MR..G. BENTHAM ON INOCARPUS. 147 
varied modifications. In Efaballia it consists of a single carpel 
with a very short terminal but excentric style, and two or three 
amphitropous ovules with a superior micropyle attached to the 
upper or inner angle of the cavity, that is, to the side from which 
the style proceeds. Only one of these ovules comes to maturity ; 
and in one already faded flower I could find only a single ovule, 
but of the usual form and attachment, and not anatropous nor 
pendulous from the summit of the cavity, as described by Endlicher. 
I do not think, however, that this skilful botanist ever examined 
Inocarpus himself. At the time of the publication of his ‘ Genera,’ 
it was rare in herbaria; and the statement that the ovule was 
pendulous from the top of the cavity must have been taken from 
Roxburgh, the only botanist since Rumphius and Forster who 
has described the plant from actual specimens ; and an inspection 
of the rude and certainly incorrect analysis in the plate of the 
Coromandel plants (t. 263) probably induced Endlicher to suppose 
that the ovule must be anatropous. The conversion of the calyx 
and corolla into an outer calycule and a simple perianth is another 
proof that Endlicher’s character was compiled from books; for an 
examination of the plant would have shown him that the two 
lobes of the calyx are not the summits of two united bracts, but 
formed by the cohesion of the normal five teeth of the calyx into 
two or sometimes three lobes, as shown by their venation, and 
sometimes by minute teeth at the apex. 
From these incorrect notions of the structure of the flowers 
which had obtained, it is not to be expected that the place of Ino- 
carpus in the natural system could have been accurately fixed. 
Jussieu, having only Forster’s and Thunberg’s characters to judge 
from, referred it to the “genera Sapotis affinia;" and, as far as 
then known, several technical points appeared to connect it with 
that order. Endlicher, however, studying apparently Roxburgh's 
figure, sought to connect it with Hernandia, in a small group an- 
nexed to Thymeles; and there more recent authors have left it, 
raising however the group to an independent order of Monochla- 
mydeæ, under the name of Hernandiacew. Even Miquel, in his 
‘Flora van Nederlandsch Indië,’ adopts this view, extracting his 
characters from Endlicher, although a slight examination of spe- 
cimens of the two plants, of which he must have had abundance at 
his disposal, would have shown him that they differed as widely in 
most of their essential characters asin habit. In the Prodromus, 
Inocarpus is excluded from Sapotacee on the authority of End- 
licher; and neither that genus nor Hernandia is alluded to 
L2 
