IV.] ECCLESIASTICAL SOURCES. 89 treatment, without novel material, can supply. Unluckily we can only, in a very fragmentary way, trace back, in commenting on Herodotus or Livy, the actual line of tradition by which they received what they report to us; as to Thucydides and Tacitus the chain is less fragmentary, and, of course, in the narrative of contemporary writers we look less for traces of earlier authority than for traces of sympathy and antipathy, personal knowledge and direct authenticity. But, as a rule, we may say the study of ancient classical History lies within a confined area, every manuscript, every inscription, every coin, and every map of which has long been known, into which rash speculation never ventures without having cause for bitter repentance, and in which anything like a new discovery, such as the recent finds at Troy and Mycenae, is so very new, that when it does come to pass no one knows what to make of it. The very definiteness of all connected with this study makes it, for educational purposes, an incomparable discipline. As a study of knowledge for its own sake, as a field of discovery and profitable speculation, as a department in which the sum of human knowledge is likely ttf be largely increased, I confess I think that it promises perhaps less than the study of later History. As a ground for fresh and remunerative exploration, I am sure it is still less promising. To return, however, to the original sources. We all know how large a debt modern and medieval History owes to the ecclesiastical writers. From the very beginning of the middle ages, Annals, Chronicles, and Histories poured in comparative abundance from the religious houses of England and the Continent, records which serve to check and correct one another at almost every turn, and which, for some of the more dramatic incidents of History, enable us to reconstruct a picture of the event, viewed by different minds from different points and in different lights, in a