74 Journal of Philology. sufficient, and for some reason, perhaps because it was very little in use, not dating by years of Rome, he counted each separate date backwards from the consulate of Vinicius, to whom, in the reign of Tiberius, he dedicated his history. The effect is absurd. Still however, if we have an epoch in the middle of events, and no other epoch independent of it, we must to a certain degree have retrograde dating : the only thing is, we ought not to have it in such a way as to make as it were the actual course of events run backwards, and to go counter to every sub-reck- oning which comes in our way and has to be involved in our narration, so that the years for instance of a man's life become more as the date expressing them becomes less, and so for others. Such a way of reckoning, and our present reckoning B. C. is such, is only fit for history written backwards from effect to qause or occasion, as we may well conceive it written, like the house that Jack built ; as if we went in the order of shewing how the battle of Cheeronea destroyed the supremacy of the Thebans which they had won at the battle of Leuctra, which again destroyed that of the Spartans which they had won at jEgospotamoi, which again destroyed that of the Athenians which they had won at (Enophyta or in consequence of Salamis, and so forth : retrograde dating will do very well for retrogres- sive history. But if we want history to be a transcript of life and action, and dating to be a help to history and to our conception of it as such a transcript, then, as we talk of the world going on and one event following another, our chronology must move as the events do : or else it is no help, but a horrible clog, to our living conception of them. An inter-eventual epoch, like the Dionysian epoch of the Incarnation, perfectly arbitrary till fixed by usage, and which then ought to be considered, by all who care for historical con- venience, absolutely inviolable, is, it will be said, like the assumed meridian which you reckon longitudes from in geography, sup- posing one such meridian absolutely fixed by usage, and sup- posing you could not reckon round to it again : you must count A. D. and B. C. as you count E. and W. longitude. But there is this difference, that there is no reason why you should count longitude Eastwards rather than Westwards, and there is a very great reason why you should count time downwards from a be-
