The observations made at Lochlee led me to ascribe an exceptionally large share to this element as a cause of the submergence of the crannog, but since then further investigations have proved that the phenomenon takes place to a similar, if not greater, extent in localities where no rising of the surface of the waters could have occurred. SECTION V. _Chronological, Social, and other Indications derived from the Relics._ Having thus glanced over the collateral phenomena bearing on the age or period of the Scottish Lake-Dwellings, it only remains to say a few words on the general evidence supplied by the relics themselves. We have already seen that the scattered remains of the artificial islands brought to light in modern times extended over a large area, embracing nearly the whole of Scotland, with the exception of its two northernmost counties and a few others lying on its south-eastern extremity. So far, however, as the discovery of actual remains illustrative of the civilisation and social condition of their occupiers is concerned, we are almost entirely limited to the results of the investigations made at Dowalton, Lochlee, Lochspouts, and Buston, all of which are within the counties of Ayr and Wigtown. In instituting a comparison between the contents of these four groups, their analogy, not only as regards the structure and local distribution of the islands, but as regards the general character of the relics, is so wonderfully alike that we have no difficulty in assuming that originally these lake-dwellings within this area were erected by one and the same people for a special purpose, and about the same time, or at least within a comparatively limited period. It is true that a considerable diversity exists as regards the number and character of the relics found in some of these localities; thus neither Samian ware nor implements of bone or horn have been found at Dowalton, though these are amongst the relics from all the Ayrshire examples. Such negative evidence, however, does not amount to much, more especially in this case, as the absence of Samian ware is more than compensated for by the presence of other articles presumably of Roman origin; and, moreover, I believe that a more careful search would have greatly increased the number of relics from Dowalton. This uniform similarity in these remains, though not entitling us to extend the above generalisation beyond a certain geographical area in the south-west of Scotland, is however sufficiently marked to enable us to dispense with any further necessity of discussing the merits of each group separately; so that whatever inferences can be legitimately derived from a critical examination of any one group may be safely applied to the whole.