lg MB. D. OLIVES ON JLTJBANTIACE.E. 



circumscribe it, as nearly in conformity with what has appeared to 

 me to be the present condition of the species in nature as I could, 

 offers, as already observed (p. 6), a very interesting study. It 

 is a remarkably variable plant, varying especially in the number 

 of leaflets, their size and form, the size of the flowers, the disk, and 

 its relation to the ovary, and also the number of cells in the ovary. 

 There are at least two principal races of this species, which are 

 generally to be easily recognised at sight. One of these has the 

 axillary cymose panicles often clustered towards the extremity of 

 the branches, forming a dense mass of inflorescence exceptional 

 in the genu?; the flowers rather larger; the pistil frequently 

 mammillate, owing to prominent immersed glands ; and the disk, 

 accrete with the base of the ovary. In a second race, the flowers 

 are very small ; the inflorescences, although of numerous flowers, 

 much reduced in size, falling usually very considerably short of 

 the leaves; the pistil but slightly glandular, mammillate or 

 smooth, and separated by a slight constriction from the disk, 

 which is sometimes, very small and connate with the very short 

 gynophore. 



I should perhaps be disposed to hold these races as specifically 

 distinct, were they not so closely united through the more Eastern 

 variety from China and the Philippines, as well as by varying forms 

 growing in the Peninsulas. I do not find, as in the case of Micro- 

 melum, that the chief races are conspicuously confined to special 

 areas, further than that the larger-flowered race described above 

 seems chiefly limited to the western portion of the area of the 

 species. The Chinese and Archipelago variety (G. citrifolia, Ldl.) 

 presents much of the fades of the first-mentioned race, with the 

 structure of disk and pistil of the second. 



MlCBOMELtTM. 



Established by Blume, in his < Bijdragen,' in 1825. He pub- 

 lished but one species, M. puhescens. Though I have not seen 

 authentic examples, I have no hesitation in recognising his plant 

 in Javanese specimens collected by Spanoghe, Lobb and Zol- 

 linger, as well as in numerous examples from other countries. 

 Curiously enough, Hasskarl * refers Micromelum to Coohia punc* 



* Plant® Javanioae Eariores, p. 280. Of CooHa, he says " Semina unum 

 alterumve evolutum, reliqua abortiva, et inde fructus dissepimenta intor a , 

 and again, in ' Cat. Hort. Bogor.' alter, p. 215, in foot-note to Micromd«* 

 (quoted as a synonym of Cookia), " Calyx parvus, urceolatus, 5-fidus (nee in 



