MR. BENTHAM ON LOGANIACE^. 71 



ovary having been described by Don as unilocular with parietal 

 placentae, contrary to the more accurate characters given by Euiz 

 and Pavon, and by Bonpland. 



The geographical range of the species is extensive, — along the 

 whole length of the Andes of South America, from New Grenada 

 to the Straits of Magellan ; and, as might be expected, there are 

 considerable variations in the foliage, although much less than in 

 some of our own shrubs, such, for instance, as our common Holly, 

 The connecting line of the petioles often shows on each side two 

 minute teeth or protuberances, from whence two prominent lines 

 are more or less decurrent along the young branches, disappearing 

 entirely on the older ones. In the small-leaved specimens gathered 

 at great elevations within or near the tropics these lines are par- 

 ticularly prominent, and characterize the D. acutangula of Dunal. 

 The southern specimens have usually a luxuriant foliage and 

 broader and more ciliate lobes to the calyx, constituting the D. 

 Hookeri, Dun. Specimens similar to these, but with unusually 

 large leaves and more numerous teeth, were originally selected by 

 Euiz and Pavon to figure as their D. spinosa ; and when Bonpland 

 had only before him the commoner Columbian form with few 

 large teeth to the leaves and narrow scarcely ciliate lobes to the 

 calyx, he did not venture to identify them as the species figured 

 in the ' Flora Peruviana,' and therefore published them as a 

 distinct one under the name of D. splendens. But all these trifling 

 differences are so variously combined in the numerous specimens 

 before me, that I cannot but regard them as mere variations of 

 one species which will retain the older name of D. spinosa. 



16. Eagkelzea, Thurib. 



Fagraeas may be almost characterized as Gardenias with a free 

 ovary. The habit and flowers are very similar ; there is in both 

 an occasional tendency to an increase in the number of lobes of 

 the corolla and consequently of the stamens ; some few species of 

 Fagrcea have even the peculiar exceptional character of Gardenia, 

 an incompletely divided ovary, the parietal placentae not reaching 

 quite to the centre. But in Fagrcea the ovary is completely free, 

 and the stipules are reduced to mere auricular expansions of the 

 base of the petiole, like those of some East Indian Taberncemontanas. 

 In other respects Fagrcea has less in common with Apocynea than 

 several other Loganiaceous genera, but it forms in the whole 

 family the nearest real approach to Gentianece. Comparing it 



