66 THE JOURNAL OF BOTAKY 



Exsiccata :— Herb. Marshall, Nos. 3651, 3654, 3657, 3658, 3659, 

 3900 (this is 3651, cultivated). 



I have seen no specimens at all closely resembling S. Drucei in 

 either public or private collections, with the exception of material 

 collected by Mr. G. C. Druce, in 1906, and cvdtivated the following 

 year; these were labelled as 8. Sternhergii Willd., but cannot be 

 referred to that species, which is bright green, glabrescent, with very 

 different foliage, sepals, and petals. I saw it on the upper part of 

 Brandon Mountain, S. Kerry, in several places, from about 2800 feet 

 to the summit (3127 feet), in July, 1911, and have grown and 

 studied it ever since. Most likely it will be found in other Kerry 

 stations; perhaps, also, on the Galtees, &c., where S. liirta Sm. 

 occurs. 



My root (No. 3900=3651) forms a very close tuft, and is now 

 about six inches across ; but some wild gatherings are a good deal 

 more straggling in habit. The petals tend to be orbicular-ovate, on 

 first expansion : but their average shape is (as described) obovate. 



8. Drucei seems best placed next to 8. incurvifolia D. Don (in 

 the group of S. cespitosa L.) ; but that has deep green, glabrescent 

 leaves, &c., and forms smaller, more crowded tufts. Grown side by 

 side, these two are obviously distinct at all seasons (the Kew " S. in- 

 curvifolia,'" kindly sent by the Curator last May, has not yet been 

 seen by me in flower, but is extremely hairy, and can hardly be 

 D. Don's species). ^S*. liirta Sm. is grej^-green, much more villous, 

 laxer in habit, with very different foliage and sepals, and larger 

 petals. 



2. S. Steeis'BEEGTI Willd., n. var. geacilis, mihi. — A typo con- 

 stanter differt habitu rigidiore : foliorum segmentis linearibus vel 

 lineari-lanceolatis, acutis, aut etiam acuminatis ; necnon sepalis angus- 

 tioribus, acutissimis, pi'o more apiculatis. 



Differs constantly from the type by its more rigid habit : by its 

 leaf-segments being linear or linear-lanceolate, acute, or even acuminate ; 

 and \)j its narrower, very acute, as a rule apiculate sepals. 



Krsiccata: — Herb. Marshall, No. 4406. 



Habitat -.—mAC^ Head {B.P. Murray; H. C. Levinge; S. H. 

 SickJiam) and Ballyvaughan (JT. C. Levincje), Co. Clare ; Brandon 

 Head, S. Kerry {Arnold Elliott, through H. S. Thompson). Doubt- 

 less it occurs elsewhere, as it appears to be the prevailing Irish form 

 of the species. 



Some of the wild specimens are much crowded, and bear very 

 numerous flowering steins ; I think that such is usually the case 

 (more or less) in this variety. My No. 4406 (from Mr. Biekham), 

 grown for about a dozen years under a north wall, is much drawn-out, 

 and represents a " forma nmhrosa,''' The Brandon Head plant, like- 

 wise cultivated, is rejnarkable for its oval-oblong petals, tinged with 

 pink, which I have not seen in <iny other example of this species. 

 Mr. Murray's, grown on a sunny rockery at Shapwick Vicarage, was 

 referred, in 1889, to Professor Babington, who wrote as follows : — 

 " I think I may call it Sterjihergii^hirta of Smith, E. B. 2291." 

 This identification ip certainlv erroneous ; the latter, of which I have 



