The "zimniki" (in Russian) or winter houses were throughout the region of one general type. They were fair-sized circular semisubterranean houses, made of driftwood and earth, and provided with a semisubterranean entrance vestibule. Their remains are characterized everywhere by a circular pit with a short straight trench depression, the same pot-and-handle type as found along the Yukon. Rarely for the construction of the houses, where driftwood did not suffice, recourse was had to whale ribs and mandibles. The "letniki," or summer houses, were constructed on the surface of wood, sod and skins, or of whale ribs and skins, approaching on one hand the summer huts of various continental tribes and on the other the "yurts" of the north Asiatic peoples. The "kashims," or communal houses, were built, much as on the Yukon, like the family dwellings, but occasionally quadrilateral and much larger. Smaller semisubterranean storage houses of driftwood and sod near the winter dwellings were seemingly general. Ruins of stone dwellings, without mortar, are said to exist in places on Norton Sound and Bay and on a lagoon near the western end of the Seward Peninsula. The few houses on the Little Diomede are made of loose unhewn stone slabs. The dwellings of the King Islanders are built on the rocky slope of the island on platforms supported by poles, all of driftwood. There is as a rule an absence of separate refuse heaps near the villages. The refuse apparently has been dumped about and between the houses rather than on separate piles. Dead villages abound. On consulting the older Russian records, however, it is seen that nearly all were still "living" as late as the early forties of the last century. Yet there are sites that were "dead" already when the Russians came, and the accumulations in other cases denotes a long occupation. The site of a dead village, in summer, is generally marked by richer and greener vegetation; same as on the Yukon. The site itself is usually pitted or humped in a line forming a more or less elevated ridge, or the pits may be disseminated without apparently much order. And there may be irregular mound-like heaps without external traces of any structure. In the older sites no trace of wood is visible; in the later rotten posts, crosspieces, parts of the covering of the house or tunnel, or even a whole habitation may be present. In the old sites the wood is hewn with stone axes; in the later it is sawed, and there may be nails. Older accumulations lie occasionally beneath more recent ones, though no interruption of continuity may be traceable. Of a superposition of villages no trace was observable. FOOTNOTES: [60] Tenth Census, VIII; reprinted in Compilation of Narratives of Explorations in Alaska. U. S. Senate Rept. 1023, Washington, 1900, 55-281.