Mu. J. G. baker's MONOGHAinr OF imiTjsu KOSKS. 2:J9 



floribus plerisque 3-6, pedunculis modice elongatis leviter aciculatis, 

 sepalis retiexis deciduis dorso nudis vel tenuitt r glaudulosis, inajori- 

 bus copiose pinuatis, fructibus late ovatis nudis serotinis discis valde 

 . incrassatis iiistructis, stylis laxe coalitis staminibus brevioribus. 



Var. SYSTYLA (Bast.). 



R. SYSTYLA, Bast, Suppl FL Maine-et^Loire{\^\2),\^,:i\', Woods, 

 Linn. Trans, xii. p. 230, Herb. 122-1^7; Smith, Eng. Flora, ii. 

 p. 395; Lindl. Mon. j9. 119; Borrer, Brit. Flora, edit. 3, p. 2^3; 

 Billot, Exsic. 1G63; Deseg. Mon. p. 24, Exsic. 3, 



R COLLINA, Eng. Bot. t. 1895, non Jacq. 



A busli 8 to 12 feet high, with elongated arching branches. 

 Prickles nuiform, scattered, the base deltoid, with shorter and 

 less-hooked points than in canina, the main ones | inch long and 

 the scar as deep. Stipules naked, or nearly so, on the back, 

 finely gland-ciliated. Leaves of the barren shoot 3^-4 inches 

 long, with 7 leaflets, the terminal one 15^-18 lines long by two- 

 thirds as broad, oblong, pointed, roundoifct the base, the upper 

 surface naked, the lower thinly hairy, pifucipally on the ribs, not 

 at all glandular; the serration quite' simple, moderately acute; 

 the petiole thinly hairy, not setose, witii 2-3 slender uniform 

 curved aciculi. Plowers generally 3-G in a cluster; the pedun- 

 cles often an inch or more long, thinly clothed with weak aciculi 

 and setse. Calyx-tube ovate-urceolate, naked or slightly setose 

 at the very base. Sepals |-|- inch long, naked or a little glan- 

 dular on the back, the main ones lengthened out at the point 

 and copiously pinnate. Corolla pale pink or rarely white, 15-18 

 lines across when fully expanded. Truit broad ovate-urceolate, 

 or subglobose, \-^ inch long by ^-\ incl) broad, not ripening till 

 October, the sepals falling before it changes colour. Column of 

 styles glabrous, generally protruded 1-1 1 line beyond the very 

 prominent disk. 



This is easy to recognize from R. canina in the typical form, 

 but is quite connected with it by intermediate gradations. I 



ha 



ve seen a form in which the column of the styles was pro- 



truded in the central flowers of a cluster, but not in the outer 

 ones, and other specimens with the styles not at all protrudeo, 

 for the rest quite doubtful between this and coUina. Our ordi- 

 nary Englisli plant, as just described, is exactly the sj^sti/la of 

 I>c\se'glise ; but sfylosa and leucochroa are mere varieties oi tlic 

 S'^itne plant, and both names have a slight priority over sj/slj/Iu. 

 I have seen this variety from Kent, Susses, the Isle of Wight, 



