formed must necessarily be very rough and imperfect. For, in the first place, we cannot always be sure of the route which he took from one town or village to another. Thus, for example, he describes two roads from Argos over Mount Artemisius to Mantinea; but there is nothing to show which he took or even that he took either. He may, like most travellers, have reached Mantinea from Argos by neither of the direct passes over the mountains, but by the circuitous route that goes by Lerna and Tegea. In the second place, it would be very rash to assume that he visited only those places where he is proved by some incidental assertion of personal knowledge to have been. Possibly or rather probably he visited many more. If he did not think it worth while to assure us that he saw the Parthenon and the Erechtheum at Athens, and the temples of Zeus and Hera at Olympia, he need not have thought it worth while to depose to having seen every insignificant shrine and image that he describes in the petty towns and obscure villages through which he passed. Thus the indications which he has given us are far too meagre to permit us to make out his itinerary in Greece with any approach to certainty. [Sidenote: Descriptions which he may have taken from books.] But if we cannot be sure that many of his descriptions are based on personal knowledge, have we any grounds for supposing that they are borrowed, without acknowledgment, from books? Such a supposition would be, on the face of it, neither unreasonable nor improbable. In the historical parts of his work Pausanias must have used many books which he does not mention, and he may have done the same thing in the topographical or descriptive parts. The grounds on which it could be proved or made probable that he borrowed his descriptions from books are various. The most obvious and certain would be the existence in an older writer of a description agreeing in form as well as in substance so closely with a description in Pausanias that no alternative would be left us but to suppose, either that Pausanias copied from this older writer, or that both of them copied from some common original. Or again it might be that the descriptions of Pausanias contained information which he could hardly have ascertained for himself, or mistakes into which he could scarcely have fallen if he had seen the things for himself. In regard to the first of these grounds it may be said at once that in the extant literature of antiquity, so far as the present writer is aware, there is no description of any place or monument agreeing in form and substance so closely with a description in Pausanias as to make it probable that he copied it. The slight and superficial resemblances which have been traced between passages of Strabo and passages of Pausanias are no more than such as may easily or necessarily arise when two writers are describing independently the same places.