A eurvey of the Phytogeography of the Arctic American Archipelago 161 
found with fuUy developed fruit, I think he attributes to much importance to these 
finds. The rare occurrence of the plant decidedly shows that it is a reHct, reduced 
to grow ill a few favourable stations, but no longer capable of spreading, notwith- 
standing the production of seed, and its relict state would be still more pronounced 
if such a distribution was combined with development of seeds every summer. But 
that is what, as far as may be judged from the available observations, takes place 
in the case of raost relict species in the Ellesmereland flora, among them Empetrum. 
Not to make the discussion to voluminous, I shall only take a few typical 
cases. The Seagull Rock in Harbour Fjord is a cliff, some hundreds of feet high, 
with a Southern exposition and sheltered from cold winds; here two species have 
their only known locaiity in the land, namely Rannnculus affinis and Chrysosphnium 
alternifoUum. Both grow there in great abundance and vigour, and both fruited 
profusely. A httle way off in the same slope I found Arenaria eiliata, not known 
from a single locaiity for several degrees of latitude to the south. On the same 
roekledges as the firstmentioned plant grew also Arnica alpina, and to tind this 
again we have to go up to the far north, to Discovery Harbour. When I found it 
in the first days of August, heads in flower and in different stages of fructification 
were seen in abundance, as well as old stalks from the preceeding year, having 
shed their achenes, most probably during the autumn and winter. Greelt (Report, 
p. 14) found it — most probably on Mount Cartmel — in flower already June 27. 
Hart, Bot. Br. Pol. Exp., speaks repeatedly of Discovery Harbour as far richer in 
species then the surronnding country, and p. 112 he says; «Iet the climate improve, 
and the Discovery Bay flora may spread by seed, &c., in all directions». The 
most favourable localities here are, as stated by Hart and Greely, the southern 
slopes of Mount Cartmel and Bellot Island. 
Now Hart, indeed, thioiks that the plants of these locahties, among which are 
also ErigeroH compositus and E. unijlorus, Androsace septentrionalis, Agropyrum 
violaceum, and Equisetum variegatuw, have arrived there »accidentally», but certainly 
the fact found is to be éxplained in quite another manner, viz., plants of a for- 
merly more general distribution have here found their last places of shelter, and 
as soon as the temperature of summer should be lowered to a certain point, several 
of them, entirely dependant on seeds even for local propagation, would disappear 
from the flora, some others might perhaps linger for some time by means of 
vegetative multiplication. Spread beyond the favoured shelters, they have once 
found, they are unable to do, even by means of seed, therefore their capacity of 
fruiting does not at all contend with their nature of relicts. The southern slopes 
and roekledges in Harbour Fjord, in Discovery Harbour, as well as several other 
similar ones in other parts of Ellesmereland, and doubtless all over the Archipelago, 
are typical reHct localities, quite comparable to the many stations of southern plants 
in northern Sweden which Gunnar Andersson and Birger have so admirably 
treated in their great memoir on the flora of that region (Norrl. Flor. Fördeln.). 
