We may ask why did the Kelts especially choose the Roman form of the name for gold, if they were then for the first time getting a name for the substance then (according to Schrader) first known to them? Before they ever reached Latium they had been in contact with peoples in Northern Italy who undoubtedly were well acquainted with gold. The Etruscans were a wealthy people, who coined gold pieces before Rome had struck coins of any kind[94]. The Umbrians on the east side, the ancient Italic race who had in the days before the Etruscan Conquest held all Northern Italy up to the Alps, which was hence known to the earliest Greek geographers by the name of Ombriké[95], were, beyond all doubt, acquainted with the use of gold, and had a name for it probably the same as the Sabine _ausum_. Why then did the Gauls remain entirely ignorant of gold and of a name for it when they had been in constant contact with those peoples who had most undoubtedly abundance of the metal and names of their own for it? Until some sufficient answer is given to the objections here raised, we must on every logical and scientific ground refuse our assent to an argument, the sole basis of which is philological. It may not be inappropriate also here to remark that it is most desirable in all historical enquiries to rely as little as possible on Etymology. From the days when the Stoics laid such importance on arguments based on the _originatio verborum_ down to the present time reasonings based on such foundations have been as a rule founded on the sand. Comparative Grammar as yet can hardly be described as a science. New principles and laws are brought to light each year, and, although of course the solid _residuum_ of what may now be regarded as more or less positive knowledge is slowly growing in bulk, those laws which were the shibboleth of Philologists a decade ago, are now rudely hurled from their preeminence. The only sound scientific method in historical research is to employ linguistic science as merely ancillary to our enquiries.