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



the flowers are said to be pink, and Lange asserts the same (1. c, p. 76), 

 where he moreover finds fault with the statement of Durand, PI. Kan., 

 thai [ho colour is yellow and supposes that the latter has come to this 

 conclusion through having seen only dried specimens (which probably 

 was the case also with Lange himself). However most of the specimens 

 of P. hirsuta that I saw at Godhavn and Egedesminde had flowers 

 that were of a very pale yellow or whitish colour, and in Ellesmereland 

 also this was far more common than the pink. This also agrees better 

 with the fact that in some cases P. hirsuta has been mistaken for 

 P. Oederi, Vahl. 



Such specimens must doubtless have induced Hart to mention 

 P. flammea, L. from Hayes Sound. I neither found it there, nor do 

 any specimens in the Nat. Hist, Mus. or at Kew confirm the statement. 

 However, it is not always easy to decide, with the help of the above 

 mentioned herbaria, how the collector or original author has himself 

 determined the specimens, as the label generally contains only the loca- 

 lity where the plant is found, not any specific name, and their present 

 arrangement can consequently very well be of a later date. Still I 

 think myself justified in concluding from the non-existence of specimens 

 in the London herbaria, that at first the plant was wrongly determined 

 but had since been put in its right place. On the other hand, some 

 authors who have rightly identified the P. hirsuta with yellow flowers, 

 have taken the form with pink flowers for another species. Such was 

 the case with a specimen from Shift Rudder Bay, leg. Feilden, with 

 "P. sudetica, L. (P. Langsdorfi, Fisch.)" on the label and also with other 

 specimens of Feilden and Hart in the Kew herbarium, some from 

 Danish Greenland, where the real P. sudetica, L. is entirely lacking 

 (cf. also Simmons, Dan. Greenl. pi.). 



A confusion of P. hirsuta with P. lanata, Cham. & Schlecht., has 

 also sometimes taken place, to which I shall have to refer again under 

 that species. Indeed the likeness between these two species is small 

 enough and the mistake is only possible for one who has not studied 

 the plants in question from nature. 



P. hirsuta differs widely from P. lanata in the slenderer growth, 

 the thinner-set spike which is far less hairy, and especially in the very 

 remarkable dilatation of the petiole and rachis of the leaves. The co- 

 rolla has a sack-like, contracted galea that almost entirely encloses the 

 stamens and the style. Sometimes the galea is cleft in the middle (in 

 dried specimens this is not uncommon) so as to give it the appearance 



