1872.) The Dolmen-Mounds of Brittany. 19 
and Jutland,—and, following the coast line, traversed Bel- 
gium, the North of France, Normandy, finally reaching 
Brittany, where the numerous dolmens attest their prolonged 
stay. Part are then supposed to have crossed over by the 
Channel Islands, which are rich in dolmen-mounds, to 
Cornwall and Devon, gradually reaching the S.E. of Ireland 
and Wales (the absence of such remains in the West of 
Ireland and in East England is very marked). Another 
portion left Brittany, and penetrated southwards along the 
coast as far as the Gironde, whence, leaving the sea-board 
to avoid the sandy plains of Gascony, ‘they followed the 
course of the Dordogne, and traversed France in the direction 
of the Gulf of Lyons. Small detached bands seem also to 
have penetrated into Savoy and Switzerland, as shown by a 
few isolated dolmens in those localities. The mountains 
seem to have delayed the onward progress of these nomades 
for some time, in the Departments of Arriége, Upper and 
Lower Pyrenees, but, at length crossing this obstacle, they 
leave traces in Portugal, through Spain,—vza Cordova, 
Granada, and Malaga,—and finally, crossing the Mediter- 
ranean, have left their tombs in the northern coasts of 
Africa, up to the very frontiers of Egypt. 
The African dolmens are probably the most recent, and, 
indeed, there are some who attribute them to the Roman 
period, and assert that they are due to the presence of a 
Roman legion raised in Armorica, who brought with them 
from Brittany their national customs and Celtic mode of 
burial; and there is a doubtful account of a dolmen in 
Algeria, inscribed to the memory of an Armorican centurion. 
Another hypothesis, supported by Messrs. Désor and 
Rougemont, is to the effect that the stream of dolmen- 
building tribes sprang from Africa originally, and were 
Euskarian rather than Celtic, and poured in a northerly 
direction over the Iberic peninsula, and thence into France ; 
in fact, the reverse of M. de Bonstetten’s track. The evi- 
dence in favour of this last theory is only negative, relying 
on the absence of tradition in Africa pointing to an invasion 
from the north, and the Celtic tradition in Ireland that the 
Irish are of African origin. 
For our own part, we believe the European dolmen- 
builders to be of a Scandinavian origin, and for this reason, 
that the Scandinavians have followed the ancient custom of 
erecting chambered tumuli to a very late period; and in 
their literature we find the records of their funeral cere- 
monies, which accord with what we should expect to find 
practised by the dolmen-builders. We are indebted to 
