70 MR. BENTHAM ON CARYOPHYLLE X. 
feia, Lepyrodiclis, Odontostemma, Honckeneya, and Merckia have 
been founded are chiefly, —unisexuality, absence of petals, a great 
development of the glands of the disk, the division of the ovary 
into 3 cells, the complete separation of the capsular valves to the 
base, or a reduction in the number of stamens, styles, or ovules. 
Unisexuality, or rather polygamy, has been mueh relied on for 
the genus Honchkeneya, made for the A. peploides. But although 
in Europe and Asia the plant is certainly most frequently uni- 
sexual, yet in America it is generally, and according to A. Gray 
universally, hermaphrodite, without there being any other distinc- 
tion between the two races. 
The absence of petals induced Linn:eus to separate Cherleria and 
Minuartia; but this absence has been since shown not to be con- 
stant, and other species closely allied to the one or to the other 
in habit have very minute petals, which again, through other 
species, pass into those of more conspicuous size ; and accordingly 
Fenzl and others have already united these two genera with 
Alsine. 
The glands of the disk are more or less developed in many Are- 
naréas, but are only taken as generic characters as being specially 
prominent in Cherleria and in Honckeneya, two species which have 
nothing else to connect them but what is common to the whole 
genus Arenaria. 
The division of the ovary into three cells is relied upon for Dolo- 
phragma, Honckeneya,and Merckia ; and if it were constant and per- 
sistent, and only to be seen in the two latter, it might be made use 
of to separate them, as they have also in common a larger, almost 
succulent globular capsule, and some affinity in habit. But the 
dissepiments are only to be found at a very early stage ; they are 
always very thin and slender, and have generally disappeared by 
the time the flower has expanded, and I at least have never found 
any remains of them when the capsule is ripe. This division into 
cells can therefore only be regarded as rudimentary ; it may be 
traced here and there throughout Caryophyllez, and, as in other 
cases of undeveloped rudimentary organs, no further systematic 
value can be attributed to it than as an indication of the normal 
type, of which the Caryophyllaceous ovary is a modification. 
The valves of the capsule separate to the base and spread out 
horizontally in Zriplateia. But the depth to which the capsule 
splits is very variable in the whole genus ; and, in the single Tri- 
plateia known, there is nothing marked in habit to distinguish it 
from some Mehringias, whilst any other exceptional characters 
