through Ohio along the Ohio river and onward in the same direction to the northeastern corner of Texas; another follows the Mississippi river; another extends from the region of the Wabash to the headwaters of the Savannah river, and another across southern Michigan and southern Wisconsin. The inference, however, which might be drawn from this fact—that these lines indicate routes of migration—is not to be taken for granted. It is shown by the explorations of the Bureau, and a careful study of the different types of mounds and other works, that the generally received opinion that the lines of migration of the authors of these works were always along the principal water-courses, cannot be accepted as entirely correct. Although the banks of the Mississippi are lined with prehistoric monuments from Lake Pepin to the mouth of Red river, showing that this was a favorite section for the ancient inhabitants, the study of these remains does not give support to the theory that this great water highway was a line of migration during the mound-building period, except for short distances. It was, no doubt, a
