It is not, however, on this slight evidence of local roadways that one would wish to base belief that the early Indians opened the first land highways in America. It is possible that they had great, graded roads near their towns and no roads elsewhere, but it is hardly conceivable. We have seen that the mound-building Indians occupied, in many instances, the heads of the lesser streams, and the argument in favor of their having opened the first land highways has been based on this interior situation which, unless the lesson of history in this case tends to false reasoning, necessitated landward routes of travel. There is, fortunately, one last piece of evidence which will more than make up for any lack of conclusiveness which may be laid to the charge of the preceding arguments. CHAPTER V WATERSHED MIGRATIONS A few descriptions of the local roadways of the mound-building Indians have been cited; reasons for believing that they used the watersheds, to a greater or less degree, as highways for passage from one part of the country to another, have been described. Let us look at the matter of their migrations. That these people did migrate there is no doubt among archæologists. The many kinds of archælogical remains now found indicate that they were divided into many different tribes, and the great distances between works of similar character show that the various tribes labored at divers times in divers places. "The longest stretch where those apparently the works of one people are found on one bank [of the Mississippi river] is from Dubuque, Iowa, to the mouth of Des Moines river. As we move up and down [the Mississippi] we find repeated changes from one type to another."[88] The direction from which these mound-builders entered the regions where their works are found, and their migrations within this region, must be decided by a careful study of the varying character of the mounds and a classification of them. This work has not been done, save in the most general way possible, though one highly important conclusion has been definitely reached. It is that the generally received opinion heretofore held by archæologists that the lines of migration were along the principal water-courses is not found to be correct, and that these lines of migration were _across_ the larger water-courses, such as the Mississippi, rather than _up and down_ them.[89]