122 ILLUSTRATIONS OF 
embryo and superior radicle ; but they differ in the estivation 
of the corolla, in possessing five, instead of four stamens, in their 
anthers being two-celled, with longitudinal dehiscence, in their 
leaves being geminate, not opposite, and in their fleshy herbaceous 
habit, not having ligneous erect stems. 
To the Stilbacee they also appear to offer some approach, on 
account of their tubular calyx with unequal teeth, their funnel- 
shaped corolla with a bilabiate border, having an induplicate 
estivation, and a superior two-celled ovarium, with a single ovule 
in each cell: but this is erect, not suspended. They have also a 
slender capsule enclosed in the persistent calyx, and although it is 
two-celled, and monospermous in each cell, the seed is erect, and 
the embryo has an inferior, not a superior, radicle. They differ in 
many other respects, and are altogether extremely different in habit. 
.There are many analogous points of structure common to Zrapa 
and Selerophylax that should not be lost sight of. In the former, 
the calyx, though only half inferior, enlarges in like manner in 
fructification, entirely grows over the ovarium, and finally becomes 
enlarged and lignescent, the lobes being also converted into spines. 
The corolla, although consisting of distinct petals, offers a plicato- 
valvate estivation. The ovarium is two-celled, with a single ovule 
suspended in each cell. Here, however, the analogy ceases, for 
in Trapa, by the abortion of one of the ovules, the fruit becomes 
one-celled, with a single exalbuminous seed, and although the 
radicle is superior, the embryo, from the diminutive suppression 
of one of its lobes, becomes pseudo-monocotyledonous, added to 
which, the habit of the plant is quite distinct, and its alliance 
very remote, 
To Tetragonia, as I have before observed, there is certainly much 
apparent resemblance, but it is altogether external, for notwithstand- 
ing the similarity of its habit, and the spiny intumescence of its 
fruit, there exists no analogy whatever in the structure of the 
flower, or of its seed, to that of Sclerophylax. 
On a former occasion (see page 67,) I have endeavoured to 
trace the relationship of the Borraginee to the Convolvulacee, 
through the intermedium of Nolanacea and the Dichondree, 
