46 



building in vogne among the ancient inhabitants was to excavate sh'ghtly, 

 to build a wall of flat stones or of bones of the largei* whales, and bank this 

 up on the outside with turf and stones. In these ancient houses, there was 

 usually a door at one side, as in most Innuit houses, and as many of the 

 Aleuts jjractice even now. The enormous yourts, entered only by a hole 

 in the top and accommodating a number of families, were of more modern 

 invention, and are rarely found among the ruined villages. From throwing 

 out debris, and the gradual accumulation of material in the course of years, 

 the house being more or less resodded every autumn, the outside embank- 

 ment in the coui'se of time became elevated from four to six feet above the 

 level of the floor. The roof was formed of whales' ribs in default of wood, 

 covered with wisps of hay tied together and laid on grass-mats across the 

 rafters ; and all this was turfed over. Hence, when the house was aban- 

 doned the straw and mats decayed, the earth and finally the rafters fell 

 in (the latter being often removed to use in some new house), the rain and 

 storms diminished the angles of the embankment, and, finally, the only 

 evidence remaining would be a roundly rectangular pit, with steep sides, 

 somewhat raised above the surface of the external soil. This might endure 

 for generations without any practical alteration, as the stone walls within 

 would prevent caving in at the sides, and the filling-up of the pit by the 

 accumulation and decay of subaerial deposits would progress very slowly. 

 As the ancient Aleuts built their houses as close together as possible, the 

 surface which is left by the disappearance of the structures above described 

 is irregularly pitted all over with depressions from four to six feet in depth, 

 and varying from ten feet square to dimensions of forty by twenty feet, or 

 even much larger. There is usually, on the highest point of the bank or 

 knoll where the village stood, a pit much larger than the others, which was 

 probably the workshop or kashim' of the settlement. Around this we usually 

 found tools and implements more abundantly than about the smaller pits 

 or remains of houses. We also found that the floors of the pits hardly 

 afforded anything until we reached the strata of the shell-heap upon which 

 the houses had been erected ; while the outer embankment, containing 

 everything which had been thrown away, was correspondingly rich. 



We therefore adopted two methods of procedure. When stormy 



