1. Baltica. This follows the traditional outline as used by many authors, e.g Ziegler et al. (1977), Torsvik et al. (1990), but is newly modified (a) to include both terranes, the southerly Malopolska and northerly Lysogory terranes, of the Holy Cross Mountains, Poland (Cocks & Fortey 1982, 1998); (b) to accept the southern extension of the area to the west of the Urals as far south as the northern part of the Caspian Sea (Cocks 2000, fig 6); (c) with the exclusion of the highest nappes of the Trondheim area, Norway (Cocks & Fortey 1982); (d) with the exclusion of the northern Taimyr Peninsular (see under the Kara Terrane below).
2. Avalonia. The faunal integrity of what are today the North American and European parts of Avalonia precludes the possibility that they represented more than one terrane in Lower Palaeozoic times, particularly when the terrane’s rapid movements over the latitudes during the Ordovician is considered. Thus ”West” and ”East” Avalonia for the two parts are not recognised as separate here and we do not recommend further use of these terms. The North American part of Avalonia (including Eastern Newfoundland, most of the Maritime Provinces of Canada and the eastern United States seaboard as far south as Cape Cod, Massachusetts – Cocks et al. 1997), have been reconstructed to the European part using the fit originally proposed by Bullard et al. 1965 (Torsvik & Rehnstrom 2001a).
3. Perunica. Often termed Bohemia. We have arbitrarily included the area today north of the Barrandian basin of Bohemia within this terrane, so that it has a northern boundary in post-Hercynian times to coincide with the south-eastern boundary of Avalonia.
4. Armorica. This includes the Armorican Massif of Normandy and Brittany, the Massif Centrale and the Montagne Noire areas of France, together with the Iberian Peninsula (apart from the westernmost South Portugese Zone- Cocks, 2000) corrected for the Cretaceous Bay of Biscay opening. Although some authors have shown differential movements within this terrane, we have no evidence to disperse them during our time slice and therefore show them together. The oldest rocks from the South Portugese Zone are middle Devonian and the area is thus excluded from our maps: in any case its terrane affinities are uncertain.
5. Laurentia. The margins of Laurentia have been modified (a) to include Spitsbergen in a position near north-east Greenland: although this archipelago represented two or even three separate terranes, the easternmost area of Spitsbergen can only be considered as having formed part of the margin of the Laurentian craton during early Ordovician times (Fortey 1975); (b) to show the margin near north-east Greenland as based on the assumption of 50% stretching of the crust in post-Devonian times; (c) to accept the Ouachita Front as the southern margin, following the line shown by Keppie & Dostal (1999); (d) to include the Arctic islands of Canada as shown by Natalin et al. (1999) – including the Bennett-Barrovia Terrane (e) with the British Isles north of the Iapetus Suture fitted to North America using the Bullard et al. (1965) fit..
6. Siberia. We follow the outline of Rundqvist & Mitrofanov (1993), which includes most of the Baikalides, which accreted onto the main craton in late Precambrian times. Our reconstructions include south and central Taimyr as part of Siberia, although this is controversial (see below). Siberia collided with Baltica and the Kazakh terranes in Permo-Carboniferous times.
7. Kara. This terrane consists of northern Taimyr and Severnaya Zemlya (Metoelkin et al. 2000). It supposedly collided with central Taimyr (today part of the Siberian Plate) in late Carboniferous-Permian times (Vernikovsky 1996).
8. North China. In general we follow the outline of Rong et al. (1995) and various other authors in the Palaeozoic divisions of China, with North China’s southern boundary at the Qin-Ling Line. The terrane includes the Palaeozoic of the Korean peninsula, but the northern margin is somewhat arbitrary because of the absence of known Palaeozoic rocks from most of Manchuria. Again arbitrarily, we take its modern western margin at the western boundary of the Huanghe Region.
9. South China. This outline also follows Rong et al. (1995) and previous authors and we include the modern south-eastern coastline and the island of Hainan (which has only Mesozoic to Recent rocks) to aid recognition of the terrane in our reconstructions.
10. Tarim. We follow the boundaries shown by Zhou & Chen (1992), with the northern boundary delimited by the Tian-Shan fold belt. The modern Qaidam Terrane to the south is entirely post-Lower Palaeozoic and is therefore omitted from our diagrams.
11. Sibumasu and Annamia. These terranes, sometimes termed the Shan-Thai and Indo-China Terranes respectively, largely follow the margins shown by Scotese & McKerrow (1990) and Fortey & Cocks (1998). Minor amendments, particularly at the northern boundary of Indochina, have been made so that that boundary is congruent with the southern boundary of South China, with which the terrane amalgamated in the late Jurassic.
12. Gondwana. The Gondwanan core (the heavy dashed line in Fig. X) consists of Africa, Arabia, Madagascar, Greater India, most of Antarctica, most of Australia, New Guinea, and most of South America. New Zealand is shown divided into North and South, following Lottes & Rowley (1990). Southernmost South America and much of western Antarctica are not shown since they were post-Lower Palaeozoic accretions. Florida was also an integral part and is included using the fit of Lawver & Scotese (1987). Gondwana formed at around 550 Ma and its early dispersal history commenced with the rifting off of Avalonia in Arenig-Llanvirn times.
13. Apulia and the Hellenic Terrane. We follow Stampfli et al. (1998) in the identification of Apulia (southern Italy) and the Hellenic Terrane (largely Greece) as separate terranes, but there is little evidence for their separate movement within Perigondwana during our time period.
14. Middle Eastern Perigondwanan terranes. The Taurides Terrane comprises most of central and southern Turkey. Although tectonically recognised as separate, the terrane was apparently situated in a position relative to Arabia comparable to that seen today. We take its boundaries from many authors, including Dean et al. (2000). The Pontides of northern Turkey are shown as a separate terrane to the west of the Taurides. The Lut, Alborz and Sanand terranes (mostly Iran), which are separated by various authors (e.g. Raumer et al. 2001), can be grouped as one for this paper (Millson et al. 1996), since there are no good faunal or palaeomagnetic data for our time period which warrants their distinction. The Afghanistan and Karaku terranes can also be treated in a similar way and are also shown as part of Greater Gondwana in our diagrams.
15. Himalayan terranes. A number of terranes within today’s Himalayas were discrete elements in Palaeozoic times and include the Lhasa Terrane (south Tibet) and the Qiangtang Terrane (north Tibet). However, these are tied to Gondwana in our reconstructions. Their outlines follow Watson et al. (1987) but the modern sutures and other evidence indicate that they did not accrete to Greater India until the early Tertiary.
16. Precordillera. This small terrane, which is today in eastern South America (northwestern Argentina) is shown here with an arbitrary outline.
17. Mexican terranes. There are at least three, and probably more, terranes which today occupy the area between the Ouachita Front of the United States and north-western South America; the most important is the Oaxaquia Terrane. We show these terranes together and linked to Gondwana in our reconstructions, in a similar way to Florida: they did not accrete to Laurentia until Permo-Carboniferous and later times (Keppie & Dostal 1999).