4 : MR. J. MIERS ON THE GENUS ATAMISQUEA. 
spond with the edges of the segments, which show by their laceration that their adhesion 
with each other and with the ribs has been complete. Within and opposite to the lower- 
most of these ribs arises a funiculus or placenta, which on reaching about two-thirds the 
height of the fruit, branches off right and left, by two short threads, towards the hilum 
of the two seeds, where they are respectively attached. The seeds are smooth, of a dark 
red colour, reniform, or of a cochleate shape, somewhat flattened on their adjacent sides, 1 
and roundish without. The testa is coriaceous, having on one side an incomplete cell, 
formed by the convolution of the inner margin about the umbilical sinus; the outer in- 
tegument is brownish, opake, and striated, and adhering to the testa forms between the 
flexure of the embryo an extension of the false dissepiment of the spurious cell, which 
serves to inclose the radicle: the inner integument is membranaceous, and marked about 
the middle of the cotyledons with a broadish thickened chalaza. The embryo is oblong, 
and bent sharply inwards at both extremities, the ends of the cotyledons and of the radicle 
being mutually turned towards each other, so that it may be said to be truly campylo- 
tropous: the cotyledons are convolutely plicated, and somewhat white and foliaceous. i 
From the facts above stated it may be inferred, that the arrangement of the floral enve- - 
lopes in this genus is contrary to the usual structure of the Capparidacee, which offer 
generally four sepals, four alternate petals, usually eight or more stamens, and a fruit, 
usually of two cells, with two or more placentæ. Sir W. Hooker, in his generic character _ 
of Atamisquea (loc. cit. p.148), regards its floral teguments as consisting of four sepals 
and four petals, in conformity with the ordinary arrangement in this family: it will be 
against the opinion of so high an authority. It appears to me however warranted by the 
they are continuous at 
their origin with the margin of the cup of the torus, while the insertion of the six narrower 
segments (petals) is upon one line, within the margin of the same cup, which is proved — 
by the fact, that when the sepals and petals fall away, the rupture of the former is marked _ 
an entire envelope that bursts into two valves, viz. in Busbeckia, Endl., Steriphoma, Spr., 
and Morisonia, Plum. The apparent inconsistency of this distribution will BR: if 
of three series, each consisting of two normal 
from the cleaving of the lobes down to their point 
per and lower pairs of petals upon the torus there 
