CRITTCAL NOTKS oN SUMK HHITANNIC SA.VIFK.VGE>; l.")7 



Western Highlands of Seotlaiid exactly agree with Maekay's (wlio 

 first found it on Brandon, in 1805) ; these I have not seen, but there 

 is some resemblance between the vegetation of those jjarts and that of 

 western and northern Ireland, owing to the mild climate. 



Kerry : — Abundant on the Brandon range ! and the Reeks ; near 

 Sneem and Waterville ; (Jlencar ; Slieve Mish range ; also reported 

 from Killarney, but Mr. Scully is sceptical. Ranges from 1500 to 

 8150 feet. Clare : — Great Isle of Arran [Aran], 1850, Melville in 

 Herb. Syme ! Tipperary : — Galtees ! ; very probably the " spon- 

 hemica'''' of the Comeraghs (Co. Waterford) may be identical. 

 Donegal ? :—Aranmore, 1881, H. C. Hart (Fl. Bonecjal). 



An allied, but distinct species, first observed by Mr. Druce in 1906, 

 of which I have seen cultivated material (1907) from his herbarium, 

 and which was gathered on the summit of Brandon Mountain, I found 

 rather frequently on the upper part of that range, in 1911, and have 

 had it under constant observation in my garden, ever since. Although 

 reluctant to increase the number of names, I hope to describe this, 

 later on. ' It has been drawn by Mr. E. W. Hunnybun. 



S. spoxHEMiCA C C. Gmelin, Flora Badensis Ahatica, pp. 224-6. 

 >S'. quinciuefida Baker, I.e., in part; Engler, I.e., p. 188; non 

 Haworth I — Unknown to D. Don, Haworth, and Smith ; but the 

 first-named had Grmelin's *S'. condensafa, which is only a form of this, 

 in cultivation. The very minute original description of S. sponliemica 

 gives a good notion of the characters ; but it is over twenty years 

 since I examined material from the locus classieus at S. Kensington, 

 and in such a case one cannot safely trust to memory. However, 

 Haworth definitely records S. condensafa as having been" found by his 

 father " in montibus Scotise " ; and I think that the true plant does 

 occur rather freely in some parts of Gi-reat Britain and Ireland. 



S. QUiXQUEFiDA Havvorth, Misc. Nat. 163 ; Ennm. Saxifr. 80. 

 S. sponliemica Baker, non Gmelin. — This was unknown to D. Don 

 and Smith, though the former ma}^ refer to it, among the synonyms 

 of S. pedatifida (p. 414) : " S. quinquefida, var. Lam. Fl. Fr. \\\. 

 p. 588 ? "; and the latter, in Engl. Fl. 280, quotes " S. quinquefida 

 Donn Cant. ed. 5. 107." Havvorth at first placed it next to S. gera- 

 nioides, remarking that it was affine to that, but far smaller; after- 

 wards he put it in ** PedatiJldcB, but inserted his >S^ viscosa between 

 them. It was thus placed by its author in quite a different group 

 h\ni\ S. plafi/pefala ; he included this in *** Quinqne/ida', to which 

 our other Hvpnoid segregates belong. 



In the Supplement to English Bolang, ed. 8, pp. 188-5 (1899), 

 Mr. N. E. Brown ably and impartially reviewed the status of this 

 jjlant, giving a full English description, based on G. Don's specimens. 

 His conclusion — certainly right, I believe — is that it is not 8. spon- 

 hemica ; and he thinks that it comes nearest to >S*. geranioides var, 

 ladanifera Gren. & Godr., though not identical. 



Last autumn I was able to have a few hours' work at the Kew 

 Herbarium, mainly \w order to examine the material in this case, 

 /S*. quinquefida is only known from cultivated examples, alleged tO' 

 have been found by the elder Don before 1801 (it is noteworthy that 



