Mil. D. OLIVEE ON AUBANTIACE.E. 21 



sufficing to retain Artliromischus of generic rank. I have there- 

 fore united it with Paramignya, to one or two species of which it 

 is Aery closely related in every respect. 



Luvfnga. 



Excepting the large trifoliolate leaves, and the inflorescences 

 generally in abbreviated cymose racemes or panicles, there is but 

 little difference between this genus and the foregoing. Further 

 material may enable us to decide more satisfactorily than at 

 present is possible, whether they should be kept separate or 

 united. L. eleutlierandra, Dalz., from the Western Peninsula and 

 Ceylon, is perhaps doubtfully distinct from L. scandens. From 

 the irregularity in the amount of cohesion between the filaments 

 of the latter species, I cannot regard the free stamens of Mr. Dal- 

 zell's plant as affording, per se, a specific character, especially that 

 in Ceylon we find free stamens associated with the narrow leaflets 

 characteristic of L. scandens. There are probably two or three 

 species of Luvunga in Borneo, judging from the imperfect speci- 

 mens which I have seen. 



Febonia. 



I am only acquainted with the species originally described in 

 the Society's Transactions (vol. v. p. 224) by Correa de Serra, — 

 F. elephantum. In dissecting the ovary at the time of flowering, 

 or just before, the placentas are found to separate very easily, pre- 

 senting the appearance of a unilocular ovary. This slight degree 

 of axial adhesion of the placentary partitions at so late a stage in 

 the development of the pistil is unusual, and I do not remember to 

 have noticed it in any of the genera which I have examined in 

 detail. I am not sure but that, either above or below, the ovary 

 may be unilocular at the time of floAvering*. 



jEgle. 

 I have nothing to add to the published accounts of this well- 

 known plant, the Bael of India, Blume describes one species in 

 his 'Bijdragen,' as cultivated in Java under the name of JE. sepia- 

 nia. So far as I can judge from his description, I agree with 

 Messrs. Wight and Arnott in considering it doubtfully distinct 

 from JE. marmelos. Blume does not mention this latter name. 



* Koxburgh (Fl. Tnd. ii. 412) says, "the germ is 1-celled, containing nume- 

 rous OTula attached to 5 parietal receptacles." 



