name was obtained from the Iberians, whom they found already in possession of Western Europe. But there is another alternative which is probably to be preferred. As we found the Albanians calling gold by a name derived from the gold coins of Florence, so the Kelts may have adopted the Latin names for gold used by their Roman conquerors. This is made almost certain by the fact that aura, in old Norse, derived from Latin aurum, became the regular word for treasure, although no one will deny that the Teutonic peoples had already gold and its cognates as terms of their own for the metal. Everyone is familiar with the influence exercised by the Roman coinage even in the countries of the East, where Rome met with a civilization hoary in age before Romulus founded Rome, and from which Rome herself had ultimately derived the art of coining. Yet by the time of Christ the Roman denarius, the penny of our Authorized Version, had already asserted itself in the Greek-speaking provinces of the East, and became in later days, when the rule of Rome and Constantinople fell before the Arab conquerors, under the form of dinar, the standard coin of the great Mahomedan Empires. Did then in like fashion the Roman form of the name for gold, which in all probability varied but little from the cognate Gaulish word, supplant at a comparatively early period that native form?
The same argument may be urged in reference to the silver. The Irish form is airgid, according to some a loan-word, being simply the Latin argentum. We have already seen that it is not possible that the Kelts, in constant contact with the Iberians who were so rich in silver, could have remained in ignorance of that metal. The Gaulish form of the name for silver was plainly in Roman times almost the same as the Latin, as is shown by Argentoratum, the ancient name of Strasburg. It is plain then that before the Roman Conquest the Gauls had a town called by the name for silver, whilst the Irish form has no nasal, the Gaulish coincides completely with the Latin. Is it not possible, that in this case too a native Keltic name, a close cognate of Latin argentum, whose lineal descendant is seen in the Irish form, may have been assimilated to the Latin form? But there is plenty of evidence from other quarters to show that
