MR. J. MIERS ON THE HIPPOCRATEACE/E OF SOUTH AMERICA. 325 
copious spiral fibres. Two other singular kinds of fruit are seen in Sicyomorpha and 
Tyloderma, which will be afterwards described; and it is probable that other forms 
still exist undiscovered. 
The greater part of the fruits, here for the first time described, were examined in the 
dried state, and therefore under unfavourable circumstances; but analytical drawings 
made many years ago, from the examination of live fruits of Clercia and Kippistia, have 
now enabled me to comprehend better the structure of the others. This knowledge is not 
so complete as may be desired; and we may hope that botanists who have favourable 
opportunities for the careful examination of living specimens will furnish more precise 
details. 
Being thus in possession of a greater number of tangible facts, we are better enabled 
to approach the question of the true affinities of the Hippocrateacee. We have seen 
how widely diversified have been the opinions of botanists on this head; and it would 
answer no good purpose to scrutinize them in detail. From this, however, should be 
excepted the latest views, as propounded by the learned authors of the new * Genera 
Plantarum,’ whose opinions are always entitled to the first consideration. In that work, 
P. 358, these celebrated botanists declare that “ between Celastracee and Hippocrateacee 
no good discrimination can be found;" and consequently they abolish the latter as a 
distinct family, and make it a tribe of the former order. As this view is now the highest 
authority on the subject, it may be well to examine the grounds upon which it must be 
either maintained or abandoned. 
I think it will be acknowledged that the facts communicated in this memoir are 
sufficiently numerous to show that the Hippocrateacee and Celastracee cannot remain 
associated in one family, and that their alliance is more remote than has been generally 
supposed. The reasons for this opinion may be briefly summarized under the following 
heads :— 
1. One of the most prominent circumstances is, that the stamens are generally aniso- 
merous in the one case, and always isomerous in the other, even when the number of 
parts vary from the usual rule: thus the Hippocrateacee with five sepals and five petals, 
have three stamens only (with one or, perhaps, two exceptions); while in Celastracee 
the stamens are invariably five, whether the petals be five or ten; or they are four when 
the petals are four. 
2. By far the most constant and valid discrimination is in the position of the stamens, 
which in Hippocrateacee are always within the disk and around the base of the ovary 
(never adnate or continuous with the disk, as asserted by many); while in Celastracee 
they are situated invariably outside the disk, not unfrequently agglutinated to it at their 
base. Goupia, placed by the authors of the ‘Genera Plantarum’ in the latter family, 
offers an exception to this rule; but good cause was long ago shown * for placing it out 
of that category. Kokoona also comes under the same rule, because the position of its 
stamens within the disk shows that it belongs to Hippocrateacee. 
3. In Hippocrateacee the anthers, peculiar and variable in form, are never introrse, 
but are either entirely or partially extrorse or lateral, generally reniform, and bursting 
* Contrib. Bot. ii. 132, pl. 74; Ann. Nat. Hist. 3rd ser. ix. 289. 
VOL. XXVIII. 2 Y 
