of Britain in the fourth century of the Christian era. We have seen that Latin writers of this period make mention of the Atecotti, usually in conjunction with the Scotti. Some have assumed that the Atecotti were a branch of the Picts. So far as positive evidence goes, it is against this assumption. Ammianus speaks of the Picts, subdivided into two nations, Dicalydones and Verturiones, and then adds that "the Atecotti, a warlike nation," and the Scotti, were engaged with these in the work of devastation. This implies that the Atecotti, like the Scotti, were distinct from the Picts. A verbal resemblance in the names led some Irish writers, from the close of the eighteenth century down to O'Curry, to identify the Atecotti with the Irish Aithech-thuatha, the ancient Rent-paying communities referred to in my third lecture. I do not think that the philologists will sanction the identification so far as it is based on verbal resemblance. The name Atecotti has not been found in any form in the native records of Ireland or Britain as the name of any nation or sub-nation or in the topography of either island. Nevertheless contemporary evidence during the second half of the fourth century shows that not only on the frontier of Roman Britain but also on the Continent there was a numerous and warlike collection of men known by this name. As in the case of the name Scotti, the conclusion I would draw is that Atecotti was a name for a general class of men not for a particular nation, tribe, or political community. The name, in its best authenticated form, is a Celtic word, consisting of the adjective _cottos_ preceded by the prefix _ate_. _Cottos_ means "old," or "ancient." The prefix _ate_, which becomes _aith_ or _ath_ in Irish of the MS. period, means "back" or "again," like the Latin _re_, and like this, too, it often has a strengthening or intensifying force. Thus, Atecotti may be taken to mean the very ancient, the primitive, the pristine folk; and so it is explained by Whitley Stokes. Who then were these very ancient people who were associated with the Scotti and were not identified with the Picts? We are reminded at once of the Irish traditions of non-Gaelic and pre-Gaelic communities which formed the main fighting strength of the kings of North Leinster and South Leinster, and of the non-Gaelic origin ascribed to Cú Chulainn, Fear Diadh, and to the kindred of Fionn mac Cumhaill and of Goll mac Morna. Of course, on this point we are far from complete certainty, but the probability, in my opinion, is that, when the Irish went to war in the fourth century, they still adhered to the politico-social distinction between the Gaelic ascendancy and the conquered plebeian race, and that this was the distinction between the Scotti and the Atecotti. The adjective _cottos_ does not appear to belong to the vocabulary of Irish, but it is found in the various Brittanic dialects and was a frequent element in Gaulish nomenclature. The Atecotti, therefore,