2. The boulders of the belts are almost without exception derivatives from the crystalline terranes of Canada. Those of the great tract especially under consideration were derived from the typical Huronian rocks of the region north of Lake Huron, and from granitic and gneissoid rocks referable to the Laurentian series of the same region. These last, however, cannot be sharply distinguished from the granitic rocks derived from other parts of the Laurentian terrane. The Huronian rocks are very easily identified because of the peculiarities of some of the species. Among these the one most conspicuously characterized is a quartz-and-jasper conglomerate. The matrix is usually a whitish quartzite. This is studded with pebbles of typical red jasper and of duller rocks of jasperoid nature, which grade thence into typical quartzite pebbles. With these are mingled crystalline pebbles of other varieties. Another peculiar erratic comes from the "slate conglomerate" of Logan. It consists of a slaty matrix through which are scattered rather distantly pebbles of granitic, quartzitic and other crystalline rocks. This is one of the forms of the "basal conglomerate" of Irving. Other varieties of this "basal conglomerate" are present. In addition to these very peculiar rocks, a quartzite of a very light greenish semi-translucent hue has a wide distribution along the tract. It is readily distinguishable from the numerous other quartzites of the drift of the interior. Some years since, on returning from my first field examination of a portion of this belt, I sent a typical series of chips from the characteristic erratics to Professor Irving, who had recently returned from the study of the original Huronian region. He returned a suite of chippings that matched them perfectly throughout, all of which were taken _in situ_ in the region north of Lake Huron.