NEW PUBLICATIONS. 97 
in habit, having broom-like branches terminating in a spine, and especially 
with the genus Bontia, with which it agrees in its hard nut, and which it ap- 
een in the co unter ot. its ongia; Mr. Miom, on tha other hand, believes 
thatitis easy to show 
to the Myoporacee than to the Bignoniacea. In pate in nearly all cases, 
the leaves are alternate they P opposite i in t Poli edis and in several 
species of Myoporum, Ep.]; t 
any rudiment of a fifth ; the ovarium is only bilocular i in two instances, which he 
has elsewhere shown ( An n. Nat Hist ind. ser. xi. 439) are : doubtíal, or ket NM 
abnormal genera of the snis i 
celled, with a single ovule suspended from the apex of each cell, and this: ripens 
into a four-celled ligneous indehiscent nut, with a seed in each cell. The only 
remaining case is Bontia, which differs from all Dess of this fanily $ in jan 
originally a bilocular ovarium, but where by t 
of en piacente, eight pesce are T each with a -- suspended 
ovu 
ars seed, with a thick osseous testa, which is lee poor with the sides 
of the cell. Mr. Miers’s knowledge of this genus is derived wholly from the 
descriptions of authors, and he finds no observations of a more recent date 
aoe those of Gertner and Jacquin; our evidence of its real structure is there- 
imperfect, but enough is recorded = show that it is a very anomalous 
Pus if it really belong to the Myoporace 
* The author next proceeded to ree tines points of structure in Ozy- 
cladus, which establish the relative value of its affinity to the Myoporacee or 
the Bignoniacee. In this genus, both the branches and Map leaves 
are distinctly A ine [so are several Myopora, Ep.], as in Bignoniacee, 
e find two other genera, where the sind terminate 
in spines, viz. Daai and Rhigozum : the flowers are bluish, a colour 
msa and the at one-celled, as 'yoporacee; the ovarium 
seated upon a five-lobed fleshy disk, which never occurs in the latter family, 
though consta igno ; it is completely bilocular, with about six 
ovules in each cell, suspended and attached by a ventral thread to a distinet 
flat dissepiment, and arranged in three superimposed pairs upon its opposite 
faces, in two lines parallel with the axis, a structure which offers a marked cha- 
— in the Bignoniacec, and unknown in the Myoporacee ; of these twelve 
vules, all become abortive, with the exception of one; the fruit is therefore 
ibo and monospermous, presenting an osseous nut, with four deep furrows 
in the apex, and divisible to the base along these stris into four valves, two of 
these sutures being more easily separable, and always corresponding with the 
margin of the persistent dissepiment, which is pressed against a and 
seed filling the whole capacity of the nut. In Myoporacee, whether the nut 
be 4-celled or by abortion 2-locular, the intervening space is always solid, and 
