This paper is a compilation of published information considered pertinent to the correlation of the 450 Silurian stratigraphic units now recognized in the British Isles. The number of units reflects both the geologic complexity of these islands and the fact that much careful description and geologic mapping was done during the past two centuries. The accompanying bibliography contains more than 800 titles. Graptolitic facies are widespread in the British Isles, and the graptolitic zonal sequence is used here to form the basis of correlation. In the shelf facies where graptolites are less common, brachiopod lineages are becoming increasingly useful in correlation. Recent work on other groups, particularly conodonts, spores, and ostracodes, provides further bases for correlation. In early Paleozoic time, Britain and Ireland were traversed by an ocean basin, called the Iapetus Ocean, which in Silurian time, however, provided no barrier to migration as most marine animals are common to both sides. The ocean-floor deposits are characterized by thin, pelagic shale sequences and thick graywacke turbidite sequences, all within the graptolitic facies. Shelf deposits occur along both margins of the ocean, in the Welsh Borderland and Wales on the southeast, and in Scotland and Ireland on the northwest. These shelf areas are characterized by clastic shelly facies with limestone locally developed. Nonmarine deposits are mainly associated with Late Silurian time, just prior to the late phases of the Caledonian orogeny, when all of the British Isles except southwest England became a mountainous, nonmarine area.

The authors are indebted to C. W. Harper, Jr:, for the use of material from his doctor's thesis on the brachiopods of the Arisaig Group. In that study, he identified many of the small collections that have helped to make correlation possible between the shore and the inland areas. W.B.N. Berry and M. J. Copeland have greatly assisted by identifying the graptolites and the ostracods, respectively, found at Arisaig. R. K. Bambach has generously contributed his knowledge of the Arisaig bivalves; he has also permitted examination of Twenhofel's field notebook. The chapter on the Knoydart has greatly benefited from the work of R. H. Denison, D. J. MacNeil, A. Suparman, and M. Tauchid. John F. Dewey gratefully acknowledges financial support from the University of Manchester, the Royal Society of London, and the Nova Scotia Research Foundation. D. Kelley has been generous with information on the Cobequid Mountains. We are indebted for funds received by W. S. McKerrow from the Nova Scotia Research Foundation, the Royal Society of London, and the Geological Society of London Tyrrell Fund, and by D. L. Dineley from the Geological Survey of Canada and the National Museum of Canada. McKerrow spent the year 1964–1965 as a post-doctoral Research Fellow at the California Institute of Technology with support from the National Science Foundation (Grants GP-2290 and GP-3743). Acknowledgment is made in the list of maps to those Massachusetts Institute of Technology students who contributed to the brook maps presented in this report, but thanks are due

The richly fossiliferous Silurian and Early Devonian rocks of the Arisaig area and the nearby related regions of Cape George and Lochaber in Antigonish County, Nova Scotia, have been known for almost one hundred and fifty years. The 5,000 ft of sediment in the Arisaig Group is the only fossiliferous section within the Province containing representatives of every stage of the Silurian. This monotonous succession of siltstones and mudstones is unique world-wide in providing an almost continuous faunal record of very shallow water conditions rather than an alternation of varying depth environments. The Eocoelia and related benthic animal communities persist from the early Llandovery through the entire Silurian and well into the early Gedinnian portion of the Lower Devonian. These rocks despite their relatively unmetamorphosed nature (never higher than the chlorite grade) have been highly folded and extensively faulted, beginning with disturbances associated with the Middle Devonian Acadian orogeny and extending into various intervals of the later Paleozoic and possibly Triassic. The complex structure is here outlined in more detail than is available in previous reports, thus giving both geologists and paleontologists a better framework with which to understand both faunal and geologic problems associated with the rocks. During the Gedinnian, the entire region became a site of nonmarine, Old Red Sandstone—type deposition, but there is no evidence for any stratigraphic break between the underlying marine and the overlying nonmarine beds. The Devonian vertebrates are similar to those of the Welsh Borderland of England. The rhyolitic and basaltic volcanics . . .

Purpose and Scope of the Investigation The Arisaig area (Fig. 1; Pl. 1) has been selected for detailed study because it provides the most continuous and best exposed sections of marine Silurian and early Lower Devonian rocks in the Appalachian Mountain system. The earliest fossiliferous rocks present are lower Llandovery in age, and it is probable that the whole of the Silurian is represented in the 4,000 ft of beds that make up the Arisaig Group; but the underlying Bears Brook Volcanic Group has no known fossils, and the Silurian rocks are faulted against the Ordovician Browns Mountain Group. It is not clear, therefore, how large a stratigraphic break, if any, is present at the base of the Arisaig Group. The marine Silurian succession passes without a break up into marine Lower Devonian; the system junction comes near the top of the Stonehouse Formation. The basal marine Devonian beds are followed conformably by the nonmarine Knoydart Formation (about 1,000 ft thick) which contains Lower Devonian (upper Downtonian to middle Dittonian) fish remains. Special studies have also been made of the Bears Brook Volcanic Group and of the structure of the area; repeated movements along the Hollow Fault have resulted in complex folding and faulting in the inland parts of the area. The village of Arisaig (Fig. 1) lies on Northumberland Strait, 15 mi northeast of Antigonish, Nova Scotia. It is now easily accessible from New Glasgow or Antigonish by Nova Scotia Route 45. The area studied extends along the coast from McAras Brook (3 mi southwest . . .

Topography and Superficial Deposits The shoreline northeast from Arisaig Point (Pl. 1) is controlled by the outcrop of the rocks of the Bears Brook Volcanic Group. Between the shore and the Hollow Fault, however, the physical features depend more on the distribution of glacial drift than on the structure of the Paleozoic rocks. The Hollow Fault, as its name suggests, lies along a valley, the base of which is never more than 400 ft above sea level. It is bordered to the southeast by a steep scarp rising to about 800 ft within 1,500 ft of the valley. This scarp marks the margin of a plateau underlain by Ordovician rocks of the Browns Mountain Group. A few hills, formed by monzonite intrusions, rise above the undulating surface of the plateau. Williams (1914, p. 41) and Goldthwait (1924, p. 28) describe the plateau as part of a widespread surface extending over the whole of the Maritime Provinces, the remains of this “Atlantic Upland” now being “preserved only on the more resistant rocks.” Goldthwait follows Daly (1901) in considering the plateau to have been planed off in Cretaceous times; the plateau is certainly post-Triassic, but it could be much younger. In describing the Arisaig region, Williams (1914, p. 44–49) divided the lowlands into “The Shore Front” and the “Carboniferous Lowland.” “The Shore Front” covers the area of Silurian and Devonian rocks around Arisaig, where hills reach a height of 750 ft west of McAras Brook (Pl. 1) and gradually descend northeastward to 650 ft near . . .

Description of Formations Browns Mountain Group The chloritized graywackes, slates, argillites, and hematite beds of this group occur on the southeast side of the Hollow Fault and were not examined closely during the present study. Williams (1914, p. 55) records the inarticulate brachiopod Obolus (Lingulobus) spissa from iron-ore beds and associated micaceous sandstones on the East Branch of Doctors Brook and north of the Little Hollow (which runs parallel to the Hollow three-quarters of a mile southeast of the West Branch of Doctors Brook). The same brachiopod occurs with the Arenig (Lower Ordovician) graptolite Didymograptus nitidus in the iron-ore beds of Belle Isle, Newfoundland (Williams, 1914, p. 55, footnote). The probability of an Ordovician age is strengthened by the fact that the Browns Mountain Group is intruded by monzonite stocks which do not intrude the Silurian, and further, because pebbles from these stocks occur in the Malignant Cove Formation; the Browns Mountain Group is therefore certainly pre-Silurian, and probably older than the Late Ordovician or lower Llandovery Bears Brook Volcanic Group. Browns Mountain conglomerates, with pebbles up to 1 cm, grits, and sandstones occur high on the east bank of Doctors Brook above stations L17 and L21 (Pls. 5A, 5B); and basic tuffs and serpentine are exposed in the Hollow Fault zone along the East Branch of Doctors Brook (stations U88 to U94, Pl. 5C). Other outcrops noted include grits on the West Branch of Doctors Brook 8,000 ft upstream from the junction of the East and West Branches, and epidotized basic rocks on the East . . .

Introduction Although possessing an apparent over-all structural simplicity, the rocks of the Arisaig area bear evidence of five phases of deformation (D 1 ... D 5 ), excluding faulting, and pose a number of vexing structural problems. The area is structurally bounded to the southeast by the Hollow Fault zone (Pl. 1; Fig. 16); this is probably a zone of repeated strike-slip and dip-slip movement. The highland country of Antigonish County, lying southeast of the Hollow Fault zone, is a low-grade metamorphic tract of Ordovician rocks, the Browns Mountain Group. The Hollow Fault zone appears to extend northeastward and southwestward for some miles, and it has been suggested (Wilson, 1963) that it may be part of a fundamental fault zone extending beyond the limits of Nova Scotia. Few minor folds can be seen in the area, but cleavages and lineations are well developed, and a detailed study of these has permitted a tenable structural hypothesis to be constructed. D 1 STRUCTURES The dominant and most persistent fold in the Arisaig area is the upward-facing Arisaig Syncline, the axial trace of which passes northeast-southwest across the southeastern part of the area (Fig. 16). Northeast of Arisaig Brook, the northern, southward-facing limb of the fold is generally overturned and the southern, northward-facing limb of the fold is gently inclined to the west, or northwest (Fig. 17). The regional fold plunge is gently to the southwest, but local, shallow, plunge-reversals occur in major parasitic folds (for example, the Doctors Brook Syncline) on the southern limb. A clear structural pattern emerges . . .

The geological history of the Arisaig region begins with the poorly dated strata of the Browns Mountain Group. Based on fragmentary inarticulate brachiopods, the Browns Mountain has been tentatively dated as Lower Ordovician. The Browns Mountain beds were tightly folded and intruded by a variety of granitic rocks prior to Silurian time. At School Brook Cove, Cape George (Boucot and others, 1959) fossiliferous Middle or Upper Ordovician quartzites occur, which are also intruded by granitic rocks. Unfortunately, the relations of the fossiliferous rocks at School Brook Cove to the adjacent Browns Mountain Group are not established. BEARS BROOK VOLCANIC GROUP Dunn Point Formation The Bears Brook Volcanic Group was formed under predominantly terrestrial conditions, probably during the Upper Ordovician. Subaerial conditions are proved by the complete lateritic soil profiles and the welded tuffs, and it seems likely that the lavas and sediments were emplaced on a basement consisting of the Browns Mountain Group intruded by monzonite stocks. During Dunn Point Formation time, andesites were extruded and flowed southward from a hinterland to the north in a series of sheets and tongues to build up a tableland. The lateritic soil profiles developed on individual andesite flows give information on the time intervals between flows and on the climate of the period. Thickness of laterite profiles is a function of the time interval between flows, and the absence of a laterite on a flow indicates rapid extrusion of successive flows. The climatic factors indicated by the lateritic soils may be summarized as . . .

During the summer of 1959, a geological reconnaissance of the Lochaber area in Antigonish County (Fig. 1) was made by several of the students at the Massachusetts Institute of Technology summer camp under Boucot’s supervision. Although the structural and stratigraphic results are less conclusive then those from nearby areas, this location is geologically strategic, and it is appropriate to list some of the facts observed. A suite of sedimentary, volcanic, and intrusive rock types lying to the west of Lochaber (PL 9) is similar to those assigned elsewhere in Antigonish and Pictou Counties to the Browns Mountain Group. To the east of the rocks assigned to the Browns Mountain Group (Pl. 10), there is a sequence of greenish siltstones and slates, together with a scattering of shelly limestones, which can best be referred to as “Arisaig Group undifferentiated.” These beds have provided enough diagnostic fossils to demonstrate that they include representatives of the following: French River Formation, McAdam Brook Formation?, Moydart Formation (green member only), and the Stonehouse Formation. Table 8 details the brachiopods found in these strata. East and southeast of the beds assigned to the Arisaig Group, a sequence of siltstones and fine-grained sandstones that are predominantly reddish are assigned to the lithologically similar Knoydart Formation. The contact with the underlying Arisaig Group is gradational on Cameron Brook (Pl. 10) as well as on the two unnamed brooks to the north of it. The gradational contact consists of an interbedding over a stratigraphic interval of . . .

Introduction This section includes a compilation and reinterpretation of previous reports on the geology of northern Cape George (Pl. 12) and a detailed account of the geology of School Brook and Cormorant Cliff Coves (Pls. 13, 14). The tip of the Cape is bounded on the east by George Bay and on the north and northwest by Northumberland Strait. The southern limit of the mapped area is a line from Skidmore Brook on the eastern shore to Horseshoe Brook (Pl. 12) on the western shore. The area includes about 16 sq mi and the southern border is 18 mi north of the town of Antigonish on the Cape George (Harbour) Road. The Cape George Road follows the shoreline around the Cape from the Marsh Cove on the east to Livingstone Cove on the west. There is almost continuous outcrop along the shoreline. Wilkie Brook [It is named Wilkie Brook on a Canadian Geological Survey Map of 1893 (Nova Scotia Sheet No. 33). A more recent map of the Canadian National Topographic Series (1953, Sheet No. 11 F/13 West Half) calls it Cape George Brook.] in the southeast part of the area and Livingstone (the 1893 Geological Survey Map calls it Livingstone Brook; the same brook is named Gillis Brook on the 1953 National Topographic Map) and McNeil’s Brook (J. R. Griffin states that McNeil’s Brook is a tributary of Livingstone Brook) in the western part also provide numerous exposures. There are scattered outcrops along many of the brooks and on . . .

The relations of the Silurian and Lower Devonian rocks of Nova Scotia to those of adjacent portions of the northern Appalachians have, until recently, been uncertain. The pre-Carboniferous grain in coastal Maine and New Brunswick is northeasterly, and in southern Newfoundland it is again roughly northeast, but in the intervening area of northern Nova Scotia (excluding Cape Breton Island) it is almost east-west. Recent work in the northern Appalachians, both published (Berry and Boucot, 1970) and unpublished, helps to resolve this situation. It is also possible to interpret the Paleozoic of the Maritimes in the light of plate tectonics (Mc-Kerrow and Ziegler, 1971). The existing relations fall easily into place if it is assumed that the Silurian and Lower Devonian lithofacies units follow the regional grain; and that the northern Appalachians have a salient (in Nova Scotia) and a recess (in New Brunswick), like many other mountain belts. Once this reconstruction is allowed, the problem of setting up reasonable paleogeographic relations (taking into account both physical and biological environmental factors) becomes relatively simple for Maine, New Brunswick, and Nova Scotia, but the detailed relations of Southern Newfoundland to Nova Scotia are not clear. Figure 24 (Eastport Belt of Figure 24 equals the Coastal Volcanic Belt of Berry and Boucot, 1970) indicates three belts which have consistently distinct facies throughout the Silurian and Lower Devonian. These form the basis for the following discussion. None of the pre-Silurian rocks of Antigonish and Pictou Counties can be dated well enough to make it. . .

The evidence of major changes associated with the Devonian earth movements in the northern Appalachians may be examined under the following headings: (1) change from marine to nonmarine facies; (2) unconformities; (3) granites; (4) pebbles in younger conglomerates; (5) structural differences in rocks above and below the unconformities. 1. The changes from marine to nonmarine facies in Nova Scotia and the adjacent areas of the northern Appalachians are summarized in Figure 30. These occurred in the Eastport and in the Arisaig belts well before the onset of any widespread folding. The first change to nonmarine conditions occurred toward the end of Ludlow times in the Eastport belt and in the Arisaig belt. But this was temporary; marine conditions were again widespread during the Pridoli and the early part of the lower Gedinnian. Persistent nonmarine environments only appeared in the Eastport and Arisaig belts after the early Gedinnian. The sea, however, persisted in the southeasterly Annapolis Valley belt from the Ludlow through the lower Emsian. There is thus no direct time relation between the marine to nonmarine facies changes and the onset of folding in Devonian times. 2. Unconformities are present above Early Devonian (Siegenian or later) beds in the two northwesterly belts (where nonmarine conditions had prevailed for some time prior to the folding); the Annapolis Valley belt had marine sedimentation continuing until lower Emsian time. After widespread folding throughout the northern Appalachians, the first postorogenic deposit in the region was the Upper Devonian Perry Sandstone, although still earlier (Middle . . .