Page:Narrative of the Discoveries on the North Coast of America.djvu/328

298 but it escaped. Esquimaux marks stood upon the heights, but no recent traces of inhabitants could be found.

We next morning cut across the eastern shoulder of Cape Alexander, to Musk-ox Lake, which lies in a valley. It is half a mile long, and empties itself by a subterraneous channel, through a steep ridge of shingle, into another basin, about half its size, which was frozen to the bottom. Crossing the ice, we forded the little stream below, which, like many others, still retained drifts of snow on its banks. Our rough route led amongst large boulders, and through wet mossy tracts producing dwarf willows. The immediate coast-line continued flat, but skirted as before by low stony hills. Some ice lingered in the bays, but the sea was quite open. At the distance of nine miles we crossed another bluff cape, composed of trap rocks, where an observation gave the latitude 68° 52′ 19″ N., variation 63° East. This was the greatest deviation of the compass from the true meridian. From Boathaven to Cape Franklin the variation increased very last, since only nine miles beyond that cape it was found to be 60°. Thence advancing eastward, it fell off to 56° 30′, and again augmented as the coast trended more northerly; while from Trap Cape to our