DIOECIA— DIANDRIA. Snlix. ^07 



lanceolate, silky; its silky stalk nearly equal to the linear- 

 oblong scale. Style not longer than the stigmas. 



S. argentea. Fl. Br. 1059. Engl. Bot. v. 19. t. I3G4. Rees's CtjrI. 



n. 98. Jnild.Sp.Pl.v.4.G93. 

 S. repens y. HtiJs. 429 ; but not n. 1G43 of Holier. 

 S. arenaiia. Lightf. 604. Iluds. erf. 1. 3G4. 

 S. lanata. Roth Germ. v. 1,418. r. 2. p. 2. 513j according to Will- 



denov). 

 S. pumila, foliis utrinque candicantibus et lanuginosis. Dill, in 



Rail Sijn. 447. t. 19./. 3 ; but not of C. Baiihin. 



On the sea shore, among loose blowing sand-banks. 



By Sandovvn castle, Kent. J. Sherard. Laugharn, South Wales. 

 Mr. lliirlock. In various other places. 



Shnib. Maij. 



Stems mostly spreading, but if sheltered erect, 4 or 5 feet high, 

 with numerous, upright, leafy brandies, beautifully downy or 

 silky. Leaves on short, stout, downy footstallcs, scattered, an 

 inch, or often less, in length, and half as much in breadth, 

 truly elliptical, with a small curved point ; the margin entire, 

 slightly revolute ; the upper side of a dull green ; at first silky, 

 then downy, finally naked, reticulated with small veins ; under 

 covered at all times with the most brilliant, silvery, satin-like, 

 close, silky hairs, very soft, almost concealing the strong mid- 

 rib and transverse veins. Stipulas ovate, entire, flat, silky, 

 more or less stalked, variable in size. Catkins before the leaves, 

 lateral, at first sessile, afterwards elevated on small bracteated 

 stalks, cylindrical, thick, obtuse, an inch long. Scales narrovv- 

 obovate, rounded at the end, bearded, their upper half black ; 

 those of the barren catkins almost linear. Stam. 2, smooth, 

 about the length of the scale. Nect. a small, obtuse, oblong 

 gland, like that of the ievtWe Jlorets, whose scale is broader than 

 the barren one. Germ, on a silky stalk nearly as long as the 

 scale, lanceolate, densely woolly, or silky, tumid at the base. 

 Style at first short ; subsequently equal to the stigmas, which 

 are cloven and blunt. Capsules becoming smoother as they 

 ripen. The length of their partial stalks appears variable. 



With the present species we enter on a peculiar tribe of Salices, 

 natives either of pure sand, or of heathy elevated situations. 

 Hudson unites them all under his repens, along with a most 

 distinct one, the rosmarin folia, whose leaves are linear. The 

 rest have elliptical foliage, mostly entire, peculiarly silky under- 

 neath. They are of humble stature, more or less recumbent, and 

 often matted, or entangled, in their growth. Though the species 

 of this tribe may be difficult to define, like those of the whole 

 genus Rosa, which my learned friend Gerard, in his Flora Gal- 

 loprovincialiSf reduced all into one ; yet a patient inquirer will 



