An Anthropogeographical Study of the Origin of the Eskimo Culture. 199 
According to verbal information from Кмор RASMUSSEN there exist in 
Melville Bay old house-ruins of an oval form which according to his in- 
formation seem probably to be of this type. How far ruins of a similar 
type occur south of Melville Bay is not known. K. BIRKET-SMITH has 
lately described some ruins from Southern Greenland which he likens to 
houses from Baffin Land which are indubivable earth-tents. He from 
GEORGE BEsT’s “True Discourse of the three Voyages .... under the con- 
duct of MARTIN FROBISHER,” first cites a very striking description of such 
semi-underground houses with a framework of whales” bones and a 
roof of skins, and then he tries to show that such houses have existed 
on the Greenland side of Davis Strait", BIRKET-SMITH, however, does not 
distinguish, as we do here, between a dome-shaped type of house which 
originates from the tent, and a rectangular house of quite a different origin. 
Therefore he can compare the house from Baffin Land with the house 
ruins from Julianehaab which for the rest he apparently correctly 
considers to be old rectangular small houses from the time before the 
common house. Yet, further, I agree with him when he expresses some 
doubt as to whether the house-ruins found by V. C. FREDERIKSEN? 
at Holsteinsborg really originate from the pear-shaped house type 
(trefoil houses, as FREDERIKSEN calls them). I could imagine that in 
reality they were of some more modern house type. 
The mentioned type of house which ought rather to be called the 
Eskimo winter tent or “earth-tent” than the dome-shaped type conse- 
quently seems to have a very sporadic distribution, which corresponds 
with the fact that it is an old form of dwelling which once had a greater 
distribution, but which is now superseded by other types of houses. In 
Alaska and at the Mackenzie, as also eastwards and especially in Green- 
land, it has been superseded by forms of houses which originally have 
come from Asia. In the districts of the Central Archipelago it is the 
snow house which has won. 
The snow house, the most characteristic form of dwelling of the 
Eskimo culture, must in its origin, however, be apprehended as being 
brother to the winter-tent. Even Мовросн? expressed the opinion that 
in all probability the bee-hive-shaped snow house must be regarded as 
succeeding the tent. In 1905* I tried to show that the snow house 
must have developed from the dome-shaped tent, as for winter use this 
was covered with snow, as it still is with the northern Indian tribes. 
But besides the dome-shaped tent there is still one form of house which 
possibly may have played a råle as a precursor, and at any rate is a 
kind of parallel to the snow house. It is the round house which is dug 
into the earth and is partly built of turf, and which occurs in North 
1M. o. G., Vol. 53, pp. 9 sqq. 
? FREDERIKSEN, pp. 391 sqq. 
3 Мовросн, II, p., 127. 
a STEENSEW.; 1, po 194, 
