~ 55 ~
time not earlier than the firm establishment of the Jewish Dispersion as a substantial factor in history, a point which in Egypt at any rate is as early as the fourth century B.C.
And here I may appropriately add what little I have to say about the further antiquity of the institution, a question, be it observed, distinct from that of the antiquity of the week in general, that is of a seven or eight-day cycle not governed by planetary associations. We find no trace of the planetary week earlier than Tibullus, but absence of evidence is not evidence of absence, and I will not say that it is absolutely impossible that the sequence may have come down continuously from a remote age. But I believe that the supposition is improbable as well as unnecessary. In the first place there is nothing improbable in the view that the idea of a planetary chronocratory took shape not long before the time when we can first trace it. When we find a usage first coming to our knowledge in a certain period of history, we are justified in believing that it is a heritage from some remote past on two conditions. One is that we find a tradition that it comes from such earlier time. There is no trace of any such tradition in the case of the planetary week. On the contrary, the one statement which we have concerning its age, that of Dion Cassius, declares that it is recent. The other condition is that the usage shews signs of being a survival of ideas which have become
