This report gives the results of a geologic study of certain features that bear upon the recent flood behavior of rivers flowing in the Massachusetts part of the Connecticut Valley. It is in part an outline of the physiographic history of the Connecticut River, a 'history that is treated in progressively greater detail as it concerns events occurring from Mesozoic time to the present, and in part a discussion of erosional and depositional processes associated with the extraordinary floods of March 1936 and September 1938.
The Connecticut River flows southward through Massachusetts in a broad lowland area of more than 400 square miles and is joined in this area by four large tributaries, the Deerfield and Westfield Rivers from the west and the Millers and Chicopee Rivers from the east. The lowland area, or :Connecticut Valley province, is flanked on the west by the Berkshire Hills, a, deeply incised uplifted plateau, and on the east by the central upland, or Worcester .County plateau, a lower upland marked by rolling topography. Most of the broad, relatively flat valley floor is underlain by Triassic sedimentary rocks. Rising above it, however, are the prominent Holyoke-Mount Tom and Deerfield Ranges, which consist in large part of dark-colored igneous rocks, also of Triassic age.
There is evidence of several cycles of erosion in central western Massachusetts, the last two of which are of Tertiary age and appear to have reached nature and very youthful stages of topographic development, respectively. Immediately prior to the glacial epoch, therefore, the Connecticut River flowed in a fairly narrow, deep gorge, which it had incised in the rather flat 5ottom of the valley that it had formed at an earlier stage. A Pleistocene crustal subsidence probably of several hundred feet, for which there has been only partial compensation in postglacial time, was responsible for the present position of much of this gorge below sea level. That an estuary does not now occupy the gorge is due to a filling by glacial debris, notably by sediments deposited in late glacial lakes. Following disappearance of the last ice sheet and draining of the associated, lakes, the Connecticut River resumed existence and began a new chapter in its history.
In those areas where the river regained its preglacial course, it now flows on sediments considerably above the rock floor of the old gorge. Where the gorge was narrow and deep, the upper parts of its walls have confined the postglacial river within rather narrow limits, as in the northern part of the state. Where it was sufficiently wide to be filled by glacial sediments over large areas, the postglacial river has meandered broadly, as in the area north of the Holyoke-Mount Tom Range. In two areas in Massachusetts and in one immediately south in Connecticut, however, the river was forced from its preglacial gorge, and its new channel has been superimposed on bedrock, with development of rapids and falls. Each of these postglacial rock channels acts as a spillway whose level controls the local base level of the river as far upstream as the next spillway. These spillways are not to be confused with other, more spectacular gorges, which are of preglacial origin and in which the present river does not flow on bedrock.
The Recent Connecticut has formed extensive flood plains and terraces through repeated sequences of erosion by lateral corrosion and downward scour, followed by deposition of .silt and sand veneers. These features, although irregular in detail, appear to be assignable to five general levels, whose means are approximately 49, 37, 30, 18, and 10 feet above present mean river level. In addition, an 80-foot terrace in the northern part of the valley was left perched, in its present position when the Connecticut abandoned its course over. a rock barrier near Turners Falls in favor of an adjacent much lower gap. The normal terraces and flood plains, slope very gently away from their riverw