1. Introduction
For more than fifty years, the stratigraphical community has been progressing, albeit slowly, towards an agreed global standard accepted by the IUGS International Commission on Stratigraphy by which to correlate the Lower Palaeozoic. Since this protracted process is now largely complete for the Ordovician and Silurian and partially finished for the Cambrian, it seems timely to communicate that standard more widely and show how it relates to the British terms which have been well known and much used internationally for the best part of two centuries.
Thus we have constructed Figure 1 to include the new global standard; the British regional units which form the basis for the revised Ordovician (Fortey et al. Reference Fortey, Harper, Ingham, Owen, Parkes, Rushton and Woodcock2000) and Silurian (Cocks, Holland & Rickards, Reference Cocks, Holland and Rickards1992) correlation charts; the earlier Geological Society correlation charts (Cocks et al. Reference Cocks, Holland, Rickards and Strachan1971; Williams et al. Reference Williams, Strachan, Bassett, Dean, Ingham, Wright and Whittington1972; Cowie, Rushton & Stubblefield, Reference Cowie, Rushton and Stubblefield1972); and the column used locally in North America for Laurentia, all calibrated against real ages. Those radiometric ages come from a variety of sources not reviewed here, and have been rounded to the nearest one million years, which we feel is appropriate for rocks of these periods in view of the margins of error usually involved, but Sadler, Cooper & Melchin (Reference Sadler, Cooper and Melchin2009) have published useful ages and graptolite and conodont zones for the Ordovician and Silurian. We have concentrated on the final global stratigraphical decisions reached and how they correlate with the older framework used in the British Isles (including Ireland) and the Laurentian part of North America, rather than dwelling in detail on how those decisions were made, which are the topics of many substantial publications.
Figure 1. Correlation chart for the Lower Palaeozoic, showing ages in millions of years (Ma); the global standard divisions; the British regional subdivisions, modified from Cowie, Rushton & Stubblefield (Reference Cowie, Rushton and Stubblefield1972) for the Cambrian, Fortey et al. (Reference Fortey, Harper, Ingham, Owen, Parkes, Rushton and Woodcock2000) for the Ordovician, and Cocks, Holland & Rickards (Reference Cocks, Holland and Rickards1992) for the Silurian; the subdivisions previously published by The Geological Society of London (Cowie, Rushton & Stubblefield, Reference Cowie, Rushton and Stubblefield1972; Williams et al. Reference Williams, Strachan, Bassett, Dean, Ingham, Wright and Whittington1972; Cocks et al. Reference Cocks, Holland, Rickards and Strachan1971); and the units used locally in Laurentia.
The divisions will now be reviewed briefly in turn. Not all are classified at the same level in stratigraphical terminology: some are series and some stages, but we present them as ratified by the International Commission on Stratigraphy. The four official series within the Cambrian range from the Terreneuvian to the Furongian; it is proposed that they should be divided into ten stages (Fig. 1), some of which are named. There are no formal Lower, Middle and Upper subdivisions of the Cambrian. In contrast, the Ordovician is divided into three series, the Lower to include the Tremadocian and Floian stages, the Middle to include the Dapingian and Darriwilian, and the Upper Series to include the Sandbian, Katian and Hirnantian stages. The Silurian is in theory formally divided into the Lower Silurian Subsystem to include the Llandovery Series and the Wenlock Series and the Upper Silurian to include the Ludlow Series and the Pridoli Series; however, since it is now known that the Llandovery alone lasted for more than half of Silurian time, we do not favour the use of those terms.
2. The Precambrian–Cambrian boundary
In the nineteenth century neither Sedgwick nor Murchison gave precise definitions of the base of the Cambrian; for the next hundred years, before reliable radiometric dating and during which stratigraphical correlation depended on biostratigraphy, the virtually unfossiliferous Precambrian rocks were poorly correlated. The base of the Cambrian was reviewed in a major symposium (Rodgers, Reference Rodgers1956) and, after much debate, it was stabilized at Fortune Head, southeast Newfoundland, Canada. The basal stratotype (GSSP) there, proposed by Narbonne et al. (Reference Narbonne, Myrow, Landing and Anderson1987), is at the first occurrence of the trace fossil Phycodes (now Treptichnus) pedum 2.4 m above the base of Member 2 of the Chapel Island Formation (Brasier, Cowie & Taylor, Reference Brasier, Cowie and Taylor1994). That horizon, which has been radiometrically dated at 543 Ma (Grotzinger et al. Reference Grotzinger, Bowring, Saylor and Kaufman1995), is at a lower level than the bases of most of the rocks previously considered Lower Cambrian in the British Isles.
3. The Cambrian System
Although originally named by Sedgwick (in Sedgwick & Murchison, Reference Sedgwick and Murchison1835) for an assemblage of fossiliferous rocks below the Silurian, the Cambrian System only reached something like its present definition when Lapworth (Reference Lapworth1879) published his paper that erected the Ordovician. Recently, the global division of the Cambrian into Lower, Middle and Upper units has been replaced by recognition of four series and (provisionally) ten stages, but at present only half of these divisions are formally named. Figure 1 shows a tentative correlation of this incomplete standard with the six named stages of the Laurentian Cambrian, as summarized by Palmer (Reference Palmer1998). These divisions are intended to facilitate correlation across the Laurentian Craton from its interior to the edge of the shelf, and their recognition exploits mass extinction events, particularly of endemic trilobites. In contrast, the international standard has drawn mostly on shelf-edge faunas, since they tend to be more widely dispersed over several terranes.
3.a. Terreneuvian Series and Fortunian Stage
The base of both series and stage is the same as the base of the Cambrian System, as defined in Newfoundland (Landing et al. Reference Landing, Peng, Babcock and Moczydłowska-Vidal2007; Section 2, above). Some strata in Britain that are now correlated with the earliest Cambrian were until recently regarded as late Precambrian (McIlroy, Brasier & Moseley, Reference McIlroy, Brasier and Moseley1998).
3.a.1. Stage 2
Stage 2 of the Cambrian is not yet defined or named, but the level at which its base will be taken is likely to be governed by consideration of archaeocyathid zones and/or the distribution of ‘small shelly fossils’ (or ‘SSF’, a variety of sclerites, many from animals of debated zoological affinity), supported by other criteria (Babcock et al. Reference Babcock, Peng, Geyer and Shergold2005, p. 103). Figure 1 arbitrarily shows the base of the Comley Series extending down into Stage 2.
3.b. Series 2
Series 2 (and the concomitant base of Stage 3) is not yet defined, but the appearance of the earliest trilobites is expected to be a significant criterion in formalizing its definition (Babcock et al. Reference Babcock, Peng, Geyer and Shergold2005, p. 103); however, other correlative criteria will be critical in recognizing Series 2. While the definition of Stage 3 is pending, it is impossible to point to definite equivalents in Britain. The Montezuman Stage in Laurentia is similarly based on the earliest trilobites there, and its base is tentatively aligned with the base of Series 2 in Figure 1.
3.b.1. Stage 4
The base of Stage 4 will likely be defined by reference to the first appearance of a trilobite, possibly a redlichiid or olenellid (Babcock et al. Reference Babcock, Peng, Geyer and Shergold2005), supported by other criteria. It is expected that, when defined, it will be possible to relate such a base to the appearance in Britain of the olenelloid trilobite Callavia (Fletcher, Reference Fletcher2006). In Britain this is near a level at the top of the Lower Comley Sandstone at Comley, Shropshire. However, it is yet to be determined how that level will correlate with the base of the Dyerian stage in Laurentia.
3.c. Series 3
Although the congruent bases of Series 3 and Stage 5 are not yet agreed, a range of possible levels, based on various criteria, has been discussed (Geyer, Reference Geyer2005). Even the youngest of these proposed levels is older than the base of the St David's Series as originally described by Cowie, Rushton & Stubblefield (Reference Cowie, Rushton and Stubblefield1972), and the older horizons are well down within the Comley Series.
3.c.1. Drumian Stage
The base is taken at the appearance of the agnostoid Acidusus (or Ptychagnostus) atavus in the base of a limestone layer 62 m above the base of the Wheeler Formation at Stratotype Ridge in the Drum Mountains, Utah, USA (Babcock et al. Reference Babcock, Robison, Rees, Peng and Saltzman2007). The most general use of the old British term Menevian corresponds fairly closely with the Drumian.
3.c.2. Guzhangian Stage
The base of this, the seventh stage of the Cambrian, is taken at the appearance of the agnostoid Lejopyge laevigata 121.3 m above the base of the Huaqiao Formation in the Luoyixi section in Guzhang County, Hunan Province, China. In Britain, the Guzhangian spans a widespread hiatus in the upper part of the St David's Series and extends up into the basal zone of the Merioneth Series of Cowie, Rushton & Stubblefield (Reference Cowie, Rushton and Stubblefield1972). The Laurentian Marjumian Stage extends from below the base of the Drumian up to the base of the Furongian Series.
3.d. Furongian Series
This series extends up to the base of the Ordovician. Its base, and that of the Paibian Stage, is taken at the appearance of the distinctive and widespread agnostoid Glyptagnostus reticulatus at a level 396 m above the base of the Huaqiao Formation in the section at Paibi, Huayuan County, NW Hunan (Peng et al. Reference Peng, Babcock, Lin, Chen and Zhu2004a,Reference Peng, Babcock, Robison, Lin, Rees and Saltzmanb). In Britain this corresponds closely to the base of the Olenus trilobite zone (Rushton, Reference Rushton1983). Thus the Furongian corresponds to the Merioneth Series, but excluding its basal Agnostus pisiformis Zone, and the Paibian roughly to the old term Maentwrog (or Maentwrogian) Stage of many authors (e.g. Harland et al. Reference Harland, Armstrong, Cox, Craig, Smith and Smith1990; Orłowski, Reference Orłowski1992). The Laurentian Steptoean Stage correlates closely with the Paibian, but the correlation of its upper limit will depend on the eventual definition of Stage 9.
3.d.1. Stage 9
The formal definition of the middle stage of the Furongian is expected to be dependent on the appearance of the agnostoid Agnostotes orientalis and/or the elviniid trilobite Irvingella major (Babcock et al. Reference Babcock, Peng, Geyer and Shergold2005, p. 104). In Britain these may correspond to a level close to the base of the Parabolina spinulosa Zone, and roughly to the base of the inadequately-defined ‘Dolgellian’ Stage of Harland et al. (Reference Harland, Armstrong, Cox, Craig, Smith and Smith1990). The Sunwaptan stage of Laurentia has its base at the Irvingella major Zone and ranges up through part of Stage 10.
3.d.2. Stage 10
This is not yet defined, but Babcock et al. (Reference Babcock, Peng, Geyer and Shergold2005, p. 104) suggested that the base might be taken at the appearance of the widespread agnostoid Lotagnostus americanus, which includes among its synonyms L. trisectus (Salter), a species recorded from the upper parts of the Dolgellau Formation in North Wales and in equivalent strata elsewhere in Wales and England. The base of the North American Ibexian Series occurs within the upper part of Stage 10.
4. The Cambro-Ordovician boundary
Lapworth (Reference Lapworth1879) defined the base of the Ordovician as at ‘the base of the Arenig’, but unfortunately that was not crisply delineated except as an unconformity at the base of transgressive sandstones such as the Garth Grit Member in North Wales (Howells & Smith, Reference Howells and Smith1997). International opinion has varied as to the level of the boundary, and in particular, whether or not the Tremadoc or Tremadocian Stage should be included within the Cambrian or the Ordovician; for example, the Geological Society of London decided some years ago (Fig. 1) to place it within the Cambrian (Cowie, Rushton & Stubblefield, Reference Cowie, Rushton and Stubblefield1972; Williams et al. Reference Williams, Strachan, Bassett, Dean, Ingham, Wright and Whittington1972). That stage is now within the Ordovician; however, although the Welsh name has been retained, its base has been transferred to a slightly different level defined in Newfoundland (see below).
5. The Ordovician System
Lapworth (Reference Lapworth1879) created the system to include the upper part of Sedgwick's Cambrian and the lower part of Murchison's Silurian, but, although its original top at the base of the Llandovery has been approximately retained, Lapworth's original base has now been extended downward to include the old Tremadoc. All the global stages listed below have been defined in the past ten years; some are defined by graptolites and others by conodonts. Our concern here is principally to relate the classical usage of Ordovician subdivisions based on British type localities to the new international classification. There is no reason why the refined British subdivisions used in the revised Correlation Chart of the Ordovician System in the British Isles (Fortey et al. Reference Fortey, Harper, Ingham, Owen, Parkes, Rushton and Woodcock2000) should not continue in use for Avalonian and Gondwanan stratigraphy, although the traditional use of five ‘series’ must be abandoned in favour of the standardized three series (Lower, Middle and Upper Ordovician) as specified by the International Commission on Stratigraphy. The former British ‘series’ will henceforth be considered as regional stages, and the former stages might be used as regional substages. The correlation between the different schemes runs into trouble only in the mid-part of the Ordovician, where the later Arenig and Llanvirn do not bear much resemblance to the new stages spanning that part of the system.
5.a. Base of the Ordovician, the Tremadocian Stage and the Lower Ordovician Series
After nearly twenty years work by a Working Group of the IGCP on the Cambrian–Ordovician boundary, the GSSP for its base and that of the Ordovician System is defined by the lowest occurrence of the conodont Iapetognathus fluctivagus, just above the base of the lindstromi conodont Biozone within Bed 23 of the Green Point Formation at Green Point, western Newfoundland, Canada. Although the appearance of a conodont species is used for the formal definition of the boundary, the appearance of the oldest planktonic graptolites belonging to the Rhabdinopora (formerly Dictyonema) flabelliformis species group is very close to, but not exactly coincident with, the same horizon. This is fortunate for international correlation because many parts of the world lack conodont-yielding limestones, especially in those areas of high Ordovician palaeolatitudes on Gondwana. Thus the appearance of these distinctive graptolites is a de facto boundary marker in many sections and maps in Europe, North Africa and South America, and is easily recognized in the field. In North Wales, the most complete fossiliferous section spanning the Cambrian–Ordovician boundary was described by Rushton (Reference Rushton1982), and may be taken as the reference in the historical type area for the Ordovician. This area is also of importance in geochronology because interbedded ashes very close to the base of the Ordovician have yielded a zircon age of 489 Ma (Landing et al. Reference Landing, Bowring, Davidek, Rushton, Fortey and Wimbledon2000).
The modern meaning of Tremadocian corresponds closely with its traditional usage, but does not equate exactly with the usual definition of the Ordovician Ibexian (quondam Canadian) Series of North America. Intercalation of conodonts, graptolites and trilobites in off-shelf strata in the western USA and China indicates that the base of the ratified Tremadocian lies above the base of the Ibexian as drawn at the base of the Skullrockian Stage in the Great Basin (Miller et al. Reference Miller, Evans, Loch, Ethington, Stitt, Holmer and Popov2003).
5.b. Floian Stage
The base of the stage is at the bottom of the approximatus graptolite Biozone at the GSSP at Diabasbrottet, near Flo, Sweden (Bergström, Löfgren & Maletz, Reference Bergström, Löfgren and Maletz2004). This corresponds closely with the base of the British Arenig Series as revised by Fortey et al. (Reference Fortey, Harper, Ingham, Owen and Rushton1995). However, not all of the old Arenig Series is embraced by the Floian Stage, and the succeeding Dapingian would also have been considered later or upper Arenig in the old sense. The defining basal horizon of the Floian is the first appearance of the widespread and distinctive graptolite Tetragraptus (Etagaptus) approximatus, which has a broad distribution within the Ordovician equatorial region in marine shale basins. It follows upon a distinctive, equally widespread late Tremadocian graptolite assemblage typified by Araneograptus murrayi. Tetragraptus (E.) approximatus did not, however, extend into platform successions, such as carbonate platforms, nor did it extend into the high palaeolatitudes such as typified western Ordovician Gondwana. However, it has been recognized for some time that a distinctive acritarch ‘flora’ known as the messaoudensis–trifidum assemblage acts as a useful surrogate for the basal horizon of the Floian stage, as reviewed in Molyneux, Raevskaya & Servais (Reference Molyneux, Raevskaya and Servais2007), in what were cool water, higher-latitude localities where Floian strata lack distinctive graptolites. Correlation into the limestone successions of Laurentia is difficult, but the base of the Floian is probably close to the base of the Tulean Stage of the Ibexian Series (Ross et al. Reference Ross, Hintze, Ethington, Miller, Taylor and Repetski1997).
5.c. Dapingian Stage and the Middle Ordovician
The base is defined at the base of the Baltoniodus triangularis conodont Biozone within the Dawan Formation at Huanghuachang, western Hubei Province, China (Wang et al. Reference Wang, Stouge, Erdtmann, Chen, Li, Wang, Zeng, Zhou and Chen2005). This has brought a conclusion to a long-standing debate about the definition of the Middle Ordovician, which had been used in different senses in different regions of the world for many years. In Europe the Middle Ordovician had been regarded as commencing with the Llanvirn Series in the classical usage, and in Laurentia at the Whiterockian, which is older, although there was confusion about the correlation of its type section in Whiterock Canyon, Nevada, with deeper water graptolitic successions. Unfortunately, the base of the Dapingian does not have a simple relationship with the British standard, and falls within the upper part of the Arenig ‘series’ of traditional usage. Because B. triangularis does not appear in the thick clastic succession attributed to the Arenig in the type Welsh areas (e.g. Fortey & Owens, Reference Fortey and Owens1987), the recognition of the base of the Middle Ordovician requires indirect evidence. Cooper, Crowley & Rushton (Reference Cooper, Crowley and Rushton2008) recorded Isograptus victoriae lunatus, which they took as basal Middle Ordovician, from just above a volcanic rock whose age is 473 Ma. However, if Stouge et al. (Reference Stouge, Wang, Li, Chen and Wang2008) are correct on the basis of graphical correlation in inferring that the appearance of B. triangularis ‘matches with the first appearance of Isograptus victoriae victoriae’ (which succeeds lunatus), then this would be a useful graptolite to use as a surrogate marker fossil since it has been recognized in the Lake District in Britain in strata underlying the Isograptus gibberulus Biozone (Rushton in Cooper et al. Reference Cooper, Fortey, Hughes, Molyneux, Moore, Rushton and Stone2004). The victoriae Biozone is correlated with the lower part of the Fennian (local) substage in South Wales (Cooper et al. Reference Cooper, Fortey, Hughes, Molyneux, Moore, Rushton and Stone2004, p. 8). The appearance of I. victoriae approximately coincides with that of the long-ranging, and very widespread graptolite Pseudotrigonograptus ensiformis, which is known from the rushtoni Biozone of the Fennian, and its range overlaps with the very widespread species Xiphographus svalbardensis, a species also known from the early Whiterockian of Laurentia. Refinement of the definition of Whiterockian may permit its base to equate with the Dapingian. In summary, the base of the Dapingian and of the Middle Ordovician falls within the upper part of the Arenig ‘series’ of former usage, and the Dapingian may informally be thought of as ‘late Arenig’. The base of the series probably lies close to that of the Fennian, but precise correlation is not resolved.
5.d. Darriwilian Stage
The base is defined 22 m below the top of the Ningkuo Formation at Huangnitang, Zhejiang Province, China, at the base of the austrodentatus graptolite Biozone (Chen & Bergström, Reference Chen and Bergström1995). This was the first of the new Ordovician stages to be defined, and the eponymous graptolite Undulograptus austrodentatus (which is an early, but not the earliest biserial form) is remarkably widespread on all Ordovician palaeocontinents. Its appearance in various European sections always shortly precedes the appearance of numerous graptolites of the traditional Llanvirn ‘series’ including the well-known ‘tuning forks’ of the Didymograptus artus species group, varied glossograptids and others. The Darriwilian includes the uppermost part of the former Arenig ‘series’, together with the classical Llanvirn and Llandeilo ‘series’. Fortey et al. (Reference Fortey, Harper, Ingham, Owen and Rushton1995) had already acknowledged the brevity of the Llandeilo ‘series’ and had amalgamated it with an extended Llanvirn ‘series’; both now pass into the Darriwilian. The base of the traditional Llanvirn ‘series’ was recognized by Nielsen (Reference Nielsen, Webby, Paris, Droser and Percival2004) as a widespread maximum flooding surface, and is therefore likely to retain its utility in international correlation.
5.e. Sandbian Stage and Upper Ordovician
The base is defined at the base of the gracilis graptolite Biozone at Fågelsång, near Sandby in southern Sweden (Bergström et al. Reference Bergström, Finney, Xu, Pålsson, Wang and Grahn2000). The first appearance of the graptolite Nemagraptus gracilis has been recognized as a biostratigraphical event of global significance ever since the time of Elles (Reference Elles1922). As a name, the Sandbian Stage was preferred by the Ordovician Subcommission over the Caradocian, which has had a long history, and which had also been redefined as commencing at the base of the N. gracilis Biozone following the revision of Fortey et al. (Reference Fortey, Harper, Ingham, Owen, Parkes, Rushton and Woodcock2000). This horizon poses relatively few problems for international correlation, and is at a convenient level. However, there is some evidence from graphic correlation that the first appearance of Nemagraptus gracilis itself can be regionally diachronous (Bettley, Fortey & Siveter, Reference Bettley, Fortey and Siveter2001).
5.f. Katian Stage
The base is at the bottom of the caudatus graptolite Biozone at Black Knob Ridge, 7 km northeast of Atoka in southeastern Oklahoma, USA (Goldman et al. Reference Goldman, Leslie, Nõlvak, Young, Bergström and Huff2007). This horizon has been recognized in the Hartfell Shales of Scotland (Zalasiewicz, Rushton & Owen, Reference Zalasiewicz, Rushton and Owen1995) and in central Wales, and thus can be approximately correlated with the type British sequence at or near the base of the Cheneyan Stage. As a result, the Katian appears long, equating to the upper half of the Caradoc as well as more than half of the Ashgill in Britain.
5.g. Hirnantian Stage
The base is defined at the base of the extraordinarius graptolite Biozone within the Wufeng Formation at Wangjiawan North section, western Hubei Province, China. Although of short duration, the Hirnantian is widely recognized as coinciding with a globally important glaciation, accompanied by a major mass extinction. Because of the pervasive nature of this event, most rock sections spanning the Hirnantian are dominated by glacial influence (and this includes the British section which gives it its name); hence the preferred GSSP in a deeper water succession at a relatively low palaeolatitude. The stage is marked also by isotopic excursions, and by characteristic shelf ‘crisis’ fossil faunas, which makes it comparatively easy to recognize on a global basis. It is essentially the same as the Hirnantian Stage, which has been used in Britain for some time (references in Fortey et al. Reference Fortey, Harper, Ingham, Owen, Parkes, Rushton and Woodcock2000).
6. The Ordovician–Silurian boundary
Lapworth (Reference Lapworth1879) defined the Ordovician System as up to ‘the base of the Llandovery’, and there has since been little disagreement. However, the type Llandovery is poorly fossiliferous near its base, and so Cocks, Toghill & Ziegler (Reference Cocks, Toghill and Ziegler1970) proposed that the base of the Llandovery, and thus also the bases of the Silurian System and the Rhuddanian Stage, should be transferred from Llandovery (Wales) to Dob's Linn (Scotland), where Lapworth had established many of his graptolite zones, which are the most accurate correlation tools at these levels. The boundary level suggested was at the base of the persculptus Biozone near the base of the Birkhill Shale Formation, following earlier workers such as Jones & Pugh (Reference Jones and Pugh1935). To examine the situation in more detail, the International Commission on Stratigraphy set up in 1974 an international Ordovician–Silurian Boundary Working Group to review the most appropriate stratotype and boundary level. That group concluded that the ‘Golden Spike’ should be placed at Dob's Linn, but at the base of the succeeding acuminatus Biozone at 1.6 m above the base of the Birkhill Shale Formation [Grid reference NT 1961 1574]. This was principally so that the widespread Hirnantia brachiopod fauna and the end-Ordovician Hirnantian glacial episode should be included within the Ordovician rather than straddling the boundary across the two systems. The boundary in several sections across the world, including the stratotype, and the distributions of various fossil groups around the boundary are described in the many papers within Cocks & Rickards (Reference Cocks and Rickards1988). More recently, the acuminatus Zone sensu lato has been subdivided into an underlying ascensus Biozone and an overlying acuminatus Zone sensu stricto, but the International Commission on Stratigraphy voted in 2007 to retain the Dob's Linn systemic boundary at the same place as previously, and indicated that the correlation of that ‘Golden Spike’ point was at the base of the ascensus Biozone.
7. The Silurian System
Murchison (Reference Murchison1835) erected the name Silurian, but, like the Cambrian, its modern concept dates from Lapworth's founding of the Ordovician in 1879. All of the Silurian is formally correlated today by graptolite biozones. In limestone sequences, particularly in Laurentia, in which graptolites are rare or absent, conodonts have also been used, but the Silurian graptolite zones are much shorter. Since the influential correlation chart of Berry & Boucot (Reference Berry and Boucot1970), North Americans have employed the British nomenclature (Fig. 1), but before then the Medinan (or Alexandrian), Niagaran and Cayugan Series were used in Laurentia. Useful global Silurian correlations were published within the volume edited by Landing & Johnson (Reference Landing and Johnson2003). In many older publications, particularly in continental Europe, Gothlandian was used as a system name, but that was internationally abandoned in favour of Silurian in 1960. Comparably, for some years the terms Valentian and Salopian were in vogue for graptolitic sequences in Britain and elsewhere, but the former has been abandoned in favour of Llandovery and the latter was essentially equivalent to Wenlock and Ludlow combined. The terms Llandovery, Wenlock and Ludlow, together with their included stages defined in Britain noted below, have been adopted as the global standards by the International Commission on Stratigraphy (Bassett, Reference Bassett1985).
7.a. Llandovery Series
Cocks, Toghill & Ziegler (Reference Cocks, Toghill and Ziegler1970) had divided the Llandovery into four stages: the Rhuddanian, Idwian, Fronian and Telychian, the latter three with their defined bases in the type Llandovery area of Carmarthenshire, Wales. However, those were deemed unsatisfactory by the international Silurian Subcommission, and thus, after further work in the Llandovery area, a revised scheme was devised by Cocks et al. (Reference Cocks, Woodcock, Rickards, Temple and Lane1984), which was subsequently voted to be the global standard (Bassett, Reference Bassett1985).
7.a.1. Rhuddanian Stage
See Section 6, Ordovician–Silurian boundary.
7.a.2. Aeronian Stage
This has its base in the Trefawr track section, near Llandovery (Grid reference SN 8380 3953), within the Trefawr Formation and at the base of the triangulatus Biozone, as proved by the graptolites occurring there (Cocks et al. Reference Cocks, Woodcock, Rickards, Temple and Lane1984).
7.a.3. Telychian Stage
This also has its base in the Llandovery area, Carmarthenshire, and was originally placed by Cocks, Toghill & Ziegler (Reference Cocks, Toghill and Ziegler1970) at the base of Jones’ (Reference Jones1925) C4 subdivision (the base of the Wormwood Formation) in the Cefn Cerig Road section southeast of Llandovery. However, after the more detailed revision by Cocks et al. (Reference Cocks, Woodcock, Rickards, Temple and Lane1984), it was moved upwards to near the base of the overlying C5 subdivision of Jones, and the revised stratotype is near the top of the Wormwood formation in a small disused quarry close to the Cefn Cerig Road [SN 7743 3232]. The stratotype horizon represents the transition between the evolving brachiopods Eocoelia intermedia and E. curtisi, a horizon which correlates approximately with the base of the turriculatus graptolite Biozone sensu lato; the latter has now been subdivided into a lower guerichi and an upper turriculatus sensu stricto Zone (Loydell, Reference Loydell1992).
7.b. Wenlock Series
Bassett et al. (Reference Bassett, Cocks, Holland, Rickards and Warren1975) defined two stages within the type Wenlock area of Shropshire, England, and these were accepted as the global standard (references in Holland & Bassett, Reference Holland and Bassett1989).
7.b.1. Sheinwoodian Stage
The base of this stage coincides with the base of the Wenlock Series and its stratotype lies at the conformable junction between the Hughley Shale Formation and the Buildwas Formation in Harley Brook at Grid reference [SO 5688 9839], at a level which approximately correlates with the centrifugus Biozone. Although characteristic graptolites have not been found at the stratotype section itself, it has yielded abundant and stratigraphically useful microfossils, particularly acritarchs and chitinozoans (Mullins & Aldridge, Reference Mullins and Aldridge2004).
7.b.2. Homerian Stage
The base lies within the Coalbrookdale Formation and is defined within a stream section in a small tributary to Sheinton Brook [Grid reference SO 6194 0204] at the base of the lundgreni Biozone. Bassett et al. (Reference Bassett, Cocks, Holland, Rickards and Warren1975) also defined two formal chronozones within this stage, the Whitwell Chronozone, whose base coincides with the base of the Homerian, and the overlying Gleedon Chronozone, whose base is exposed in a track near Eaton Church [SO 5016 8999] at the base of the nassa Biozone; however, the International Commission on Stratigraphy decided that international correlation was not yet fine enough for such chronozones to be useful.
7.c. Ludlow Series
This term was originally coined by Murchison but was defined more precisely and split into four divisions, the Eltonian, Bringewoodian, Leintwardinian and Whitcliffian, by Holland, Lawson & Walmsley (Reference Holland, Lawson and Walmsley1963). The Silurian Subcommission decided that there were too many to be appropriate as international stages, and thus two international stage names were eventually accepted (Holland et al. Reference Holland, Lawson, Walmsley and White1980).
7.c.1. Gorstian Stage
The stratotype base of this stage and also the Ludlow Series is defined at an old quarry in Pitch Coppice, near Ludlow, Shropshire [Grid reference SO 4723 7298] at the conformable junction between the Much Wenlock Limestone Formation and the Elton Formation at a level which correlates with the base of the nilssoni Biozone. The stage includes the Eltonian and Bringewoodian of the former classification.
7.c.2. Ludfordian Stage
The base is at Sunnyhill Quarry, near Ludlow [SO 4950 7255] at the conformable junction between the Bringewood and Leintwardine formations at a level estimated to correlate with the base of the incipiens Biozone. The stage includes the Leintwardinian and Whitcliffian of the former classification.
7.d. Pridoli Series
The diacritical marks of the Czech ‘Přídolí’ are often omitted from general publications, and we follow that practice here. Because of the relatively short duration of the series, it has not been divided into stages. Its base is within the Pozáry Formation at Pozáry, near the southwestern boundary of Prague, Czech Republic, and coincides with the base of the ultimus Biozone (Kříž et al. Reference Kříž, Jaeger, Paris and Schönlaub1986). Through a complex correlation web in which ostracods and conodonts are important, that level essentially correlates with the base of the long-established Downton Series of Britain, named after Downton Gorge, near Ludlow, Shropshire (White & Lawson, Reference White and Lawson1989; Miller, Reference Miller1995).
8. The Siluro-Devonian boundary
In 1960 the Siluro-Devonian Boundary Working Group (originally termed ‘Committee’) was the first such to be established, and its conclusions were internationally endorsed by the nascent International Commission on Stratigraphy in 1972. These were later summarized in Martinsson (Reference Martinsson1977), which includes many descriptions of boundary sections across the world and the distributions of key floral and faunal groups. Even Murchison (e.g. Reference Murchison1866) had recognized ‘Passage Beds’, later termed the Downton Series, above the Silurian in the Welsh Borderland and underlying the unequivocally Devonian Old Red Sandstone. For many years, most British stratigraphers took the base of the Devonian at the thin Ludlow Bone Bed, which lies at the very top of the Ludlow succession in the type Ludlow area, and above which the Downtonian is largely non-marine. In the late nineteenth and early twentieth centuries, graptolites were thought to have become extinct at the end of the Silurian, and thus graptolite-bearing beds were naturally considered to be Silurian in continental Europe. However, when distinctive Old Red Sandstone fish were correlated with those later ‘Silurian’ French, Belgian and German deposits, it became clear that graptolites had persisted through the Lower Devonian. Following much debate, the stratotype boundary was eventually agreed at 8 m above the base of the Lochkov Formation within the classic Barrandian sequence at Klonk, near Prague, Czech Republic, and at the base of the uniformis graptolite Biozone. A more recent borehole drilled across the boundary has documented further the graptolite and sedimentological data (Chlupàč & Hladil, Reference Chlupàč and Hladil2000). In Britain, the Downtonian is therefore at least mainly of Silurian age, although the boundary between it and the overlying Dittonian part of the Old Red Sandstone has not yet been precisely correlated with the stratotype basal Devonian section in the Czech Republic.
9. Postcript
Geologists have hitherto often despaired at the lack of stability and international agreement on standard correlation terms. However, despite some tension which has occurred at times during the long period of gestation while the standards outlined above have been formulated and agreed, that period seems happily now nearly over, and we can hope that the schemes here will be of permanent use in industrial as well as academic applications.
Acknowledgements
We are very grateful to numerous colleagues in Britain and abroad for discussion, particularly those who have served with us on the Cambrian, Ordovician and Silurian Subcommissions, and also the members of the Ludlow Research Group. Charles Holland merits special acclaim for his adroit steering of the Silurian subdivisions. We thank Abigail Brown for drafting the figure, and Jan Zalasiewicz and two anonymous referees for comments.
