It is only when the supporter of the authenticity of a work of singular and unique antiquity can begin with an epoch of comparatively recent date, and argue backwards through a series of continuous works, each older than the other, to one still older than any, that he can reasonably accuse the critic who demurs to his deductions of captiousness. In this way the antiquity of the oldest Chinese annals is invalidated: in this way the date of the Indian Vedas (1400 B.C.). But the great classical literatures stand the test, and from the present time to Claudian, from Claudian to Ennius, and from Ennius to Archilochus we trace a classical literature with all its works in continuity; each pointing to some one older than itself. Even this forbids an excessive antiquity to Homer. Again--the likelihood of forgery must be continually kept in mind; so much so, that even in the unexceptionable literature of the classics, if it could be shewn that any age between the present and the eighth century B.C., were an age in which the Greek drama, the Greek epics, the Greek histories, or the Greek orations could be forged, a great deal would be subtracted from the proofs of their antiquity. I do not say that it would set them aside; because everything of this kind is a question of degree; but the argument in their favour would be less exceptionable than it is. For it cannot be too strongly urged that the preservation of records of high antiquity, in and of itself, is naturally and essentially improbable. More than half of the antiquities of the world have been lost; and this alone gives us the odds against an instance of survivorship. This has been insisted on by more than one archæologist--more cautious and candid than the majority of his brotherhood. Whoever doubts this should look around him. How few nations have a literature! How thoroughly is the non-development of a permanent literature the exception rather than the rule! And, even when records come into existence, how numerous are the chances against their preservation. Destruction is the common law: continuance a happy rarity. For extraordinary phenomena we must have extraordinary proofs. From the present time to the eleventh century we may trace the native Welsh literature continuously; but no farther. If any thing be older than the laws of Hoel Dhu, they must be so by four centuries, with nothing in the interval. This is the measure of the value of Welsh evidence to the events of the fifth century. Writers, however, in Latin existed earlier. Still, this is unsufficient to be conclusive to the validity of a fact in the fourth. Such a statement must be tested by its own intrinsic probability. It cannot come before us invested with the dignity of a historically authenticated event. What this is will soon appear.