1898-1902. No. 2.] VASCULAR PLANTS OF ELLESMERELAND. 71 



latter name instead of S. groenlandica, L., a name which is quite as 

 old as the name caespitosa (Sp. plant., Ed. I, p. 404). Even if it has 

 not from the beginning had reference to the entire range of forms of 

 the species, it must nevertheless be used for it and not for one of the 

 forms only as is done by Engler, 1. c, and others. In Engler & 

 Prantl, Pflanzenfam. Ill, 2a, p. 55, Engler again has used S. deci- 

 piens, Ehrh. for the main species, iS. caespitosa, L. is said to exist only 

 "in den arktischen Landern und auf den Rocky Mountains". I have 

 not been able to find the original description of Ehrhart's S. decipiens; 

 it does not exist where Engler (1. c, p. 186) quotes it from, Ehrhart, 

 Beitr. Naturk., V., p. 47. But even if it should apply better to the entire 

 series of forms within the range of the species, the oldest name must, 

 in all circumstances, be kept up, and for the above-mentioned reasons I 

 consider that, as there are two equally old names given by Linnaeus, the 

 right course to lake must be to give preference to the name S. groen- 

 landica, which can only apply to the arctic and sub-arctic plant here 

 in question, and not to any of the similar species of central European 

 mountains. In the herbarium of Linnaeus, some parts of which I have 

 had occasion, through the kindness of the Secretary of the Linnaean 

 Society of London, Mr. B. Daydon Jackson, to inspect, there is no "S. 

 groenlandica" to be found, and "*S'. caespitosa' is only represented by 

 one specimen from Idwell in Carnarvonshire, which is the plant com- 

 monly understood by that name, and another quite different one, without 

 any note about its origin. That Lapeyrouse (F1. Pyren.) and others have 

 used the name S. groenlandica wrongly, is of no consequence. 



Under his S. caespitosa, Engler enumerates a great many diffe- 

 rent forms, that are very difficult to keep distinct. One of the best 

 characterised, however, seems to be the one which Rob. Brown in Chlor. 

 Melv., p. 16, describes as S. uniflora (he had already used the name 

 without description in his List of pi. in Ross, Voyage I). The author 

 himself says about it : "Nimis affinis S. caespitosae Linn. ; vix distincta 

 species". Engler has it, as far as I understand him right, as a form 

 under S. decipiens var. caespitosa (cf. 1. c, p. 190). Too much stress, 

 however, must not be laid upon the characteristic expressed in the name, 

 but the description must be given approximately the form it has in 

 Lange, 1. c, p. 62. 



aS. groenlandica, L. var. uniflora, (R. Br.) m. Humilis, pulvinata; 

 foliis radicalibus aggregatis, trifidis, cuneatis, breve petiolatis, laciniis 

 obtusis; foliis caulinis Hnearibus vel inferne lobis lateratibus angustis 

 instructis; flore unico (vel 2—3): laciniis calycis obtusis; petalis albis, 



