A GALLOWAY STONE-AGE VILLAGE. 93 
dwellings so as to have them more or less surrounded by water 
or marsh. They did not construct their houses of stone half- 
underground or wholly subterranean like the earth-houses or 
weems, nor did they follow the methods of the hut-circle men 
who lived in circular wood and wattle huts built on the surface 
of the ground. 
While the bottom of each pit was at some depth under the 
surface, it should not be forgotten that the supposed flooring 
was sufficiently near the pre-historic surface to make it necessary 
to have a good proportion of the cubical contents of the chamber 
above the level of the surface if the chamber were to be 
habitable. There is no reason to believe that the huts were 
open to the heavens, or that the excavated soil was removed 
from the spot. It then follows that the upper part of the dwell- 
ings was more or less mound-like in character, the heaped-up 
earth from the original excavation having assisted towards this 
appearance. It seems difficult to escape from such a conclusion. 
The shape of the mound would naturally follow the plan, 
whether round or oval, of the structure covered by the mound. 
Again the structure would be in harmony necessarily with the 
plan of the foundations, the flooring, and the walls, all of which 
were apparently in the shape of a longish oval. It would appear 
then that the mounds were somewhat long in shape, though 
perhaps not so pronouncedly so as the plan of the wooden 
substructures. 
In considering this peculiarity of the Stoneykirk remains, it 
is interesting to consider that most of the ancient British dwell- 
ings, built of wood and wattle, which have been examined and 
recorded, are round, and, further, that they belong to the 
Bronze, the Early Iron, and later periods. The huts which 
comprised the marsh village at Glastonbury, occupied a few 
years before the Romans arrived in that district, were roughly 
circular and about 20 feet in diameter. With the exception of 
the Isle of Wight specimens, the pit-dwellings found in England 
appear to have been circular, as were also the crannogs and 
hut-circles. 
No stones were used in the Stoneykirk structures, and stones 
sufficiently large for building walls are rare in the locality. It © 
seems safe to say that the walls were of turf and wood and 
wattle. If the places were roofed, it is safe to assume further 
