1903 - 4 .] Date of Upheaval of Raised Beaches in Scotland. 269 
caldrons, including both the spheroidal and conical types. The 
former were always constructed on a uniform plan, the special 
feature of which was a lower rounded portion made of thin 
bronze, and an upper band of iron to which the lower was 
riveted, and to which also were fastened two large suspension 
rings. (See Gross, Oppidum Helvete , p. 45 and pi. xiii.) 
It will be remembered that one of the Cockburnspath caldrons 
was supposed to have had its mouth strengthened by an iron band. 
Similar caldrons made of iron have been found in Ireland, two 
being among the collection of relics from the Lisnacroghera cran- 
nog, which also contained a number of Late Celtic objects ( Lake 
Dwellings of Europe , p. 386). It would thus appear that there 
was an evolutionary sequence in the manufacture of these caldrons 
in the British Isles : first, those made of bronze ; second, those 
made of bronze and iron; and third, those made exclusively of 
iron. On the Continent, caldrons were generally found associated 
with sepulchral remains, except those from Oppiduvi La Tene, 
but in the British Isles they were undoubtedly used for culinary 
purposes. In protohistoric times in Ireland they were so highly 
prized that they are often referred to as heirlooms in families, and 
as forming part of the special property of kings. Tradition tells 
us that among the treasures brought to that country by the Tuatha 
De Danoan was the Goire an Daghdlia , or Magic Caldron. On 
these grounds I see no reason why the Kincardine caldron, though 
belonging to an earlier date, should not have been used as a Roman 
camp-kettle; and the association of the Cardross bucket with a 
military camp, traditionally believed to be Roman, lends additional 
support to this view. The general argument on this phase of the 
subject may be thus briefly stated : — The finding of bronze caldrons 
of pre-Roman types, and of a wooden roadway, presumably of 
Roman construction, in association with the debris of great forest 
trees, some of which showed over 300 ring-growths, all buried 
beneath a bed of peat from S to 14 feet thick, affords something 
more than presumptive evidence that the site of this forest had 
become dry land at least some centuries before the Christian era. 
But before attempting to assign a more precise date to this 
upheaval, it is desirable to know something of the terrestrial move- 
ment which caused it, especially as to the rate of its action. Was 
