All rock units, many fabrics, and two major faults mapped in the Hadlyme and Willimantic areas, Connecticut, are identified in a 1.45-km-deep research hole drilled at Moodus, Connecticut. Correlation of surface information and existing shallow drill-core data with the Moodus data was partially based on the description of core and cuttings from the deep hole, with supplemental chemical analyses, petrography, and a geochemical log. Rocks to a depth of 709.9 m are a mixture of quartz-diopside granofels and biotite-quartz-plagioclase schist identical to outcrops of Hebron Formation and are assigned to the Merrimack terrane. A 46.1-m section of biotite granitic gneiss within the Hebron Formation at Moodus correlates with the Canterbury gneiss. Below the Hebron Formation is 26.2 m of mylonitic muscovite-biotite-plagioclase-quartz schist, and 79.2 m of mylonitic, strongly layered garnet-sillimanite-muscovite-biotite-plagioclase-quartz schist and biotite-hornblende-plagioclase-quartz gneiss. These units correlate with the Yantic and Lower members, respectively, of the Tatnic Hill Formation, included in the Putnam-Nashoba terrane in eastern Connecticut. The rocks below 815 m in the Moodus hole are dominated by biotite- and hornblende-bearing plagioclase gneisses, but also contain thin layers of granitic gneiss, amphibolite, and pegmatite. The total lithologic assemblage is very similar to and correlated with the sequence of rocks of the Avalon terrane south of the Honey Hill fault. Subdivision of these rocks into the Hadlyme and Mamacoke Formations of the Waterford Complex is proposed. Chemical variability identified in the geochemical log is used to divide these gneisses into 44 layers 12 to 30 m thick, and is interpreted to reflect primary compositional fluctuations of volcanic and intrusive protoliths. Fabrics in the core and chips from the research hole indicate that mylonitic fault zones occur above (125 m thick) and below (145 m thick) the Canterbury gneiss. Phyllonitic fabrics overprint these mylonitic rocks and lesser deformed rock higher in the Hebron Formation, at the base of the Canterbury gneiss, and throughout the Avalon section. Brittle faults indicated by the presence of breccia, microbreccia, and slickensided chlorite surfaces with associated chlorite-zeolite-calcite-quartz–filled fractures occur throughout the well and are relatively common in the Avalon section. The range of temperatures and deformation mechanisms reflected by these rocks is similar to those identified in the Honey Hill fault zone. This suggests that the long history of reactivation of the Honey Hill fault zone extends as far northwest as Moodus and is represented there by a wider zone of deformed rock.

Major-element chemical analyses of 130 samples of high-grade plagioclase gneiss, together with geologic data, suggest that metadacitic gneisses of the late Proterozoic Avalon and Putnam-Nashoba terranes include four extrusive and two intrusive units formerly mapped as Monson Gneiss and Middletown Formation. Evidence for an igneous rather than sedimentary protolith for these gneisses comes from the mineralogic and chemical homogeneity, and the consistency of the chemical variations within individual bodies of gneiss with fractional crystallization processes. Evidence for the extrusive origin of some units comes from interlayered contacts with metasediments and metarhyolitic (alaskitic) and metatholeiitic (amphibolitic) volcanics, and from a 20- to 50-m-scale chemical cyclicity, typical of modern ash-flow tuffs. Evidence for an intrusive origin of two bodies comes from abundant amphibolite and calc-silicate xenoliths, from a massive structure lacking compositional layering, and from kilometer-scale chemical zoning or homogeneity. The identification of some units as lithodemic orthogneisses indicates that intrusion as well as extrusion led to the lithologic sequences in the Avalon terrane of southeastern Connecticut. It also precludes the lithostratigraphic correlation of these orthogneisses with paragneisses. Given this geochemical and geologic evidence, we believe that previously advanced stratigraphic arguments that major recumbent folds relate gneiss bodies now interpreted to be intrusive and extrusive should be reexamined. We believe that the part of the Avalon terrane consisting of metavolcanic rocks may be a right-side-up volcanic pile, and that major recumbent folding may not be present.

Revised correlations of critical units in south-eastern Connecticut contradict earlier stratigraphic correlations that led to the interpretation of major recumbent fold napping in the Avalon terrane. The revisions are required in part by the discrimination between lithodemic and lithostratigraphic units. We find no compelling evidence that schistose Putnam-Nashoba rocks exist within the Avalon terrane. On the contrary, we interpret these blastomylonitic schists and gneisses within the Avalon terrane around the Lyme dome as ductile fault rocks, derived from an Avalonian granodioritic orthogneiss. Rocks originally mapped as Brimfield Formation (Merrimack terrane) west of Old Saybrook are shown to correlate with the Middletown Formation of the Bronson Hill terrane. This new correlation makes all plagioclase gneisses of the Killingworth dome isolated from rocks of other terranes by the Middletown Formation, and provides evidence against the assignment of any plagioclase gneisses south of the Bronson Hill terrane near the Killingworth dome to the Avalon terrane. In contrast to the interpretation of fold napping, we propose that a major fault nappe exists in the Avalon terrane. The fault zone separates rocks of the Lyme dome from overlying allochthonous rocks of the Waterford Complex in the “Selden Neck block,” and fault splays isolate paragneisses of unknown age. Splays of the fault extend west, where they cut an extension of the Honey Hill fault zone, narrow slivers of Merrimack terrane rocks, and units in the Bronson Hill terrane. Important deformation in the zone probably occurred from middle Carboniferous to middle Permian, although earlier deformation is possible. Subsequently, the entire zone was folded around the Lyme dome. The tuning of the deformation suggests that it was related to the Hercynian assembly of Pangaea.

Location Stop 1. The most spectacular exposures of any ductile faultin southeastern New England are those of the Willimantic fault inroad cuts along the westernmost interchange of the 5 mi (8 km) segment of unfinished I-84 at its junction with U.S. 6 and Connecticut 66, 3.5 mi (5.7 km) west of Willimantic, Connecticut (Fig. 1). The outcrops at 41°43'24„ north latitude and 72°16'west longitude are in the town of Coventry, Tolland County, east-central Connecticut (Fig. 2). They lie in the northwest quadrantof the Columbia 7½-minute Quadrangle, but are not shownon Snyder's (1967) bedrock map because the cuts were firstopened in 1972. This interchange maybe reached from Hartford, Connecticut by following I-84 east to U.S. 6, then east to theinterchange from Middletown, Connecticut by following Connecticut 66 east to the interchange; from Providence, Rhode Island, by following U.S. 6 west to the interchange; or from Norwich, Connecticut by following Connecticut 32 north to U.S.6, then west to the interchange. Once at the interchange, park in the commuter parking lot between U.S. 6 and Connecticut 66, and walk up the median of I-84 to the cuts (see Fig. 1). These cutsare in part of a federal highway system! Stay well off the pavementand away from traffic! Please do nothing to jeopardize theuse of these exposures by future geologists.