Emerging from the dunes at a point north of the Iroquois river in Jasper county, northwestern Indiana, a well characterized belt of surface boulders stretches westward to the State line, just beyond which it curves about to the south and then to the east, and re-enters Indiana a little south of the northwest corner of Benton county. It soon turns abruptly to the south and reaches the Wabash river near the centre of Warren county. The immediate valley of the Wabash is thickly strewn with boulders from the point where the belt reaches it to the vicinity of West Point on the western line of Tippecanoe county. The uplands, however, do not give any clear indication of the continuity of the belt, and the connection is not altogether certain. There is an inner well-marked belt that branches away from this in the central part of Benton county and runs southeasterly into the northwestern quarter of Tippecanoe county, beyond which only scattered boulders occur, which leaves its precise connections also in doubt. But starting from West Point, which is less than a dozen miles from the point where the two belts cease to be traceable with certainty, a well-defined belt, one or two miles wide, runs southeasterly across the southwestern corner of Tippecanoe county and the northeastern quarter of Montgomery county to the vicinity of Darlington, beyond which its connection is again obscure, although boulders occur frequently between this point and the northwestern corner of Brown county, where boulders are very abundant. So also, patches of exceptionally abundant boulders occur in the west central part of Clinton county. These may be entitled to be regarded as a connecting link between the train which enters northwestern Tippecanoe county and that of northwestern Boone county, as scattered boulders of the surface type, but of not very exceptionally frequent occurrence, lie between them. However this may be, a belt of much more than usually frequent surface boulders stretches southeasterly to the vicinity of Indianapolis, and probably connects with a very well-marked belt lying near the south line of the southeast quarter of Marion county and in the northeastern part of Johnson county. There is also a well-defined tract in southeastern Hendricks county, running east and west, without evident connection with the foregoing tracts, though it may be the equivalent of the Darlington belt. There is also a somewhat unusual aggregation in the form of irregular belts in southeastern Johnson county, in the vicinity of Nineveh, and in southern Shelby county. The belt south of Indianapolis is probably to be correlated by scattered boulders only slightly more abundant than those of the adjacent region, but of the surface type, stretching northeasterly to near the center of the west half of Henry county, where a well-marked belt again sets in. From this point the tract runs northeasterly nearly to the north limit of the county, where it turns