10 MB. D. OLIVER OK AURANTIACEJ!. 



gularly, or leaving a clean horizontal scar. In one species, Ata- 

 lantia monophylla, this indurated scar or pseudostigma has been 

 mistaken for a stigma, and a species accordingly founded on it 

 {A. platystiyma, It. W.). With regard to the style and ovary, 

 though the relative proportional lengths vary considerably through 

 the order, the species of each of the more natural genera are 

 tolerably uniform. 



The fruits of the various species oifer marked contrasts ; with 

 the familiar Hesperidium we find but little in common either in 

 the succulent 1- or 2-seeded berry of Glycosmis, or in that, often 

 almost dry, of Clausena and Micromelum. None of the fruits 

 of the Aurantiacece, that I am aware of, dehisce ; and in this we 

 find one distinguishing mark from a prevalent character of the 

 allied Xanthoxylaccce. The seeds usually present a pair of equal, 

 fleshy, more or less planoconvex, oval or oblong cotyledons, unless 

 mutual pressure, or a plurality of embryos, alter their form, as is 

 frequently the case. The most remarkable difference from the 

 typical structure is in the embryo of Micromelum ; the cotyledons 

 (in M. pubescens) being foliaceous, broad, almost or quite reniform 

 or cordate, and contortuplicate, with a conspicuous radicle. 



The inflorescence in Aurantiacece, upon the whole, I regard as 

 essentially determinate and cymose, as is generally manifest in the 

 multifoliolate genera (e. g. Glycosmis, Micromelum) ; in some spe- 

 cies oiAtalantia, and in Citrus*, this determinate character is less 

 obvious, though I think it still obtains f. 



Upon the geographical distribution of the Aurantiacece I have 

 not much to observe. The genera and species greatly preponderate 

 in India, the Malayan Peninsula and Archipelago, four or five 

 genera extending north to China and Japan ; four or five differing 

 widely from each other, representing at least two (perhaps each) 

 of the three sections of the order usually adopted, reach Australia— 

 viz. Citrus australasica, Planch J, Triphasia glauca, Glycosmis ar- 

 borea, var. and Micromelum pubescens, var. and one or two undeter- 

 mined species. Ten genera, containing fifteen or sixteen species, 



* Conf. Payer, Organ. Yegct. p. 113. 



t A. DeCandolle, in a valuable memoir upon the Beyoniacem (Ann. Sc. Nat. 

 Ser. iv. Bot. vol. xi. p. 100), intimates his opinion that axillary cymose inflores- 

 cences are infrequent associated with truly alternate leaves. 



X Tliis differs from typical species of Citrus in its free stamens. Specimens 

 in fruit, closely resembling C. auslralis, are in Sir W. Hooker's herbarium from 

 Cunningham, under the name of Limonia australis ; though Cunningham's me- 

 morandum states the only flower he could find was decandrous and the stamens 

 free; it may be possibly an Atalantia allied to A. missions. 



