132 DECAMD. DIGYN. 



HAD. Rooks in the western Highlands, G. Don. p. Scotch Alps, 

 G. and D. Don. y. Banks of a rivulet not far from Airly Castle, 

 Angus-shire, G. Don. &. Highest mountains of Angus-shire, G. 

 t)on. Fl. June, July. 7/ . 



Varieties without end might be produced of this plant, which, as I 

 have already intimated, I can find to differ in no essential point from 

 the last species, if it be not in the always 3- or 5 -fid leaves, and 

 perhaps the .somewhat greater length of their segments. Nothing 

 is more variable than the pubescence on the stems and leaves, and 

 the extent of the sterile shoots. The > Jlowers,, too, upon each stem 

 vary in number, and the petals somewhat in shape. The leaves 

 are, in all, more or less acute in the lower ones, and generally tipped 

 with a short bristle in the upper ones, as in S. hypnoides. In none 

 of these marks, therefore, can I see any thing to found specific cha- 

 racters upon, as Smith and Don have done. I am not even satis- 

 fied that the S. pahnata of E. B. is distinct. It is certainly the de- 

 cipiens of Ehrhr. and Sternberg 5 and an impartial examination of 

 the beautiful figure by the last mentioned author (Sax. t. 23.), 

 with the equally good one of his S. Sternbergii (our hirtri), t. 24, 

 will, I think, satisfy any one that they also are but slight varieties of 

 the same plant. The former is the more hairy of the two a . 



13. S. pedatifida (pedatifid Saxifrage), lower leaves and those 

 of the rather short sterile shoots upon very long foot-stalks^ 

 divided into 3 deep linear lanceolate acute spreading segments, 

 the lateral ones bind, the superior leaves of the much branched 

 flowering stern linear undivided. . B. t. 2278. .D. Don, 

 MSS. ined. S. quinquefida, Haw. in Misc. Nat. and Donn 

 Hurt. Cant. 



HAB. Rocks near the head of Clova, G. Don. It is stated also to be 

 found in the Highlands by Mr. J. Macku;/, but without any parti- 

 cular station being given in E. B. Fl. May, June. 3/ . 



A very distinct species from any British one, nor does it appear to be 

 noticed in Sternberg's fine work, though coming near to his S. la- 



a Since the above remarks were written upon this most intricate fa 

 mily of the Saxifrages, it was with much satisfaction that I saw, in the ar- 

 ticle Saanfraga in Rees'a Cycfofldfdia, that Sir James Smith has, with that 

 degree of candour which so often accompanies his writings, acknowledged 

 that he considers S. hirta to be only a var. of the Linnaean. Ceespitosa. A 

 comparison with Linnaeus's description, and especially with the admirable 

 description and figures of Gunner in his Fl. Norvegica, will, I think, make 

 it clearly appear that they are the very same. The figure of S. ccesp. mE.B. is 

 however a dwarf and densely leaved var. the same as D. Don's condemata, 

 only more pubescent, and precisely the S. grccnlandica of Linn, and Gunn. 

 Norv.t. 7>f' 1.: under which variety, as Smith has himself considered the 

 groenlandica'm the Cyclop, to be, he should have quoted the E. B. ccespitosa. 



I must do Sir James Smith the further justice to state, that in the same 

 work, he doubts if his pahnata (E. B. t. 455.), the declpiens of Ehrhr., be 

 truly distinct from ccespitosa. His S. platypetala he says is distinct from its 

 allies, unless it be a variety of hypnoidcs. Of S. elongellahe tells us that the 

 greatest peculiarity is in its solitary 1-flowercd peduncle. But some of my 

 specimens from the discoverer himself, Mr. Don, have 2 and even 3 flowers 

 upon the same stalk. 



