DKECIA. DIAND. 277 



Fertile Ft. Perianth single, tripartite. Stijles 2. Caps. 

 2-ce!le(l, cells 1-seedctl. 



10. Hydkocharis. Barren FL Cal. tiij)artite. Petals 3, ** the 

 o interior filaments beaked," Sni. 



Fertile FL Cal. tripartite. Petals ?>. Styles (i, each with 

 2 stigmas. Caps, inferior, coriaceous, roundish, ()-ceIled, 

 many-seeded. 



(Empetrum, Ord. Tkiandk.) 



(Dr.CANDRIA. 



Lychnis dioica, CI. DiiCAND. Pentag.) 



(ICOSANDRIA. 



Rubus Chanicemonis, Icos. Poi.yand.) 



(PoLYANDRIA. 



Stratiotes atoiJes, Poi.yand. Pentag. Populus nigra, 



Ord. OCTAND.) 



7. MONADELPHIA. 



11. JuNiPKRUs. Barren FL Scales of the catkin subpeltate. 

 Perianth 0. Stam. 4—8, I -celled. 



Fertile FL Scales of the catkin tew, united, at length 

 fleshy, and surrounding the 3-seeded berry. 



12. Taxus. Barren FL Perianth single at the base, Stam. lui- 

 nierons. ylnthers\)ehiilc, 6 — 8-celIed ; cells opening beneath. 



Fertile FL Perianth single, urceolate, scaly. Style 0. 

 Drupe fleshy, perforated at the extremity, 



{Salix ruLra, Ord, Diandr.) 



1. DIANDRIA. 



1. S.>LIX^ 

 * Leaves serrated, smoothish, espcclat'ij above. 

 1. S. purpurea (bitter purple IVillow), monandrous decumbent, 

 leaves lanceolate broadest upwards serrated glabrous, germens 

 ovate very pubescent sessile, stigma nearly sessile. JLightJ. 

 p. 598. E. B. t. 1388 

 Hab. Banks of the Esk, near Netherby, in Eskdale, Lightf. N.bank 

 of the river Tweed, opposite Melrose, Maugh. Ft, March, Apr. Jj . 



" The following specific characters of this most intricate genus are taken 

 in every instance, where not otherwise nientioned, from specimens that are 

 the best authority for the E. Bot. species ; that is, from individuals j^athered 

 in the late Mr. Crowe's garden near Norwich, in company with Mr. Crowe 

 himself, or with Sir James Smith. If my characters differ in some measure 

 from those published by Smith, it will shovv how variable are these plants, 

 and how careful we ought to be in not multiplying the species unnecessari- 

 ly. Future observations, I thinlt', will warrant the reduction of the present 

 very extensive list of Britioh willows. 



