stone, may become mixed in the soil with those of a long subsequent date; and second, that had these stone implements been in common use in Roman times, their presence among Roman re- mains would have been the rule and not the exception, and we should have found them mentioned by Latin authors. Moreover, if their use had survived in this manner into Roman times, we should expect to find them still more abundantly associated with tools of the Bronze Age. We have, however, seen how rarely this class of stone instruments is found with bronze.
As to the stone celt discovered at Ash, Mr. Douglas remarks it may not "be improbable that this stone instrument was deposited with the dead, as an amulet; and which the owner had found and preserved with a superstitious reverence." In a tumulus in Flanders,[1] six celts were found placed upright in a circle round the interment, but from the difference in the condition of their surface they appeared to be of different ages, so that it has been suggested that they also were gathered from the surface of the soil and placed in the tomb as amulets. We shall subsequently see that flint arrow-heads were frequently thus preserved in Merovingian cemeteries.
In many cases in Germany,[2] stone axes, for the most part perforated, are said to have been found in association with objects of iron; but the proofs of the contemporaneity of the two classes of objects are not satisfactory. The religious veneration attaching to the Thor's hammers may, however, have had to do with their interment in graves, at a time when they had ceased to be in ordinary use. Moreover, the axes may have been preserved to ward off lightning.
Another argument in favour of these instruments having remained in use in Britain until a comparatively late period, has- been derived from the circumstance of the words stan-æx and stan-bill, occurring in Æfric's Saxon glossary. These words are translated by Lye[3] as a stone axe, a stone bill—terms which have naturally been regarded as referring to axes and bills made of stone, which, therefore, it might be reasonably inferred were in use at the time when the glossary was written, or about A.D. 1000. On examination, however, it appears that no such inference is warranted. The glossary is Latin with the Saxon equivalents annexed to each word, and the two words referred to are
