262 ON THE GENUS VILLARESIA. 
is seen, corresponding with the furrow in the albumen, which, 
rising from the base of the cell, ascends to near the summit, bear- 
ing on its extremity the remains of the abortive ovules. This is the 
cionosperm above mentioned, elongated and attenuated with the 
growth of the fruit; it is exactly the same as occurs in Santalacee, 
which has a similar albumen impressed with a longitudinal furrow 
on one side, a similar delicate solitary integument without chalaza or 
raphe, and a similar free external thread-like cionosperm. The em- 
bryo i is also like that in Santalacee, and very different from that in Zea- 
cinacee. 
It is therefore extremely unphilosophical to persist in uniting toge- 
ther, in one family, two groups of plants so utterly discordant in every 
respect. M. Baiilon, in his admirable memoir on the Loranthacee 
(Adansonia, iii. 85), fully agrees in this opinion, especially as regards 
the position of Jcacinacee in the Celastral cohort. He enters fully into 
the nature of the structure of the Olacacee as being entirely opposed to 
that of Icacinacee, and maintains (l. c. p. 99) its intimate relationship 
with Loranthacea, and especially towards those genera having a free 
central placenta, such as “ Primulacee, including Ardisiacee, Theo- 
rastee, Myrsinacee, and Aigiceracee.”’ The consideration of this 
affinity, pointed out by me many years ago, would lead us into too wide 
a digression from our present subject. 
The authors of the new ‘Genera Plantarum’ omit all notice of the 
important fact of the constant occurrence of a dorsal raphe in the 
seminal integuments of the Zcacinacez, a development which forms a 
strong connecting link between that family and others of the Celastral 
alliance. There is indeed so close an approximation between Icacinacee 
and Aguifoliacee, that wherever the one is located the other must 
necessarily accompany it, for the only material difference between them 
is a contrary mode of estivation. The occurrence of a dorsal raphe I 
believe to be universal throughout the Celastracee, Hippocratacee, 
Aguifoliacee, and Icacinacea, and to be almost peculiar to them : this 
combined with a marked uniformity in their floral and carpological 
features, consolidates them into one very natural cohort. The line 
of demarcation between them is shown by the following simple rule, 
against which I do not remember a single exception. 
A. Stamens inserted within the disk, 
Ovules erect, with a dorsal raphe. Petals imbrieated . . Hippocratacea. 
