EL2EAGNACE2E. 487 



and some Comhretacece. A. L. de Jussieu,' in 1789, only followed 

 Ad anson, adding to his order Elaagnacece (Chalefs) Quinchamalium 

 and Colpoon, which are also Santalacea. A. Richard reduced this 

 order to its present limits in 1823, in a special Monograph, 2 wherein 

 he describes, besides El Magnus Shepherdia and Ilippophae, the plant 

 Conuleum, which, as we have seen, 3 is a Monimiad of the genus 

 Siparuna. However, De Schlechtendal, reversing the order 

 Elceagnacea* for the Prodromus, 5 retained the genus Conuleum, and 

 added as doubtful genera Octarillum of Loureiro, 6 and Acxloxicon of 

 Ruiz and Pavon. 7 This little order has since remained unaltered; 

 it contains some thirty species, of which five-sixths belong to 

 Eheagnus. This genus inhabits the temperate regions of Europe, 

 Asia, and North America. Shepherdia is peculiar to America ; 

 Ilippophae to the Old World ; each genus contains two species. 



All the Elaagnacece are arborescent or frutescent ; 8 all have their 

 organs covered with peltate or stellate scurfy hairs, often silvery or 

 scurfy ; all have exstipulate leaves, naked leaf-buds, small incon- 

 spicuous flowers, possessing one or two whorls of stamens with 

 introrse anthers, and a single carpel with anatropous ovules. Of 

 the variable characters we consider some of primary importance, and 

 have used them to divide this group into two series, whereof one 

 is only a doubtful member of the order. 9 This is Aextoxicece, 

 wherein the floral receptacle is scarcely concave, the ovary contains 

 two collateral descending ovules, and the perianth is triple. 10 The 

 Elceagnece have on the contrary a sacciform receptacle which persists 

 around the fruit, to which it forms a fleshy, often drupaceous indu- 

 sium ; their perianth is simple, and their ovule is solitary, suberect. 

 The other varying characters are best fitted for generic distinction. 

 The leaves are opposite in Shepherdia, alternate in Ilippophae and 

 Elceagnus; the flowers are hermaphrodite in most species of the 

 latter genus, dioecious in the former two. The perianth may consist 



1 Gen., 74, Orel. i. to come near Elceagnus ; but the structure of its 



i In Mem. de la Soe. d'Hist. Nat. de Paris, ovary is quite unknown to us. The seed is de- 



i. 375-408, t. 24, 25. 3 Vol. i. 305. scribed as arillate (?). 



4 Lindl., Introd., ed. 2, 194; Veg. Kingd., ' Prodr., 131 (1797). 



257. — Elmagnea R. Br., Prodr., 350. — Kndl., s The branches often taper into spines bearing 



Gen., 333, Ord. cxi. — Elceagnidece DrJMORT., only rudimentary leaves, or leafless. 



Anal., 15, 18. 5 XIV., 606-616. 9 Aextoxicon has been referred to the doubtful 



6 Fl. Cochinch., 113. — Ekdl., Gen., n. 2083. Euphorbiacece by ENBLICHEB, to llicinece by 



— Schlchtl., Prodr., 615. By its tubular te- MlEES, to Monimiacere by DECAISKE. A. in: 



tramerous perianth, its four stamens, and its Candollk (Prodr., xvi. (ill) accepts none ot 



simple gynseceum, the place of this genus appears these affinities. '" See p. I s .">, note 1. 



