A survey of the Phytogeography of the Arctic American Archipelago < 1 
(Journ. I, p. 273), Navy Board Inlet, Lyall (KH), Possession Bay, Fisher (Journ., 
p. "62), Port Bowen, Parry (KH, LS); Melville Island, Sabine (NHM), Beverley 
(KH), M'Clintock (KH), M'Millan (ex Macoun, App. Arctic), Hennessey (ib.), Winter 
Harbour, Pakry (Journ. I, p. 182), Bushnan Cove, Parry (1. c, p. 199), M'Clintock 
(Add. Papers, p. 171), Depot Island, MTJltntock (Furth. Papers, p. 577); Bathurst 
Island: Evans Point, M'Dougall (Add. Papers, p. 279); Cracroft Sound, Osborn 
(Furth. Papers, p. 234); Cornwallis Island: Assistance Bay, Sutherland (KH); 
Griffiths Island, Markham (Add. Papers, p. 2o9); Nortli Devon': Cuming Creek, 
Borden (ex Macoun, 1. c), Powell Creek (KH), Beechey Island (KH), betwcen 
Beechey Island and Northumberland Sound (KH), Lyall. Mount Belcher, Cape Vera, 
West Fjord, Simmons, Boat Cape, Schei (ex Simmons, Stray Contr.); North Kent, 
SiMMONs (1. c.); Graham Island, Schei (ex Simmon.s, 1. c); Heiberg Island: Whit- 
sunday Bay, Big Island in Heureka Sound, Schei (ex Simmons, 1. c.) ; EUesmereland, 
very common, Hayes, Hart, Greely, Bay, Isachsen, Simmons, Bartlett (CH); Wolf 
(ex Rydberg, List Peary). 
Geographic area. Greenland, arctic and subarctic parts of North America, 
Rocky Mountains, eastern and arctic Siberia, Altai, Novaja Semlja. 
The long list of distribution not only shows that this plant has been observed 
in a great raany places, which raay be partly accounted for by its being the lar- 
gest lignaceous species of the Arcbipelago, or of the northern parts at least, but it 
certainly also implies its being present in every region of auy extent. Besides it 
is also a very variable plant, as already Hooker, in App. Parry II, has pointed out. 
I have seen places in Ellesinereland, where a considerable series of more or less 
different forms could be gathered growing together, and. my material from the 
SvERDRUP expedition would, I think, have offered unHmited opportunities of esta- 
Dhshing new names to a raanufacturer of bad species. I raight also adnjit, that 
there may be several forms of specific value included in the S. arcHca, such as it 
is here defined, but I never can admit that the right way is to put a new name 
on every form that happens to be included in a collection, especially when it is to 
be treated bv a ta^onomist, who has never seen the plant in question growing and 
is, cousequently, not aquainted with its variabihty and with the connecting links 
that may exist in its range of distribution, or even in a single locahty. A new 
species-name onlv ought to be established for a plant distinguished by a series of 
good (not relative) characters. But that is not the case with the forms hitherto 
described within the specific range of S. arcfica, Pall. 
The original siberian willow of Pallas, indeed, is described so as to include 
only the most broadleaved forms, but it is closely connected with the more narrow- 
leaved, which is most common in the american parts of the Polar Regions, and 
which Rob. Brown gave the same name as Pallas had used, without, however, 
knowing the plant of Pallas. Now most authors which use the name S. arctiea 
quote Brown as author, but I have not thought necessary to enter them in the 
synonymic on that account. S, aHghrmn Cha.n.. Brownii, Anderss., S. groen- 
