ANCIENT SETTLEMENT ON TREWOETHA MARSH. 69 



... .If there are several families in occupation of the same house, 

 which is the rule, the main bench is divided by low partition- 

 boards into separate stalls for each family." " Each of the 

 families," he says elsewhere, "had its own partition marked off 

 from the common couch, and in each stall so formed, man, wife, 

 and children would be closely packed, a four-foot space thus 

 having to accommodate husband, two wives, and six or more 

 children," (F. Nansen, ''Across Greenland," 1890). 



We have in the largest hut precisely this arrangement, — it 

 is divided into two stalls for two families, each with its own 

 hearth. 



The houses must have been low, roofed with rafters brought 

 together in the middle, and covered with thatching of rushes 

 and turf. The smoke escaped through the roof. There is not a 

 trace of chimney or of window in any one of the huts. 



The upright slabs of granite pretty well mark the height of 

 the walls, they are usually thin and pointed at the top, and could 

 hardly support walling laid above them i n courses. Moreover, 

 the fact of the locker and ovens being either on the floor-level 

 or raised but a very few inches above it show that the inmates 

 worked in a crouching position. The doors were generally never 

 higher than 3 feet. I do not lay much stress on the low door 

 in Hut F, as that was fallen, and we may possibly have made 

 some mistake in re-erecting the stones, but other huts have the 

 jambs unf alien. 



That the huts were not in continous occupation through a 

 long tract of time I think certain, from the condition of the 

 kitchen middens. Of these we explored two, and both showed 

 no traces of sequence of deposits. Moreover, it was evident 

 that the surface of turf and peaty soil had been removed round 

 the houses, for the layer of ash and pottery lay on that ; and 

 there was growth of peat and turf above it to the height of from 

 2-ft. 6-in. to 3-ft. 6.in. 



We found no evidence that the inmates of these huts were 

 engaged in tin mining. No moulds, no dross. On the contrary, 

 there was every appearance that they had been a pastoral people. 

 The hones found were all small, far too small to have served for 

 sharpening sickle or scythe. 



