REVISION OF DRYOBALANOPS AND DIPTEROCARPUS. 99 
lignosa, toro inserta, eoritg Siro acuta, valvis tribus 
parte superiore adprimum dehiscens. Sentinia fere unici 
cotyledones ae valde iincatiilen, involute et ovarii 
parietum reliquiis tanquam columellz e tori fundo orte 
a tee. 
are comparatively few'i in nu oe there is a "tendency towards their 
becoming connate, and in soc din ote this is ae tise so far that they 
Out of this really very simple structure De on ~—. in his 
“Mémoire,” an elaborate arra quae sitet hich the 
Candctle ts sees this curious soe although he inserts paren- 
e obvious explanation with a note of interrogation.} 
Balen on hus probably also koe led by De Vriese to state that “les 
. sont libres,” I can, however, see no reason why 
a should describe them (p. 213) as “‘  -seriata.” Bentham and Hooker 
are silent upon these points. 
De Vriese also figures the ogg with the valves equal. Asi 
some other pair 0 genera, such as Shorea and Vatica, the 
inner valves really con ndiderably shorter than the outer. This is 
clearly hows in Y Baillon? s figure (l.c., p. 203) ; but he, in common with 
eh the other authorities, makes the ‘valves equal in his description of 
genus, 
The m ture fruit has been made the subject of a rather elaborate 
paper fy Guile mans.§ He exposes the error of De Vrie 
ting to Dryobalanops a perispe rm, and accepts the obvilies explanation 
of Korthale that the so-called ‘‘ columella” is the residue of the rm 
of the fruit (see fig. 4). A eer structure may be found in Frar 
and in the immature fruit of Fagus. The peculiarity here is that “the 
* Hook. Journ., 1c., p. 38. + Prodr., xvi., pt. 2, 6 
{ Hist. des Plantes, iv., 203.  § Ann. des Sc. Nat., ey sér. v., 1856, 
n2 
