be excused for suggesting that in works on comparative religion and morals this principle is not always sufficiently realised, and that the panorama of religious or quasi-religious practice from all parts of the world, and found among peoples of very different stages of development, with which we are now so familiar, needs constant testing by increased knowledge of those peoples in all their relations of life. At any rate, in dealing with Roman evidence the investigator of religious history should also be a student of Roman history generally, for the facts of Roman life, public and private, are all closely concatenated together, and spring with an organic growth from the same root. The branches tend to separate, but the tree is of regular growth, compact in all its parts, and you cannot safely concentrate your attention on one of these parts to the comparative neglect of the rest. Conversely, too, the great story of the rise and decay of the Roman dominion cannot be properly understood without following out the religious history of this people--their religious experience, as I prefer to call it. To take an example of this, let me remind you of two leading facts in Roman history: first, the strength and tenacity of the family as a group under the absolute government of the paterfamilias; secondly, the strength and tenacity of the idea of the State as represented by the _imperium_ of its magistrates. How different in these respects are the Romans from the Celts, the Scandinavians, even from the Greeks! But these two facts are in great measure the result of the religious ideas of the people, and, on the other hand, they themselves react with astonishing force on the fortunes of that religion.