8 MB. D. OLIVEB ON AUBANTIACEJE. 



it would be useless to attempt more. The genera which I feel 

 to be thus opprobrious are Limonia and Triphasia. Liinonia in- 

 cludes, of Indian species, L. acidissima and L. alata. These plants 

 differ very considerably in habit as well as in floral characters, as 

 stated below. Koemer placed the former plant in his genus 

 Herperetliusa, and had it not been the original Linnean Limonia, I 

 should have been disposed to allow it there to stay ; but as, if 

 one of the species be separated from Limonia, L. acidissima must 

 remain, I have not ventured to separate L. alata, failing any plant 

 in the herbaria which I have examined more clearly congeneric 

 with it. As to Triphasia, T. trifoliata appears worthy of generic' 

 distinction, but T. monophylla, DC, I consider a trimerous Ata- 

 lantia, and T. glauca, Ldl., an Australian plant, awaits further 

 collectings before it can be permanently settled. It is but doubt- 

 fully a congener of T. trifoliata. 



The structure of the flowers in Auraniiacece offers hardly any 

 remarkable or anomalous feature. They are all regular and usually 

 symmetrical ; the ovary, however, frequently independent, in this 

 regard, of the outer whorls. In Piplostylis (now merged in Clau- 

 sena), a 3-celled ovary appears the prevalent condition (it varies 

 with 4, 5, or even 2) with outer whorls normally pentame- 

 rous. With the exception of cohesion of filaments in two genera 

 noted above, and the usually more or less gamosepalous calyx, 

 either cohesion or adhesion in the three outer whorls is excep- 

 tional. I find in Atalantia monophylla a tendency in the petals to 

 adhere slightly to the base of the stamina! tube : this is the only 

 case of notable adhesion that I have remarked. The sepals ex- 

 hibit all grades of cohesion, and vary from rotundate or ovate 

 broadly imbricating segments, separate to the base in Glycosmis, 

 to the cupulate or urceolate, often almost entire calyx, of some 

 species of Citrus, Paramignya, Luvunga, &c. In some forms of 

 Murraya exotica the calyx lobes are proportionally narrower than 

 m the rest of the order, and must be almost or quite valvate in 

 aestivation. The petals, normally 3, 4, or 5 in number, glandular, 

 farm, and mostly fleshy (least so in Clausena, which is rather 

 characterized by thinner petals), are free; in aestivation usually 

 imbricate, though in some Micromehms they become quite valvate, 

 with often a trace of obliquity at the margins ; in Clausena, though 

 clearly imbricate, they are usually not so much so as in the order 

 at large. In a Paramignya of Griffith's from Malacca, I find, in 

 respect to aestivation, a very unusual condition. The corolla here 

 appears truly valvate, indeed almost a little induplicate-valvate 



