MR. BENTHAM ON LOGANIACEiE. 75 



(usually ten) of the lobes of the corolla and of the stamens, whilst 

 that of the lobes of the calyx is four only. But all the other 

 characters are too near to those of Fagrcea to justify its removal 

 into another tribe, more especially as, according to Blume, Fagrcea 

 itself has occasionally six or seven lobes to the corolla. 



There appears to be but one species of JPotalia known from 

 tropical America; at least I can find no difference between 

 Martin's Cayenne specimens, which are evidently Aublet's P. 

 amara, and Spruce's Rio Negro ones, corresponding with Martius' 

 P. resinifera. Both are low weak shrubs (1 to 3 feet high 

 according to Aublet, about 4 feet high according to Spruce), 

 of which the short flowering branches are said to wither and fall 

 off with the inflorescences, as is the case with a great number of 

 other shrubs and even trees. The plant figured in Griffith's 

 I Icones Plantarum Asiaticarum,' t. 383. fig. 1, as a Fotalia, appears 

 to be a species of Ehretia allied to E. longiflora, Champ. 



18. Anthocleista, Afz. 



The great difference between this African genus and JPotalia 

 consists in the greater development and singular arrangement of 

 the placentae, well figured in Hooker's Icones, t. 793, 794 (Niger 

 Flora, t. 43, 44) . They appear to be two parietal placentae twice 

 bifid and connected together by a spurious dissepiment dividing 

 the ovary into two cells, thus giving the appearance of two pairs 

 of opposite bifid placentae placed at some distance from each other 

 on the dissepiment, whilst in JPotalia there are but two bifid pla- 

 centae in the centre of the dissepiment. This difference is, how- 

 ever, not greater than those observable in the placentation of 

 different species of Fagrcea, and had the genus Anthocleista not 

 been already established, I should certainly have considered it as 

 a second species of JPotalia. 



On a further examination, I see no reason to alter the opinion 

 I had already expressed in Hooker's ' Niger Mora,' that the three 

 supposed species of Anthocleista, A. nobilis and niacrophylla of 

 Don, and A. Vogelii of Planchon, are in fact but one, the di- 

 stinctive characters given being liable to variation even in the 

 same specimen. 



19. Stetchnos, Linn. 



The genus Strgchnos appears to have no very exact parallel in 

 either of the allied families JRubiacece or Apocynece. In the former, 

 the combination of a succulent indehiscent many-seeded fruit with 



