350 JOHN BOIT
statement that she stood along the shore about three or four leagues from land shows that it was the mainland the depths he gives could not be gotten at that distance from the Queen Charlotte Islands' shore. Through Hecate Strait, up which she is sailing, there is a current of one to two knots in a northeasterly direction. Thick fog comes on and a heavy gale from the southeast, which con- tinue for two days, during which the Columbia beats to and fro. On the 3rd August the fog lifts, an indifferent observation is taken, and the ship is found embayed. This spot, it is suggested, was the stretch of water immediately to the eastward of Prince of Wales Island. Captain Gray then, says Hoskins, determined to stand "to the northward through what was supposed to be a strait between the continent and some islands." Upon the foregoing hypothesis this strait would be between Prince of Wales Island and the Gravina Islands. After being driven in that direction by the southeast gale for. three hours, land was seen in every direction but the westward; this means that in the fog Kasaan Bay, which lay ahead, seemed open water. The ship pursued her course and two hours later "the land was seen close aboard"; then a passage opened to the northwest. This, it is submitted, is Clarence Strait; but, being directly to leeward, it was thought unwise to enter it, and an effort was made to find a harbour on the weather shore. Reaching that shore the eastern iBoit says a small opening appeared, the Columbia made for it, and anchored under a point of land in Port Tempest. If the other suggested identifications be correct, this opening will be the western
Revillagigedo Channel is about 55 25' north and 131 45' west; the difference is no greater than is to be found over and over again in the journal, though the longitude is usually too far east.
Hoskins describes Brown's Sound, as the stretch of water in the neighborhood of Port Tempest was called. The above suggested site of Port Tempest fits his description exactly. After stating that the sound has many arms, ne proceeds to specify them; the geography of the suggested locality answers the requirements in an appealing and convincing manner. The first branch that Hoskins mentions is that upon which Port Tempest was situated; it trended east inclining to south as far as the eye could reach; this is Revillagigedo Channel. There was another arm extending to the north; this is Behm Canal. Then another running in a northwesterly course "up which the natives informed me was a village called Cahta"; this is Kasaan Bay, on which the Indian village of Cahta or Carta exists today. There was another arm leadin^ west northwest, in which the horizon was clear to the limit of vision, and which he took to be a strait; this is Clarence Strait. Finally there was another arm stretching to the west southerly, "up which I was informed by the natives was their village of Sushin"; this is "Choi- mondeley Sound, on which is the abandoned Indian village of Sushin or Sushan, Chasina or Chachina! Having disposed of the large branches, Hoskins adds that there were other smaller inlets; these are Baker Inlet, Skowl Bay, etc.
The general geography of the suggested site of Port Tempest bein~ thus shown to correspond with Hoskins' description, it will now be shown that the suggested spot fits also with the details that he furnishes. At Port Tempest, he tells us there were "also two small islands which afforded some little shelter"; and at the western end of Revillagigedo Channel is Guard Island, which is described in the Alaska Coast Pilot, 1883, p. 82, as "consisting of rocks uniting at low water two low, rocky, high-water islets, one west from and considerably larger than the other, and both bearing shrubs and a few trees." Vancouver also men- tions them as "two small islands with some trees upon them, S 40 W, half a league distant" from Point Vallemar. Voyage, vol. 4, p. 184, 8vo. ed. Again, Hoskins says that from Port Tempest "the land to the northward was about four miles and that to the southward one mile distant." Here also the geography agrees. From Guard Island, at the western entrance of Revillagigedo Channel, to Cape Camaano, the nearest land to the north, the distance is four and a half miles, and from that island to the western end of Gravina Island Point Vallemar, the nearest land to the southward is about two miles.
finally, in these small matters, Hoskins and Boit both mention the
At that distance lies Ward Cove. "A creek forming a small bank at its mouth
falls in at the head of the cove"; Alaska Coast Pilot, 1883, p. 81. The method
of fishing shows that the "river" was only a brook, for the men waded into the
water, threshing it with long poles and scaring the fish down the fall, where
they were gaffed with harpoons, boat hooks, etc.
Leaving the geographical portion of the subject, a word may be said ethno- graphically. The Indians met in the vicinity of Port Tempest, as Hoskins records, spoke the same language and had many of the customs of the natives of Queen Charlotte Islands. They were therefore plainly the Kaigani Haida, an intrusion into the Tlingit territory, occupying the southern part of Prince of Wales Island
