45 



modern rock-slieltcrs utilized for burial-places. There are nine villat^e- 

 sites on Captain's Bay alone. 



Chika Rocks, Ahutan Pass. — Here are remains of a small, but populous, 

 settlement, but no sliell-lieaps. 



On the islands to the eastward of Unalashka these remains are so 

 niunerous as not to be practicable to enumerate, except such as we actually 

 visited or have been specially reported to us, namely: Sannakh Islands, 

 village-sites very numerous; False Pass, two localities for village-sites; Port 

 Moller, Aliaska Peninsula, shell-heaps extending over twenty acres, village- 

 sites much less extensive. Unga Island, at Delaroff Harbor; Korovin 

 Island, Nagai Island and Simeonoff Island, amongthe Shumagins. Chiachi 

 Islands; Chignik Bay, Aliaska Peninsula, extensive village-sites; Chirikoff 

 Island; and so on to Kadiak Island and Cook's Inlet. 



Tlie population of the islands was estimated at fifty thousand by 

 Shelikoff", and, in view of the evidences of habitation, the estimate could 

 not have been excessive at one time, though perhaps too great at the time 

 he visited the islands. The present population is about two thousand. 



The village-sites or shell-heaps are indicated, as far as the eye can 

 distinguish vegetation, by their brilliant green covering of herbage, which 

 is only dimmed when covered by snow, and even in the height of spring is 

 brighter and more verdant than the adjoining slopes. 



This is the result of the fact that the shell-heaps are great mounds of 

 the most fertile material, which thousands of years Avould not suffice to 

 exhaust by the ordinary draughts of nature. Bones, shells, and all varieties 

 of rejectamenta having been deposited here for centuries, the covering of 

 soil which has accumulated over them is incomparably rich, and it has even 

 been suggested that the solid beds of compacted fish-bones, which are to be 

 found in some localities, might be quarried and exported as a fertilizer. 



Nothing is to be got from these deposits without extensive excavation 

 and patient search. 



Our usual method in investigating these accumulations was as follows : 

 The shell-heaps, especially those surmounted by village-sites, usually pre- 

 sent an undulating appearance, which from some neighboring elevation is 

 at once seen to result from the followino' cause : The method of house- 



