2 ME. D. OLITEE ON ATJEANTIACE^E. 



Aurantiacea? which require a brief allusion here, one of the earliest 

 and most important is certainly the essay of Correa do Serra in 

 the ' Annales du Museum'*, entitled " Observations sur la Famille 

 des Orangers ct les Limites qui la circonscrivent." Considering the 

 period, a remarkably good appreciation of natural affinities is ma- 

 nifest, both in the observations upon those characters which he con- 

 ceived to be available for the purpose of generic grouping, as well 

 as in the redistribution of the genera of Aurantiacece there pro- 

 posed. In this paper, Atalantia and Glycosmis, new and natural 

 genera, were founded ; while, on the other hand, one or two pre- 

 viously published genera (Bergera, Chalcas, Clausena) were merged. 

 The opinion of Correa de Serra that Bergera is not generically 

 distinct from Murraya I coincide in, though I should scarcely per- 

 haps have united them, had I not had the opportunity of examining 

 an Archipelago plant of Lobb and Cuming, from the Philippines, 

 which supplies an important connecting link between them. I do 

 not, however, agree with him, and with Sprcugel, that Clausena 

 should be, in like manner, united with Murray a. The addition of 

 new forms since Correa wrote, tends to establish Clausena as one 

 of the more distinct generic types of the order. As in Clausena, so 

 also in most of the genera, new species from India, Eastern Asia, 

 the Archipelago, Australia, and Africa, have materially altered the 

 aspect of the family ; and as these accumulating materials have 

 been too often published from insufficient data and without the 

 advantage of herbarium comparison, and have been also allocated 

 on very artificial, if not on arbitrary grounds, the order would 

 seem about to fall into a sort of ill repute in consequence of the 

 best types for the systematist having thus become almost entirely 

 swamped. 



Correa de Serra f regarded the genus Citrus as expressing the 

 type and centre of affinity of the Aurantiacece. If in the species 

 known at the present time, the numerical proportion of those pre- 

 senting numerous important structural features in common be the 

 main element in determining the type, it can scarcely be said that 

 the multilocular ovary, numerous ovules, and numerous usually 

 polyadelphous stamens of the genus Citrus permit it fairly to repre- 

 sent the order. The number of species having normally five cells, 

 or fewer, to the ovary, ovules solitary or in pairs, and stamens very 

 rarely normally exceeding ten, I estimate approximately at about 

 sixty (53 to 75) ; these at present are referred to ten genera. Of 

 polyandrous, multiovulate species, I reckon but about ten or eleven, 

 * Tom. vi. (1805) p. 376. f Loc. cit. p. 378. 



