COXTllIBUTIONS TO BOTANY. 
45 
able under careful examination : the suspension of the ovules is 
from a short and free central placenta that rises above the ab- 
solute union of the dissepiments, in the very top of the ovarium, 
and the partitions about this spot, though they all converge as 
far as, and even touch the central placentai-y column, are yet 
really free from it. Although the five ceils are here completely 
established throughout the entire length of the ovarium, still for 
a very short distance at the apex, there exists a communication 
between the cells, through the almost imperceptible chinks exist- 
ing around the margins of such apical portions of the dissepiments 
as are really disconnected with, although touehing the placentary 
column. The definition of Ad. Jussieu* of ‘Moculis ad apicem 
inter se perviis,^^ though quite true, is not a correct expression 
of the structure of the ovarium of Humirium. This offers much 
support to the views of Mr. Bentham in regard to the affinity of 
the Humiriacea with the Olacacece, but in other respects there 
seems less relation between the two families; for if in the ovarium 
of Humirium, the confluence of the dissepiments had been com- 
pleted about the central placentary column to the very summit, 
the position of the Humiriaceee in the system would have been 
close to the Aquifoliacea. From the structure of the ovarium 
and other characters, the conclusion now appears to me irresist- 
ible, that wherever the Stijracea may be stationed in any natural 
arrangement, the Humiriacea must be placed in contiguity with 
them. The Humiriacece will therefore form one of those osculant 
relations, existing everywhere in nature, which can only be re- 
presented by the circular system, and never by any linear 
arrangement; in the former method the Cionosperma would 
touch the Dryales, through Humirium, while in the linear system 
their location will fall to a distance. The position of the Humi- 
riacecR has never been satisfactorily determined : Von Martins, 
who first suggested the order in 1826, considered it allied to 
Meliacece, though doubtfully : Jussieu, in entertaining the same 
view, had similar misgivings in regard to this affinity. Dr. 
Bindley, in his ‘ Introduction to Botany,’ held them more nearly 
related to the Aurantiaca, and Endlicher placed them at the 
head of a class called Hesperides, in association with the Meliacece, 
Aurantiacce, &c. Meissner stationed them also in the Hespm-ides 
of Endlicher, adding at the same time to this class, the Olacacece ; 
and finally. Dr. Bindley in his ‘ Vegetable Kingdom’ renounces 
his former views, and fixes them in a most singular association 
with the Ericaceae, Epacridaceae, Monotropacece, &c. ]\Ir. Bentham 
(Binn. Trans, xviii. p. 682) considered “ that among dichlamy- 
deous plants, they come nearest to the Olacineae ; ” but in this 
* Flor. Bras. Merid. A. St. Hilaire, vol. ii. p. 88. 
