of the 
Ebenacece with the Olacacece was, I believe, first pointed out by 
J ussieu, but few botanists have attended to the suggestion ; from 
the indications just mentioned, it will probably be found, that a 
more fitting position for the Ebenacece in the system exists among 
* A preciselj- similar structure is found in Diospyros CandoUeana, ac- 
cording to Wight’s ‘leones,’ plate 1222, fig. 8 to ll. In several other 
instances in this family, the albumen is depicted in the same work as heing 
distinctly ruminated, so that this may probably be a general character of 
the order, although so remarkable a feature is not noticed in anj- botanical 
work. Gaertner however hints at the fact, but only in one instance out of 
the many species of Diospyros he describes; D. tetrasperma, which has 
its “ albumen radiato-striatum, quasi fibrosiun.” 
