4 Originally the White Rock Formation was defined in the Wolfville area (Figs. 1, 2) mainly on the basis of an upper and lower quartzite bed separated by slate (Faribault 1909, 1921; Crosby 1962). The formation was extended to the southwest into the Torbrook, Bear River, Cape St. Marys, and Yarmouth areas based primarily on the occurrence of one or more such quartzite beds (Faribault 1909, 1921; Smitheringale 1960, 1973; Taylor 1965, 1969). The forma tion in the Torbrook, Cape St. Marys and Yarmouth areas also includes volcanic units absent in the Wolfville and Bear River areas, as noted by Smitheringale (1960) in his redefinition of the formation.

5 The relationship of the White Rock Formation to the underlying Halifax Group has been variously interpreted. Smitheringale (1960, 1973) thought that slate units in the White Rock Formation were indistinguishable from those in the Halifax Group and concluded that there was no hiatus in sediment accumulation between the two units, although he noted that erosion could have occurred locally. Depending on the map area, he placed the base of the White Rock Formation at the bottom of the lowest quartzite unit or of the volcanic package. Schenk (1972) proposed a conformable contact and identified a diamictite (interpreted as a til-lite) in the lowermost part of the White Rock Formation in the Wolfville and Cape St. Marys areas. Subsequently, Schenk and Lane (1981) placed the diamictite in the uppermost part of the Halifax Group and considered it to be of Late Ordovician age. Long (1991, 1992, 1994) again assigned the diamictite units to the White Rock Formation but reinterpreted them as non-glacial in origin, citing evidence such as absence of dropstones and lack of lateral continuity. He also showed that the diamictite does not occur in a consistent stratigraphic position; instead he considered these clast-supported conglomerate beds to be gravity-flow deposits formed in a non-glacial slope setting. Long (1991, 1992) suggested that they are of Early to Middle Ordovician age based on inferred stratigraphic continuity with the Halifax Group. Keppie (2000) showed the White Rock Formation to extend down into the Middle Ordovician and to be locally conformable or disconformable on the Halifax Group. In the Torbrook area the provincial geological map (Keppie 2000) showed the White Rock Formation overlying the Goldenville Group, although we have confirmed that in fact it overlies the Halifax Group, as elsewhere (Fig. 1).

6 Culshaw and Leisa (1997) interpreted the White Rock Formation in the Cape St. Marys area to lie in a synclinorium with extensive shear zones in the Halifax Group on both limbs, and the whole area was referred to as Cape St. Marys shear zone. Later, Culshaw and Dickson (2015) showed that the White Rock Formation forms a simple syncline and re-stricted the Cape St. Marys shear zone to its western limb and, based on structural evidence, showed that the original contact was an angular unconformity now embedded in the ductile shear zone.

7 In the Yarmouth area, the White Rock Formation has also been inferred to occupy a synclinal structure with mainly sedimentary components, including prominent quartzite units, toward the bottom and mafic and felsic volcanic components toward the top (Taylor 1965, 1967; Sarkar 1978; Lane 1979). The relationship with the underlying Halifax Group was variously inferred to be conformable or uncertain, but Culshaw and Leisa (1997) documented shear zones on both the western and eastern margins, which they termed the Cranberry Point and Chebogue Point shear zones, respectively. Cooling ages of muscovite in the Cranberry Point shear zone, as well as in the Cape St. Marys shear zone to the north, indicated a middle Carboniferous (Alleghanian–Variscan) age for the deformation (Culshaw and Reynolds 1997).

8 Fossils are sparse in the White Rock Formation. Lane (1975) reported the occurrence of shelly fossils in the Yarmouth area, and cited A. Boucot (personal communication, 1971) for identifying a castellated rhynchonellid brachiopod that has a maximum age of Caradocian but is probably much younger. Ferguson (in Williams et al. 1985) incorrectly located this fossil occurrence in the Cape St. Marys area. Lane (1979) reported shelly fossil hash in a quartzite outcrop in the Wolfville area, but this observation was not confirmed during the present study. In the Torbrook area, Ludlow graptolite specimens, identified as Monograptus sp. cf. M. tumescens (Smitheringale 1960), were collected from slate above the upper quartzite in the White Rock Formation and assigned to the Kentville Formation. Smitheringale (1973) later assigned these slates to the upper part of the White Rock Formation. Subsequent work (e.g., Bouyx et al. 1997) reassigned these graptolite-bearing slates to the Kentville Formation.

9 The Kentville Formation was defined in the Wolfville area (Fig. 2) as a unit of interbedded marine slate and siltstone that conformably overlies the White Rock Formation and conformably underlies the New Canaan Formation (Ami 1900; Crosby 1962). The Kentville Formation was also mapped in the Torbrook area (Smitheringale 1960, 1973; Taylor 1965, 1969) but not in the areas to the southwest (Smitheringale 1960, 1973; Taylor 1965).

10 Graptolites and other fossils were reported in the Kentville Formation from the Fales River section, the area east of Spinney Brook, and Torbrook River (Fig. 3) by Smitheringale (1960) and dated as Ludlow. Bouyx et al. (1997) also described graptolites and other fossils collected from the lower part of the Kentville Formation in Fales River and assigned a Late Wenlock to Early Ludlow age (i.e., around the Middle–Late Silurian boundary). Samples collected from the top of the Kentville Formation in Spinney Brook yielded Pridoli (Late Silurian) microfossils (Bouyx et al. 1997). Late Silurian vertebrate and crinoid remains in the upper part of the White Rock Formation in the Bear River area indicate that the upper part of the formation in that area is age-equivalent to the upper Silurian Kentville Formation in the Torbrook area (Blaise et al. 1991; Bouyx et al. 1997).

11 Strata overlying the Kentville Formation in the Tor-brook area and the White Rock Formation in the Bear River area were named the Torbrook Formation by Hick-ox (1958). They consist of interbedded marine mudstone, siltstone and sandstone with minor fossiliferous limestone and ‘iron-formation’ (Smitheringale 1960). Volcanic rocks occur in the formation in Spinney Brook (Jensen 1975). Where the Kentville Formation in the Bear River area was not recognized, the contact between the White Rock and overlying Torbrook formations was placed above a massive quartzite, based on the presence of fossils interpreted to be Early Devonian (Lane 1979), but shown later to be late Silurian (Bouyx et al. 1997). In the Torbrook/Spinney Brook area (Fig. 3) the contact between the Torbrook Formation and the underlying Kentville Formation was defined as the first appearance of thinly bedded sandstone (Jensen 1975). The age of the Torbrook Formation is well constrained by fossils. It contains abundant fauna that Boucot (1960), Cumming (in Smitheringale 1973) and Jensen (1975) considered to be Gedinnian to Siegenien, and maybe as young as Early Emsian (Early Devonian). In the sandy basal part of the Torbrook Formation on Spinney Brook, Bouyx et al. (1997) collected samples bearing Urnochitina urna, a chitinozoan of Pridoli age. This fossil indicates that the Torbrook Formation extends down to the Upper Silurian and overlaps in age with the New Canaan Formation, consistent with the field observations during this mapping project (e.g., White et al. 2017b). In the Bear River area, the fossil assemblage includes brachiopods, crinoids and bivalves, and indicates an Early Devonian age (Smitheringale 1973; Blaise et al. 1991; Bouyx et al. 1997). The youngest detrital zircon grain from a sandstone sample from near the top of the Torbrook Formation in the Torbrook area yielded a U–Pb age of 388 ± 8 Ma (Murphy et al. 2004). If reliable, this age suggests a maximum depositional age of latest Early Devonian, perhaps as little as 10 million years prior to intrusion by the South Mountain Batholith and permitting only a narrow window for deformation and metamorphism associated with the Neoacadian orogeny.

12 Crosby (1962) assigned the metasedimentary and mafic metavolcanic rocks that conformably overlie the Kentville Formation in the Wolfville area to the New Canaan Formation. The New Canaan Formation contains crinoids, brachiopods (e.g., Conchidium) and corals that indicate a possible Middle Silurian age (Wilson in Crosby 1962). Boucot et al. (1974) placed the formation in the latest Silurian (Pridoli).

13 Schenk (1995b) correlated the upper part of the volcanic sequence in the Yarmouth area with the New Canaan Formation of the Wolfville area. However, this interpretation is not supported by U–Pb ages from the felsic volcanic rocks (Keppie and Krogh 2000; MacDonald et al. 2002; White et al. 2017a), which show that the volcanic rocks in the Yarmouth area are early Silurian, whereas the New Canaan Formation is of late Silurian to Early Devonian age based on fossils.

14 Rocks in the Rockville Notch Group were regionally metamorphosed to lower greenschist facies (locally to lower amphibolite facies in the Yarmouth area) and deformed into northeast-trending folds with associated axial planar cleavage during the Middle Devonian Neoacadian orogeny (van Staal 2007; White et al. 2001, 2007, 2017b). In addition, the group was intruded by the ca. 373 Ma South Mountain Batholith and related granitoid rocks (e.g., Clarke et al. 1997, 2004), which resulted in a well-developed contact metamorphic aureole containing mineral assemblages characteristic of up to hornblende-hornfels facies (e.g., Campbell 1999). Although protoliths are typically easily recognized, appropriate metamorphic names are used here to describe these rocks.

15 In the Wolfville area, the Rockville Notch Group consists of the White Rock, Kentville, and New Canaan formations and occurs on the limbs of the Canaan synclinorium, which comprises the New Minas and White Rock synclines (Fig. 2). The White Rock Formation in the Wolfville area overlies either the Hellgate Falls Formation or the Elderkin Brook Formation of the Halifax Group (Fig. 2), and varies in thickness from about 100 to 200 m. In the White Rock syncline, the White Rock Formation is composed of a series of quartzite beds 1.5 to 10 m thick (Fig. 4a), separated by thin (10 cm to 5 m thick) beds of black slate. These slate beds are like those in the underlying Halifax Group (Crosby 1962; Smitheringale 1960, 1973), although in most outcrops they contain phosphate nodules which were not observed in the Halifax Group slate (White 2010c; White and Barr 2010).

16 Along strike the quartzite and slate beds vary in thickness with an overall lensoid distribution. It is unclear whether this ‘pinch and swell’ character is the result of the original sedimentary environment or of later deformation. Above the quartzite package is a unit of grey to light-green cleaved metasiltstone to slate, locally interlayered with thin (5 to 15 cm thick) beds of pale sulphide-rich metadolostone. Above the slate package is another series of quartzite beds (1 to 5 m thick) interlayered with green-grey metasiltstone and metasandstone. The contact with the overlying Kentville Formation is gradational, with the quartzite beds thinning and then disappearing over 50 m.

17 In the New Minas syncline, the stratigraphy in the White Rock Formation is slightly different. The base of the formation is marked by a series of up to six quartzite beds with thicknesses of 2 to 10 m, interlayered with dark grey to red-brown metaconglomerate to breccia (Fig. 4b), black metasiltstone and slate. In some localities quartzite overlies the Halifax Group, whereas in others metaconglomerate overlies the Halifax Group. The metaconglomerate is the diamictite/tillite of Schenk and Lane (1981), which they included in the upper part of the Halifax Group rather than in the White Rock Formation. The quartzite continues above this unit to the base of the Kentville Formation but is interlayered with grey cleaved metasiltstone to slate and lacks meta-conglomerate. In this area, it is difficult to map a middle ‘slate’ unit or the double quartzite member as defined by Crosby (1962) and Smitheringale (1960, 1973). As in the White Rock syncline, the quartzite varies in thickness along strike with many of the beds displaying a lenticular morphology, tens to hundreds of metres across.

18 Like the White Rock Formation, the overlying Kentville Formation occurs on the limbs of the New Minas syncline and White Rock syncline (Fig. 2). The formation consists of pale grey-green, laminated metasiltstone and slate overlain by black slate (Fig. 4c). The lower slate unit locally contains pyritiferous limestone nodules. The upper contact is marked by a basal metabasaltic flow in the overlying New Canaan Formation, well exposed in Elderkin Brook. The metabasalt is dark green-grey and amygdaloidal, and grades laterally into mafic lithic metatuff and metavolcanic breccia. In contrast, Clark (1977) placed the base of the New Canaan Formation below a mainly calcareous metasiltstone and slate unit. However, these rocks are included here in the Kentville Formation. Overlying the metavolcanic rocks is a succession of dark grey to black siliceous metasiltstone, metasandstone, and metalimestone, all of which are locally interlayered with amygdaloidal metabasalt. The metalimestone and metasandstone locally contain fossils that indicate a latest Silurian (Pridoli) age (Boucot et al. 1974).

19 Clark (1977) subdivided the formation into four metasedimentary and metavolcanic units. James (1998) remapped the area and logged core drilled by Smith (1992). He subdivided the formation into seven informal units; three mainly metavolcanic, three metasedimentary, and one mixed metavolcanic/metasedimentary, the details of which are too small to be indicated on Figure 2. James (1998) noted that towards the top of the formation, the metasiltstone contains abundant sulphide minerals and supported the interpretation of Smith (1992) that this mineralization is similar to sediment-hosted exhalative (SEDEX) deposits.

20 Crosby (1962) estimated that the thickness of the New Canaan Formation exceeds 300 m. Clark (1977) subsequently estimated that it is 725 m, although he included a slate unit in the lower part of the formation. James (1998) included the slate unit in the underlying Kentville Formation and placed a minimum thickness of the New Canaan Formation at 530 m, an interpretation consistent with our observations.

21 In the Torbrook area the Rockville Notch Group occurs on both limbs of the Torbrook syncline, overlying the Elderkin Brook and Hellgate Falls formations of the Halifax Group (Fig. 3). Along the southeastern limb of the syncline, the base of the White Rock Formation is marked in places by a rhyolitic metatuff or flow that is overlain by amygdaloidal and pillowed metabasaltic flows. Locally the basalt displays peperitic texture with metalimestone as the interstitial material (Fig. 5a). The rhyolite at the base of the formation in Fales River yielded a U–Pb zircon crystallization age of 442 ± 4 Ma (Keppie and Krogh 2000); within error it spans the Ordovician-Silurian boundary, currently placed at 443.4 ±1.5 Ma (Cohen et al. 2013 updated 2017). Along strike the metavolcanic rocks are missing and the basal unit is a quartzite and metasiltstone package up to 5 m thick. However, in the Factorydale area (Fig. 3) the basal quartzite/metasiltstone is interlayered with minor (<1 m thick) felsic crystal metatuff (Cullen et al. 1996). On Fales River, a well-bedded, heavily bioturbated, quartz arenite (~50 m thick; Fig. 5b) with numerous grazing and burrowing trace fossils is stratigraphically above the basalt. This unit was not observed elsewhere.

22 Along the northwestern limb of the syncline the base of the White Rock Formation is well exposed along the west bank of Nictaux River (Fig. 3) and marked by a 5-m-thick section of black Mn- and P-rich metasiltstone interlayered with thin (5 to 40 cm thick) phosphatic ironstone beds. To the southwest these Mn-rich metasiltstone and ironstone beds were once mined (the Parker and Armstrong ore beds of Fletcher 1905a, b). These rocks are overlain by a 3-mthick quartzite bed. Because the base of the White Rock Formation was previously defined at the first quartzite, these Mn-, P-, and Fe-rich beds under the quartzite were included in the upper part of the Halifax Group by MacDonald and Ham (1994) or of the Goldenville Group by Keppie (2000). However, along strike to the northeast, this basal section is dominated by thick amygdaloidal metabasalt flows containing a 2-m-wide quartzite bed, indicating that these rocks are part of the White Rock Formation. The quartzite is capped by a thin (<0.5 m thick) Mn-rich ironstone that was previously mined (in the Foster Pit of Fletcher 1905a, b). A thin-section from the black Mn-rich meta-siltstone on Nictaux River contains a jaw/toothwhorl with two cusps, of questionable acanthodian affinity; if correctly identified, this fossil suggests an age no older than Wenlock (S. Turner, written communication, 2016), inconsistent with the apparent position of these rocks at the base of the White Rock Formation; additional work is thus required to clarify stratigraphic relations in this area.

23 The upper part of the White Rock Formation varies. In Fales River it appears to have the ‘double quartzite member’, with two quartzite beds separated by black featureless metasiltstone. The black slate is like that in the overlying Kentville Formation in the area. In Nictaux River a green mafic lithic metatuff is at the top of the formation. This unit was observed only at low-water conditions. Smitheringale (1973) described a basaltic flow in a similar stratigraphic position in the Nicholsville area, but based on its texture, the unit is more likely to be one of the numerous gabbroic sills that occur in the area. Like the White Rock Formation in the Wolfville area, many of the quartzite and metavolcanic beds are lenticular along strike, as was noted by Smitheringale (1973). The White Rock Formation in the Torbrook syncline area varies in thickness from 300 m to 550 m.

24 The Kentville Formation in the Torbrook area ranges in thickness from 500–600 m and consists of dark grey to black or pale green-grey, featureless to laminated metasiltstone and slate interlayered with thin laminae to thinly bedded pale metasandstone and metasiltstone. In the Fales River and Nictaux River sections, close to the bottom of the unit, black metalimestone nodules are locally common (Fig. 5c). The contact with the overlying Torbrook Formation is gradational. In Spinney Brook the green slate in the upper part of the Kentville Formation becomes increasing interlayered with thicker, coarser-grained meta-sandstone. This change marks the base of the Torbrook Formation. In Nictaux River the upper contact is marked by the presence of a black fossiliferous metalimestone that forms the lowermost unit in the Torbrook Formation. Graptolites and other fauna of Ludlow age were reported in the Kentville Formation in the Fales River section (Smitheringale 1973) and the area east of Spinney Brook (Jensen 1975). Bouyx et al. (1997) also described graptolites and other fossils collected from the base of the Kentville Formation in Fales River and assigned an age of Late Wenlock to Early Ludlow (Middle– Late Silurian transition). Samples collected from the top of the Kentville Formation in Spinney Brook yielded Pridoli (Late Silurian) microfossils (Bouyx et al. 1997).

25 The Torbrook Formation occurs in the core of the Tor-brook syncline and is about 1500 m thick. The Spinney Brook section was described in detail by Hickox (1958), Smitheringale (1973), and Jensen (1975). The basal part of the section in Spinney Brook consists of a coarsening-upward metasandstone package interlayered with metasiltstone. Rare macrofossils occur in the finer grained metasandstone. In the Nictaux River the base of the Torbrook Formation is marked by a fossiliferous limestone and black metasiltstone. The remainder of the formation consists of interbedded metasandstone, metasiltstone, slate and metalimestone that are typically heavily bioturbated and fossil-rich. Phosphate nodules are common. A charac-teristic feature of the Torbrook Formation is the presence of quartz-rich and oolitic chamositeand francolite-bearing ironstone beds (Marshall 2011) that are locally fossiliferous and phosphate-rich (e.g., Fletcher 1905b). Mafic lithic metatuff is present near the middle of the formation (mv on Fig. 3), as Jensen (1975) also noted at a similar stratigraphic level in Spinney Brook. The formation contains abundant fauna that Boucot (1960) and Cumming (in Smitheringale 1973) considered to be Gedinnian to Siegenien age, and perhaps as young as Lower Emsian (Early Devonian). In the sandy basal part of the Torbrook Formation on Spinney Brook, Bouyx et al. (1997) collected samples bearing Urnochitina urna, a chitinozoan of Pridoli age. This fossil indicates that the Torbrook Formation extends down to the Upper Silurian and overlaps in age with the New Canaan Formation, consistent with the field observations during this mapping project. The top of the formation may be as young as 388 ± 8 Ma, the age of the youngest detrital zircon grain reported by Murphy et al. (2004).

26 The Rockville Notch Group in the Bear River area was originally interpreted to be preserved in a syncline (Taylor 1969; Smitheringale 1973; Lane 1975; Doyle 1979), but White et al. (1999) showed, as a result of new outcrops and more detailed mapping, that younging is consistently to the east to where the rocks are intruded and cut off by the South Mountain Batholith (Fig. 6). The White Rock Formation in the area consists mainly of grey to dark grey slate and metasiltstone, with minor light grey to white metasandstone. The slate tends to be darker, less laminated, and has a higher proportion of silt than in the underlying Bear River Formation of the Halifax Group. The Elderkin Brook and Hellgate Falls formations of the Halifax Group are missing in the Bear River area (Fig. 6). The metasiltstone in the White Rock Formation is laminated to massive and forms beds 5 to 20 cm thick. The metasandstone forms laminated to massive beds and discontinuous lenses 1 to 10 cm thick. However, one notable metasandstone layer is 30 m thick, locally laminated and cross-bedded, and discontinuous along strike. Rare grey metalimestone beds, less than 15 cm thick, occur at the base of the formation, and calc-silicate lenses are present in places. The top of the formation contains lenses of intraformational conglomerate less than 3 cm thick. It appears to be gradational with the overlying Torbrook Formation, which occurs in a belt between the White Rock Formation and South Mountain Batholith. The upper part of the White Rock Formation in the Bear River area yielded upper Silurian vertebrate and crinoid remains (Bouyx et al. 1997), indicating that the upper part of the formation in that area is age-equivalent to the Kentville Formation in the Torbrook area. However, the total thick-ness of 1000–1200 m is more than the combined thickness of the White Rock and Kentville formations in the Torbrook and Wolfville areas.

27 The Torbrook Formation consists predominantly of metasiltstone and calcareous metasiltstone, with minor slate, metasandstone, metalimestone and rare ironstone. Fossil-rich horizons are common in the metasiltstone, which is interlayered with grey slate. White quartzite beds become more abundant up-section, and are up to 2 m thick and locally cross-bedded. In earlier studies (e.g., Smitheringale 1973), these quartzite beds were presumably interpreted as belonging to the underlying White Rock Formation, and hence the presence of synclinal structure was inferred. Thin marble beds and nodules and calc-silicate layers are also interlayered with the metasiltstone and quartzite. The ironstone is exposed only in old mine trenches northeast of Bear River and consists of black, faintly laminated to massive iron-rich metasandstone. The total thickness of the formation is about 1550 m, but the top is not exposed. In the Bear River area, the fossil assemblage includes bra-chiopods, crinoids and bivalves, and indicates an Early Devonian age (Smitheringale 1973; Jensen 1975; Blaise et al. 1991; Bouyx et al. 1997).

28 The Rockville Notch Group is represented only by the White Rock Formation in the Cape St. Marys area, where it is interpreted to be preserved in a synclinal structure (Fig. 7). On the coast it outcrops only on the western limb of the inferred structure and, along with adjacent rocks of the Halifax Group, it is highly sheared (Culshaw and Leisa 1997; Culshaw and Dickson 2015). At the exposed western contact, sheared and deformed slate and metasiltstone of the Halifax Group (Bear River Formation of White 2010a) are overlain by highly strained felsic metavolcanic rocks (Fig. 8a). The lithic metatuff at the base (Fig. 8b) apparently corresponds to the diamictite (tillite) of Schenk and Lane (1981), which they described as a pebbly, 1-m-thick mudstone. Long (2016) described these fragments as bombs, and U–Pb dat-ing confirmed the igneous origin of the unit and yielded an age of 442.9 ± 2.1 Ma (White et al. 2017a). The metatuff is in sharp contact with sheared and altered mafic metavolcanic rocks that are in places amygdaloidal. The mafic rocks are in sharp contact with thin felsic metatuff and thick quartzite, in which locally preserved cross-bedding confirms younging to the east. The quartzite is overlain by a section of mafic metavolcanic rocks more than 20 m thick. The section includes several flows with amygdaloidal tops and locally well preserved peperitic features in which the intermingled sedimentary rocks are quartz-rich metasandstone; the section includes a thin layer of cross-bedded metasandstone about 1 m thick. The metavolcanic rocks are in turn overlain by quartzite and tuffaceous quartzite in a section about 40 m thick. They grade upward into slate with abundant quartzite beds and lenses. The slate was interpreted by Long (2016) to belong to the Kentville Formation. It locally contains unidentified trace fossils. A mafic rock exposed in a separate outcrop in the beach consists of pyroxene-bearing gabbro and appears to be a younger dyke or sill, in contrast to the volcanic interpretation of Long (2016), who assigned it to the New Canaan Formation. The White Rock Formation is not exposed on the eastern limb of the syncline, although the distribution of units in the Halifax Group is consistent with that structure (Culshaw and Leisa 1997; White et al. 2001). The White Rock Formation, as exposed on the shore at Cape St. Marys, is about 140 m thick.

29 As at Cape St. Marys, the Rockville Notch Group in the Yarmouth area is represented by only the lowermost part of the White Rock Formation (Fig. 9). It forms a northeast-trending belt with a maximum width of about 15 km in the well-exposed coastal section. The southeastern margin is in faulted contact with slate that may belong to the Halifax Group (Fig. 9). The fault is inferred to be a brittle structure that juxtaposes chlorite- and biotite-grade rocks of the Halifax Group against staurolite-grade, typically schistose rocks of the White Rock Formation. However, in the coastal section north of Chebogue Point, the metamorphic grade in the White Rock Formation appears lower than in the same unit to the north, and the rocks are less sheared than those of the adjacent Halifax Group. The brittle fault is located within the broader Chebogue Point shear zone of Culshaw and Liesa (1997), previously referred to as the Deerfield shear zone by Keppie and Dallmeyer (1995), that affects both the White Rock Formation and the older rocks.

30 On the northwest margin, the position of the contact between the White Rock Formation and Halifax Group is poorly defined because of lack of exposure, both inland and on the coast. The most southerly outcrops of Halifax Group slate at Cranberry Point are separated by nearly 3 km, with no outcrop from the most northerly outcrop of metasedimentary rocks assigned to the White Rock Formation. The contact position is inferred by a strong aeromagnetic signature that appears to be associated with the Halifax Group. The Halifax Group and possibly the adjacent White Rock Formation in that area are deformed in the Cranberry Point shear zone (Culshaw and Leisa 1997).

31 The White Rock Formation in the Yarmouth area has been metamorphosed to amphibolite facies, and the combined effects of metamorphism and deformation have obscured primary sedimentary and volcanic features in many outcrops. However, macroscopic relict igneous features such as phenocrysts, amygdales, lithic and crystal clasts, and bombs indicate that the metavolcanic units include both flows and metatuffaceous rocks. Possible pillow structures are preserved locally and, overall, the sequence appears to have been deposited in a mainly subaqueous setting, with minor subaerial eruptions suggested by coarsely vesicular mafic flows and rare ignimbritic felsic units that contain vestiges of flow-banding and flattened pumice fragments. Some basaltic flows preserve amygdaloidal tops, and sedimentary and reworked metatuffaceous units locally preserve cross-bedding and graded bedding that enable the determination of younging direction. The White Rock Formation in the Yarmouth area previously has been interpreted to occur in a synclinal structure (Taylor 1967; Lane 1979; Hwang 1985; Culshaw and Liesa 1997; MacDonald 2000). However, White et al. (2001) and MacDonald et al. (2002) reported that the rocks appear to be consistently right-way-up and younging to the southeast, without repetition of distinctive lithologic units. Hence the rocks are divided into six fault-bounded members with clear stratigraphy within each member, but uncertain stratigraphic relations between members (Fig. 9). The members may be in part laterally equivalent. Overall, the rock types are like those preserved in the Cape St. Marys section, but much thicker (minimum ~ 6 km).

32 At the base of the Chegoggin Point member is a tuffaceous metasandstone and metasiltstone beds with calcsilicate beds and lenses. Upward, these rocks become interlayered with laminated phyllite and slate, the latter containing as yet unidentified trace fossils. The phyllite and slate grade upward into metasiltstone with distinctive garnetiferous schist layers up to 2 m thick that contain locally abundant chloritoid, and are associated with sulphide-rich metasandstone layers. Thick-bedded, massive to well-laminated quartzite occurs near the top of the unit (Figs. 10a, b). Above the quartzite is a mainly mafic metavolcanic member. The lower part consists of mafic crystal and lithiccrystal metatuff interlayered with minor intermediate to felsic crystal to crystal-lithic metatuff and metabasaltic flows with relict pillow-like textures (Fig. 10c). It also includes a boudinaged quartzite layer like those in the underlying unit. Zircon grains from a sample of felsic crystal-lithic metatuff yielded a concordia age of 435.8 ± 2.1 Ma (White et al. 2017a, b). The metavolcanic rocks are overlain by tuffaceous metasandstone interlayered with thinly bedded metasiltstone and metasandstone. These rocks grade upward into andesitic lithic to lithic-crystal metatuff.

33 In its lower part, the Overton member consists of mafic flows and lithic metatuffs. The middle part consists of well-laminated metasiltstone and metasandstone, locally interlayered with amphibole-rich tuffaceous metasandstone and mafic flows and crystal metatuff, and includes a massive rusty quartzite layer 2 m wide. The metasandstone sequence contains calc-silicate lenses, abundant trace fossils such as Chondrites, Teichichnus, and Palaeophycus, and rare brachiopod and gastropod shells (Figs. 11a–d). The trace fossil assemblage does not provide any significant age constraints or palaeogeographic information (S. Jensen, written communication, 2017). A castellated rhynchonellid brachiopod from this area has a maximum age of Caradocian (Late Ordovician), but it is probably much younger (A. Boucot, personal communication, 1971 in Lane 1975). Several brachiopod and gastropod specimens were recollected from the same location and the fossil assemblage was reassigned to “no older than Early Silurian” (A. Boucot, written communication, 2001). Upsection, the metasedimentary package is in sharp contact with a thick package of mafic lithic to lithic-crystal metatuff and associated flows interlayered with minor tuffaceous metasandstone. The uppermost part of the member contains intermediate crystal metatuff and metaconglomeratic layers.

34 The Cape Forchu member is dominantly mafic metavolcanic rocks. The lower part consists of massive flows and mafic lithic metatuff, locally interlayered with tuffaceous metasandstone. The middle part of the unit is well exposed at Cape Forchu and consists of thinly bedded mafic lithic metatuff and tuffaceous metasandstone with well-preserved original volcanic and sedimentary features, and minor calc-silicate lenses. The upper part of the unit is separated from the middle part by a shear zone and consists of massive mafic flows and lithic to crystal-lithic metatuff. Locally the flows contain pillow-like structures and brecciated flow tops.

35 The Dayton member comprises mainly metasedimentary rocks. The lower part consists of thickly bedded metasandstone with abundant calc-silicate nodules and interbedded biotite-spotted slate. The middle part consists of biotite-spotted slate interbedded with metasiltstone and, near the top of the unit, the metasedimentary rocks are interlayered with mafic lithic metatuff and flows and tuffaceous metasandstone.

36 The Sunday Point member is in gradational contact with the metasedimentary rocks of the Dayton member and is a sequence of mafic lithic metatuff and flows interlayered with minor tuffaceous metasandstone. The mafic flows locally contain peperite-like structures, and many have amygdaloidal upper parts, 1–2 m wide, that indicate younging to the east. Cross-bedding and graded bedding in metasand-stone layers also indicate younging to the east. The member contains two distinctive felsic crystal-metatuff layers, each about 10 m thick. The lower felsic layer can be traced inland for several kilometres in sporadic outcrops. The upper felsic layer yielded a U–Pb (zircon) age of 438 +2/-3 Ma (Mac-Donald et al. 2002).

37 The Government Brook member contains a mixture of metasedimentary and metavolcanic rocks. Through most of the unit, the metasedimentary rocks are staurolitebearing micaceous phyllite and schist locally interlayered with thin horizons (<30 cm thick) of metasandstone and amphibolite. The metavolcanic rocks are mafic to rarely intermediate lithic metatuff interlayered with tuffaceous metasandstone, and occur in a central belt flanked by metasedimentary rocks. Exposures along the coastal sec-tion north of Chebogue Point are unlike inland exposures of the unit in that they lack phyllite and schist, and consist of metasiltstone and metasandstone with interlayered mafic rocks that could be either flows or sills. No evidence for younging direction was observed, and these rocks appear to be in faulted contact with more deformed slate of the Halifax Group to the east (Fig. 9).