Ethnographical collections from East Greenland. 361 



SO large, that the resembhince with a pear is no longer present; 

 Petitot compares the form with a Grecian cross ^). This only applies, 

 however, to the inner construction of the house. The outer is dome- 

 shaped, covered by earth and snow. The light comes in through a 

 window, which is placed on the top of the dome, an arrangement 

 unknown further east, so far as I am aware, but it agrees with the 

 house form of the western Eskimo, where the entrance is often 

 through the roof. This does not mean, that these houses lack the 

 usual, long entrance (underground in the west). 



It is a question, whether there exists a genetic relation between 

 the dome-shaped on the one hand and the cross-shaped or pear- 

 shaped type on the other. I am of the belief, that both forms are 

 equally typical of the winter dwellings of the Eskimo, as they occur 

 side by side over the greater part of the Eskimo territory, the 

 former however predominating towards the west (Bering Strait), the 

 latter towards the east (from Mackenzie River to Davis Strait). The 

 material has practically determined the development of the dome- 

 shaped type; the use of whale ribs in place of tent-poles must ne- 

 cessarily lead to this form (cf. Frobisher's description above p. 359, 

 footnote) and it thus belongs naturally to the districts where the 

 Eskimo have been keen whale hunters. But the other type, the 

 house of the Mackenzie River Eskimo, is probably more deeply 

 rooted in the early history of the Eskimo people, as it (as I think) 

 must be connected with the Siberian "earth-tent", the winter-dwelling 

 of the Ob-Ugrian peoples (Ostyakes, Wogules, Samoyedes and others) 

 as described in recent years by Sirelius^). The rectangular, central 

 structure in the Mackenzie house, with the sloping beams of the 

 side wall leaning against its uppermost wooden frame-work and 

 with the flat roof above, might be an American expansion of the 

 "earth-tent". The fact, that it is not known from the northern part 

 of Alaska is of no importance, since archæological investigations 

 with excavation of ruins have not been carried out there. The 

 characteristic roof-supports in the house of the eastern Eskimo 



^) Petitot: Vocabulaire Français-Esquimau (1876) pp. XXI — XXII; Les Grands 

 Esquimaux (]887j pp. 49—50; figs. IV and V. 



2) Sirelius (1906) pp. 74—104, (1907) pp. 120—127 and (1909) pp. 17—113. — 

 H. P. Steensby in his work on the origin of the Eskimo Culture (1905) has 

 made a one-sided comparison of the Eskimo house, viz. the house-type of the 

 Mackenzie River, with the hut of the Prairie Indians (Mandan Indians). The 

 resemblance does not seem to me striking and the argument with regard to 

 the origin of the Eskimo type looses its value, because the author has not taken 

 the Siberian house and tent forms into account. Comp, my review in "Geografisk 

 Tidsskrift" (Copenhagen) 1907—08, pp. 219-220. 



