6 MR. D. OLIVER ON AURA.NTLA.CK.E. 



access, and I have generally abstained from an attempt to base new 

 or modified diagnoses upon them. 



One result, especially in the case of one or two species, of this 

 extended correlation of specimens, has been peculiarly interesting ; 

 showing, as I consider, different varying tendencies in different 

 geographical areas occupied by them. In the case of Mioromelim 

 pubescens, e. g. the form originally described by Blume, growing 

 in Java, presents moderately-sized flowers (2-3 lines), with a cu- 

 pulate, almost quite entire or slightly undulate-dentate margin, the 

 leaflets usually from 2£-3£ inches long, by 1-1^ broad, the style 

 exceeding or not shorter than the ovary. The same form we have 

 from Penang. In Ceylon wc find the calyx more decidedly lobu- 

 late (dentate-lob ulate), the style comparatively long, usually rather 

 exceeding the ovary, the leaflets commonly narrower in proportion, 

 varying usually from 1|-3| inches in length, by ^ to 1^- inch in 

 breadth. I have not seen specimens from the western peninsula 

 of India*. In the north-east, in Sikkim, Nipal, and Assam, a 

 strong form occurs with rather larger flowers varying from 2-3| 

 lines, in very broad corymbose cymes — the calyx triangularly lobate; 

 leaflets 3^-6 inches long, by usually 1^-2 inches broad ; the style 

 moderately long, about equalling, not exceeding, the ovary. 



Eeturning to the islands : in the Western Pacific, — Friendly 

 Islands, Fijis, and New Caledonia, — another form occurs with 

 flowers altogether smaller, scarcely reaching 2 lines ; calyx den- 

 tate-lob ulate, the style short, shorter than, or scarcely equalling, 

 the ovary, which is frequently 4-locular, the leaflets usually broader 

 in proportion to their length, and more decidedly ovate — from 

 l£-3£ inches by l-2£ inches. This is the Mioromelwn glabrescens, 

 Bth., Limonia minuta of Forster, as pointed out by Dr. Asa Gray. 

 Then again in Australia and its adjacent islands (Endeavour Kiver, 

 Wide Bay, Cape Upstart, Port Essington, Cairn Cross Island, &c), 

 we find a form very nearly allied to the last named, often drying a 

 yellowish brown colour, and with the ovary varying with 4-5 or 

 3 cells ; the young fruit too varies in amount of pubescence. A 

 corresponding series of phases, other characters being affected, is 

 offered by Glycosmis pentaphylla, a most protean and difficult spe- 

 cies, — in respect to leaves only, varying from a simple (1-foliolate) 

 form— the lamina 10-12 inches long by 2-4 inches broad— to 

 pinnate (3-5 leaflets), each lamina 2-3 inches by |-1 inch. The 

 number of loculaments in the ovary, though clearly normally 



* It is noteworthy that it lias not yet, as far as I am aware, been found 

 there. 



