The same kind of diversity of structure also obtains in 
Pedaline as proposed to be limited at fol. 934. In that 
order, at least in Martynia Zanquebarica, which is either 
a species of Josephinia, or a nearly related genus, the two 
opposite placentas meet in the centre, where they form a 
cohesion, but being 2-lobed, and their lobes revolute, a 
cohesion is again formed with the sides of the endocarpium, 
where the edges of the lobes touch it; whence the cap- 
sule is 6-celled, the two anterior and posterior empty cells 
being formed by the space occasioned by the divergence 
in opposite directions of the lobes of the placenta, and the 
4 lateral seminiferous cells owing their existence to the 
accretion of the lobes of the placenta to the sides of the 
ovarium. Now, in Martynia, there being no folding back 
of the lobes of the placenta, which, on the contrary, are at 
right angles with their centre, and no cohesion taking place 
between the placentas themselves or their edges and the 
sides of the ovarium, the necessary consequence is, that 
the pericarpium is unilocular, with Spurious cells; whence 
it may be predicated that Martynia bears the same rela- 
tion to Pedaline as Eccremocarpus to Bignoniacee. 
Perhaps Eccremocarpus may 
sidered the connecting link 
Gesneriez, agreeing with the 
ovarium, glandular pubescence, 
ties of appearance, and with the former in their more 
essential attributes. Cyrtandracez of Dr. Jack, which 
led capsule as described by that 
d not a 4-celled one as erroneously 
f some remarks upon Didymocarpew 
order under another d I 
most properly be con- 
between Bignoniacee and 
latter in their unilocular 
and some other peculiari- 
with the latter 
2 considered to 
\< same relation 
to Bignoniacex, as Eceremocarpus 6n the part of the 
latter bears to Gesneriez, 
Jie, 
