MR. G. D. HAVILAND: REVISION OF THE NAUCLEE.E. 9 
are broad and flat; when the placentas are linear with central 
attachment (Uncaria), the capsules open at the side, the two 
ends for the most part remaining joined, and the seeds are small 
with the long filiform processes at either end. In this last case 
the flowers, and especially the fruit, are generally pedicelled, 
so that they are less crowded than in the other forms, in which 
both the flowers and the fruit are always sessile. 
Interfloral bracteoles also give characters of value in generic 
grouping. When the capsules open at the top, the bracteoles 
are paleaceous with spathulate tips, subglabrous, and so nume- 
rous as to completely separate the different flowers. More 
frequently (Cephalanthus and Adina and some species of Uncaria) 
the bracteoles are filiform and hairy with small spathulate tips, 
and in this case there is generally but one for each flower or 
even less. In some species of JVauclea there are between the 
flowers short thick glabrous cones reaching not much higher 
than the ovary. In other species bracteoles are wholly absent ; 
but in these what seem to be rudiments of flowers are some- 
times found between the perfect flowers. The receptacle is 
generally very hairy. 
The inflorescence also provides characters of value in generic 
grouping. The simplest form is a solitary peduncle bearing the 
head of flowers at its apex, jointed and bearing two pairs of 
bracts at the joint; one pair of these bracts is apt to become 
foliaceous in charaeter, aud the other stipular. They may be 
seated near the top of the peduncles, when they are generally 
large and protective to the flower-head in bud, or they may be 
seated near the middle of the peduncle, in whieh case they are 
smaller; and it will be often found that they protect the flower in 
very early bud, the upper part of the peduncle lengthening dispro- 
portionately to the lower at a later stage : or, again, they may be 
seated below the middle of the peduncle, when they are rudi- 
mentary and functionless; in this case the young flower-head is 
often protected by large stipules at the base of the peduncle. 
In the genus Uncaria it is common to find such simple peduncles 
terminal and axillary, the terminal ones being generally ternate 
and the lateral ones always solitary, though exceptionally thev 
are branched ; some of the lateral ones in this genus are gene- 
rally reduced to compressed recurved barren hooks. Ln the other 
genera it is very rare to find simple peduncles both terminal and 
axillary; in one species of Sarcocephalus terminal and axillary 
