A CLASSIFICATION OF GARDEN ROSES. 281 



bright green, moderately firm in texture, 2-3 ft. long, ^-f in. 

 broad low down, tapering very gradually to the apex, smootli on 

 the margin, with 15-20 close veins on each side of the midrib. 

 Peduncle as long as the leaves, ^ in. diam. low down. Eaceme 

 moderately dense, 3-4 in. long ; pedicels ^-^ in. long ; bracts 

 oblong, obtuse, about as long as the pedicels. Upper flowers bright 

 red, lower bright yellow. Perianth cylindrical, IJ-li in. long, 

 J in. diam. ; segments ovate, obtuse, l-12th in. long. Style 

 finally exserted ; stamens included ; ovules 10-12 in a cell. — 

 Baker in Journ. Linn. Soc. xi. 393, excl. syn. Tritoma Burchellii 

 Herbert in Bot. Eeg. t. 1745. 



Hab. Southern provinces, Thunherg ! BurchcU ! The plant 

 grown for many years under this name at Kew has a peduncle 

 spotted with purple, and much shorter pedicels than Burchell's 

 type, and connects it with K. Uvaria. 



18. K. CAULESCENS Balicr ; Hook. fil. in Bot. Mag. t. 5946. — 

 Caulescent, with a stem |— 1 ft. long, as thick as a man's thumb. 

 Leaves ensiform, acuminate, 2-3 ft. long, 2-3 in. broad at the 

 base, moderately firm in texture, very glaucous, with 40-50 very 

 distinct ribs, acutely keeled, the margin conspicuously denticulate. 

 Peduncle as long as the leaves. Eaceme dense, 4-6 in. long, 

 2-2|^ in. diam. ; pedicels |- in. long ; bracts oblong-lanceolate, J in. 

 long. Perianth an inch long, | in. diam. above the ovary, ^ in. at 

 the throat ; segments ovate, obtuse, l-12th in. long and broad. 

 Stamens and style all much exserted, 



Hab. Province of Albany, on the Stormbergen Mountains. 

 Introduced by Mr. T. Cooper into cultivation about 1860. Painted 

 by Miss North from a plant procured near Grahamstowu now in 

 cultivation in the Cactus-house at Kew. 



A CLASSIFICATION OF GAEDEN EOSES. 



By J. G. Baker, F.E.S.* 



A GOOD general monograph of the genus Rosa is needed. 

 Lindley's 'Monograph' was published in the year 1820, and since 

 then a great number of new species have been discovered, and a 

 very large number of books and papers have been written bearing 

 upon the subject in one way or another. The difficulty which one 

 finds at the outset in using Lindley's book is that his primary 

 groups are characterised so briefly, and that the points of contrast 

 which they present are not fully and clearly brought out into view. 

 The following key shows the best way of getting over these diffi- 

 culties which, after having had a large number of specimens 

 through my hands during the last thirty years, I am able to 

 suggest. The list of species is only intended to be exhaustive so 

 far as garden roses are concerned, with the addition of a few well- 

 marked types not yet brought into cultivation. What I have aimed 



* From the ' Gardeners' Chronicle,' August 15th, 1885, p. 199. 



