1898-1902. No. 2.] VASCULAR PLANTS OF ELLESMERELAND. 33 



several assurgent stems. This is also the case with most specimens 



from other parts of Arctic America; and even specimens collected by 



Chamisso at Unalaschka (in the Copenhagen herbarium) show a close 



resemblance; but I have also seen specimens collected by Kjellman at 



St. Lawrence Island, which have larger and more open flowers with 



longer tube, and somewhat longer upper lip, but for the rest, not 



differing from mine. Even my specimens have the upper lip somewhat 



longer than the lower one, which, although the case also with the 



original specimens, is not mentioned in the description of Rob. Brown. 



Besides the section-character, the bidentate galea, P. ardica differs 



from P. lanata in its laxer stems, in the broad rachis of the leaves — 



which reminds one very much of P. Mrsuta — in a shorter and less 



dense spike, in the less developed wool, in the long tube and upper lip, 



in the dark purple colour of the flowers and in the shape of the capsule 



which is much longer and less oblique, protruding with one-third of its 



length over the calyx, taking almost the middle place between that of 



P. lanata and P. Mrsuta. On the other hand, there is a resemblance 



in the protruding of the style from the galea. 



To P. Mrsuta it shows a great resemblance in habit and in the 

 shape of the leaves, even though the rachis is less broadened than in 

 that species. The flowers are much larger, longer, more open and the 

 galea is not nearly so cucullate contracted as in P. Mrsuta. Even the 

 protruding style forms a distinguishing character from the ordinary 

 specimens of that plant, where the galea i^ not split open. 



The purple colour of the flowers also forms a distinguishing 

 character, as far as Ellesmereland specimens of P. Mrsuta are con- 

 cerned, but I have seen some specimens from Spitsbergen that resemble 

 P. ardica so much, that I was obliged first to examine them more 

 closely to make sure that the plant really was P. Mrsuta. The flowers, 

 also, were unusually large. It is probably of this form that Nathorst, 

 (Nya bidr., p. 11) speaks, although he characterises the flowers as pink 

 (enbart Ijusroda). I was also reminded of P. sylvatica when I first saw 



P. ardica. 



The species grew on svampy ground with a dense vegetation of 

 Carices, grasses etc., and at the edges of pools, and was in full flower 

 in the beginning of July 1899, the only time I saw it. The shape of the 

 fruit could be judged from numerous capsules remaining from the 

 previous year, that doubtless had been ripe in the autumn. 



Occurrence. Only found at the outer part of Hayes Sound 

 (Buchanan Strait): Lastraea Valley (abundant, 858), Eskimopolis (847, 



