¥ 
ON THE GENUS VILLARESIA. 259 
in Icacinacee. The stamens usually exceed the number of ‘petals, and 
when several sterile ones are present, the fertile stamens stand opposite 
the petals, the sterile filaments being indiscriminately or only partially 
opposite or alternate with them. In Ximenia, Heisteria, and Endusa, 
which have 8-10 fertile stamens, 4 or 5 are opposite to that number 
of petals; in Opilia, Strombosia, Anacolosa, Cathedra, Quinchimalium, 
Arjoona, and Myoschilus,* which have only 5 or 6 fertile stamens, they 
are all opposite to as many petals; in Schopfia, where the petals are 
more than half gamopetalous, the stamens stand opposite to the seg- 
ments of the tube ; in Agonandra the fertile stamens are opposite the 
petals, and the sterile stamens alternate with them. Ptychopetalum 
and Athesiandra (the latter appearing to me a good and distinct genus) 
have 8 or 9 fertile stamens, three of them being immediately opposite, 
the rest indiscriminately face the sides of the petals. In Aptandra the 
stamens are monadelphous, the filaments being united into a long 
cylindrical tube. From these facts it will be seen that as regards the 
position of the petals, the number, quality, insertion, or position of the 
stamens, there is an absolute want of any analogy between Olacacee 
and Icacinacee. 
Tt has already been stated that the calyx in Icacinacee remains con- 
stantly unchanged, but in Olacacee it is far otherwise. In Olax and 
Heisteria, where it is free from the disk, it enlarges with the fruit, and 
in the form of a lax membranaceous covering, it encloses the fleshy 
drupe in a greater or less degree. In Liriosma it also increases con- 
siderably, but it becomes confluent with the fruit, which thus assumes 
the appearance of an inferior drupe. In Cathedra it becomes multi- 
plied by fresh developments, so that the fruit is found seated within 
3 or 4 concentrie cups around the free persistent disk. In most other 
cases, the disk being confluent with the ovary, it thickens and en- 
larges with it, converting it into an apparently inferior drupe, some- 
last three genera are generally placed in Santalacee, but they appear to 
o 
The 3 
me truly O/acaceous, having a distinct whorl analogous 
free or adnate to the disk upon which the petals are seated, and which is wanting in 
Brown called 
Santalacee cannot be far separated from Olacacee, although the prevailing system 
of classification, founded on the condition of e calyx, unnaturally places them 
idely apart. i 
8.2 
