46 
CONTRIBUTIONS TO BOTANY. 
couclusioD, as I stated before, be had probably in view bis tribe 
Icacinea, where in Pogupetalum there is sometimes a similar large 
deshy connective, forming a conspicuous appendage much exceed- 
ing the length of the anthers, and it has also an ovarium of 
several cells with two ovules suspended from near the summit. 
In Stemonurus we find similar glandular cilia upon the filaments. 
In Ptychopetalum we have double the ordinary number of sta- 
mens, and in several genera of the same tribe, we perceive a 
hypogvnous cup, with ten free lobes, investing the base of the 
ovarium. They resemble the Icacinacea also much in habit, 
having similar coriaceous exstipulate leaves, and terminal or axile 
inflorescence of small crowded flowers, each flower being sup- 
ported on an articulated pedicel. Notwithstanding these distant 
indications, the real affinity of the HumiriacecB appears to me to 
be nearest the StyracecB. 
While speaking of the latter family, I will offer a passing ob- 
servation upon the anomalous genus Diclidanthera, placed doubt- 
fully by Von Martins in Ebemcece or Styracece, by Lindley in 
Styracece, by Don in Ebenacetp, and by Endlicher in Styracece. 
Prof. A. DeCandolle in his 'Prodromus’ (vol. viii. p. 245) has 
given several i-easons why it should be excluded from the last- 
mentioned order, but has not assigned to it any other position, 
and since then no other botanist has ventm-ed to indicate its true 
locality in the system. I shall be able to demonstrate that its 
corolla is not gamopetalous, as generally stated; or at least 
that its petals are easily separable from each other, being only 
slightly agglutinated together by the feeble adhesion of the fila- 
ments to them. In its habit, its stipular alternate leaves, the 
linear form of its petals, their mode of aestivation, the stamens 
always double the number of the petals, the valvular and hinge- 
like dehiscence of its anthers, the suspension of a single ovule 
from the summit of each cell of its ovarium, the form and direc- . 
tion of the embryo with large foliaceous cotyledons in fleshy al- 
bumen, Diclidanthera will be seen to approximate closely to the 
Hamamelidacete, from which family it dift'ers only in some slight 
characters, principally in the absence of the fleshy hypogynous 
disk that in Hamamelis and its congeners serves to agglutinate 
the base of the ovarium with the lower portion of the tube of the 
calyx, and which thus renders it semi-inferior. 
Whether for this reason, it will form the tj^e of a distinct 
Order, is a point to be determined when the family of the Ha- 
mamelidacecE has been better investigated ; but its proximity is 
certainly here, and in the meanwhile it may be desirable to 
place it in a separate tribe of that order. 
I shall also be able to add some new evidence confirmatory 
of the observations of Dr. Arnott relative to the structm-e of 
