429 
divide, to the headwaters of the Stewart, Macmillan, and Ross rivers, 
where certain kinds of fur are more plentiful. In the spring they 
return to the Gravel river and build moose-skin boats, in which they 
descend tliat river to the Mackenzie. The Mountain Indians have 
hunted on the Gravel river for a long time; there are meat-drying 
racks everywhere along the stream banks. Some of their signs are 
very old, showing evidence of stone implements having been used. 
It was probably a long time before they grew bold enough to 
cross the divide, but even now they are careful not to go far down 
the streams on the western slopes for fear of meeting the fierce Yukon 
Indians; so that mutual fear and distrust have established a dead 
line over which representatives of neither side pass ” (Keele, Joseph; 
“A Reconnaissance Across the Mackenzie Mountains on the Pelly, 
Ross, and Gravel Rivers, A^ukon and Northwest Territories”; De- 
partment of Mines, Geological Survey, Canada, No. 1097, p. 11 F, 
Ottawa, 1910). 
Eskimo. Since Eskimo remains seem very rare along the north 
shore of the gulf of St. Lawrence, and the Iroquoians apparently 
controlled the coast in summer as far east as Kegashka, I have 
assumed that in 1525 Eskimo territory terminated on the mainland 
somewhere about the strait of Belle Isle, not at the Mingan islands 
opposite Anticosti, where they fought with the Micmac in the seven- 
teenth century. Furthermore, I have marked them along the north- 
west coast of Newfoundland on the strength of some recent investiga- 
tions by W. J. Wintemberg, who discovered between Bonne bay and 
Flowers cove numerous prehistoric Eskimo sites (belonging to the 
Dorset culture mentioned on page 230), but not a single Beothuk 
site. The Eskimo may have frequented, also, the northeast coast of 
the same island, for it was probably they, not the Beothuk, wRo 
worked the soapstone quarries at Fleur-de-Lys. They may have 
abandoned Newfoundland, however, before 1525. 
