4 MB. D. OLITEB ON AUBA.yTIACEJE. 



the order*. Sclerostylis I think cannot be maintained, and most 

 of the Indian species referred to it by recent writers I unite with 

 Atalantia. Micromelum offers one or two peculiar features, and 

 is, I consider, a good genus. These genera I notice in further 

 detail in the remarks upon the several generic groups. 



It is to Messrs. Wight and Arnott that we owe the first good 

 digest of the genera. This appeared in their ' Prodromua Planta- 

 rtim Peninsuke Indiae Orientalis ; ' some farther observations are 

 added by Dr. Wight in his ' Illustrations of Indian Plants.' 

 Although they confined themselves to Indian members of the 

 order, yet as nearly all the genera are represented in India, their 

 observations have a general interest. The sections which they 

 propose, Limonece and Clausenece, dependent upon the ovules being 

 collateral or superimposed in the biovulate species (uniovulate 

 species are included in the former section), are adopted by Meisner, 

 by Endlicher in his ' Genera Plantarum,' and other botanists. I 

 may observe that, to estimate the value of character afforded by 

 the relative position of the ovules in Aurantiacece demands con- 

 siderable caution. In some genera the ovules are constantly, and 

 apparently entirely, superimposed, the upper ovule ascending or 

 peltate, the lower pendulous {Micromelum pubescens) ; in others, 

 the ovules are usually superimposed more or less obliquely, at 

 times becoming almost or quite collateral. Indeed, in the same 

 species, sometimes in the same ovary, both partially superimposed 

 and collateral ovules may be found (Mnrraya crenulata and Li- 

 monia alaia). Other species, again, appear to have their ovules 

 invariably side by side, as in the biovulate species of Atalantia. 

 In perhaps all of the biovulate Aurantiacece the ovules are attached 

 to the placenta very nearly or precisely at the same level ; when 

 they are superimposed, the lower is furnished with a very short 

 funicle, and is almost always dependent or pendulous, while the 

 superior ovule is peltate, or more or less laterally attached. In 

 the collateral-ovuled species the pair may be alike pendulous or 

 laterally affixed. 



Altogether, after the dissection of a very large number of flowers, 

 I am of opinion that, like many other absolute, isolated, diagnostic 

 characters, this breaks down in too many Cases to render it of 

 subordinal, or indeed of generic, value, unsupported by additional 



* I have felt a like difficulty in respect to the species described by Blanco in 

 his < Flora de Filipinas.' Though most of his plants are probably in our her- 

 baria, and familiar to me, yet I cannot certainly identify them with his de- 

 scriptions. 



