122 H. G. SIMMONS. [sec. arct. exp. fram 



sparingly. This form is at home in somewhat moist clay-fields and 

 slopes, up towards drier places it goes over into forms 1. and 2. This 

 most resembles C. Edmondstonii, and I have seen specimens of it 

 collected by Hartz at Scoresby Sound, which (Fan. og Karkr., p. 327) 

 are mentioned as connecting var. caespitosum with the typical C. alpinum. 



4. A small-leafed, condensed form, growing in semi-orbicular tufts 

 or dense mats. It agrees entirely with Malmgren's description of his 

 variety, has succulent leaves which are quite glabrous or at most furn- 

 ished with a few ciliae. I always found it sterile. Nathorst says 

 (Nya bidr., p. 22) that var. caespitosum is more common in the northern 

 coasts of Spitsbergen than G. alpinum, and that it is rarely flowering, or 

 appears to be so, because it flowers late. Andersson & Hesselman on 

 the contrary, speak of C. alpinum as flowering later than C. Edmond- 

 stonii and its variety. I think that these different statements are due 

 to the confusion under var. caespitosum of two separate plants, viz., 

 the analogous varieties of G. alpinum and G. Edmondstonii, of which 

 the latter probably flowers earlier and more abundantly, whereas the 

 former seems to flower very sparingly and then late in the season. I 

 have not had an opportunity of observing it in greater numbers after 

 the beginning of August, but even old inflorescences were sought for 

 in vain in the extreme f. pulvinata. The form 3. also, was rarely 

 found in flower, and only late in the summer. The previously-described 

 glabrous form of G. alpinum, var. glabratum, Retz., has long inter- 

 nodes and narrow leaves, and flowers normally, that is to say, it is 

 different from the form here in question in every respect. I think that 

 the Ellesmereland form is characteristic enough to get a name and there- 

 fore call it f. pulvinata. It is this form, and not the var. glabratum, 

 which is the counterpart within the formseries of G. alpinum of the 

 var. caespitosum within that of G. Edmondstonii. 



The plant which Hart, Bot. Br. Pol. Exp., p. 28, mentions as "G. 

 laiifoUum L. {G. caespitosum Malmgren)" belongs, principally at least, 

 to the last-treated form, partly probably also to the form 3. I have 

 seen several specimens in the Nat. Hist. Mus. as may be seen under 

 the occurrence of the forms. The f. pulvinata grows on wet clay fields 

 and its large dense tufts are quite filled with loam. Doubtless this form 

 is directly produced by the local circumstances. It also grows in com- 

 pany with other plants of the same growth, such as Alsine Bossii, 

 Saxifraga oppositifolia f. pulvinata, and others, and is very singular 

 in appearance, but it goes gradually over in the form 3., which is again 

 connected with the other two. I never saw it with flower and of course, 



