INTRODUCTION
hanc oram novissimi maris tunc primum Romana classis circumvecta insulam esse Britanniam affirmavit, ac simul incognitas ad id tempus insulas, quas Orcadas vocant, invenit domuitque. dispecta est et Thule, quia hactenus iussum et hiems appetebat. Footnote 2
It was then that a Roman fleet for the first time circumnavigated this shore of the remotest sea and proved that Britain is an island. At the same time it discovered and conquered the islands called Orcades which were unknown up to that time. Also a clear sighting was obtained of Thule because that was the limit of their orders and winter was approaching.
There are three basic facts in these two sentences: for the first time Britain was found to be an island; the Orcades were previously unknown; a clear sighting was achieved of Thule. The first two of these statements are incorrect. Caesar had stated that Britain was an island while Pomponius Mela c. 40 years before Agricola's maritime expedition and Pliny the Elder, who died in the eruption of Vesuvius in a.d. 79, knew of Orkney.Footnote 3 The purpose of this analysis is to examine the third statement.
THE THULE OF PYTHEAS
In the second half of the fourth century b.c., Pytheas of Massalia (Marseille) made a journey into north-west Europe.Footnote 4 He travelled up the west coast of Britain and recorded that further north lay a place called Thule, according to Cleomedes and Geminos of Rhodes visiting it himself.Footnote 5 We only know of his journey because later writers mentioned it, some in disbelief, others acknowledging the veracity of his account. Even today, there is disagreement about the location of Pytheas’ Thule, the preferred locations being Iceland and the west coast of Norway.
One of our basic references for the location of Thule is the Greek geographer and historian Strabo, who lived during the reign of Augustus.Footnote 6 He did not believe the account of Pytheas, calling him an utter liar. A further record is provided by Pliny the Elder in his Natural History.Footnote 7 Their statements of Pytheas’ account are similar and can be supplemented by the comments of other writers:
(a) Thule is the furthest north of all named islands;Footnote 8
(b) it is six day's sailing north from Brettanike;Footnote 9
(c) it is near the frozen sea; one day's sailing from Thule is a frozen sea; beyond Thule the sea is sluggish and frozen;Footnote 10
(d) it lies where the summer tropic is the same as the arctic circle; there are no nights in summer, when the sun is passing through the sign of Cancer, and conversely no days in winter, some think this happens for six continuous months; Geminos of Rhodes stated that he believed that Pytheas did journey to the north, quoting Pytheas as stating that ‘the barbarians pointed out to us on several occasions the places where the sun lies down. For it happens around these places, that the night is extremely short, two hours in some, three hours in others, so that after the setting, although only a short time has elapsed, the sun straightaway rises again’;Footnote 11
(e) at Thule there was neither sea nor air, but a substance composed of these resembling a sea-lung and he saw this himself.Footnote 12
The conclusions that can be drawn from the above are that Thule lay on the Arctic Circle and that it lay close to the ice cap, or at least seas containing ice floes. In her assessment of the ancient sources, Roseman notes that the summer tropic and the Tropic of Cancer are the same today as they were at the time of Pytheas, but that the north pole would have lain 32o 7’ to the east, while ‘polar ocean areas are subject to temperature inversions, which can increase the effect of refraction to an extraordinary amount’, the result being that ‘the information that Pytheas reported for areas near Thoule (sic) could have been obtained from much further south than the fixed Arctic Circle at 66o N’.Footnote 13 She concludes that the various strands of evidence ‘all allow reasonable certainty that he got as far as 62o N’ but otherwise she refrains from committing herself to a specific location.Footnote 14 62o N is the same latitude as the Faroe Isles, but a degree south of Iceland.
Cunliffe states that the evidence that Thule is Iceland ‘seems unassailable’.Footnote 15 His conclusion rests on the ‘six days north from Britain (not east to Norway), the day of twenty-two or twenty-three hours, the Arctic Circle and the place where the sun sleeps, and the frozen sea’; Norway does not meet those criteria. McPhail prefers a location in Norway for Thule on the basis that Strabo described Thule as inhabited which Norway was and Iceland was not, while Eratosthenes included the place within the inhabited world.Footnote 16 Scholars may not agree whether Thule was Iceland or Norway but the consensus is that it was not Shetland. As Cunliffe puts it, those ‘who support Shetland do so partly because they think it inconceivable that Pytheas could have made a more distant exploration and partly because Tacitus evidently believed Shetland to be Thule’.Footnote 17
BERRICE AND BELGAE
There are two other points of relevance. Pliny stated that one may sail to Thule from Berrice; Mela recorded that Thule is situated opposite the shore of the Belgae.Footnote 18
Pliny completed his survey of Britain with the statement, ‘there are some who mention other islands too, the Scandiae, Dumna, the Bergi and, the largest of all, Berrice, from which one may sail to Thule’.Footnote 19 Scandia appears twice in Ptolemy's Geography where it is recorded that ‘east of the Cimbrian peninsula there are four islands called the Scandian islands, three of them smaller … but one of them very large and the most eastwards at the mouth of the river Vistula … it is properly called Scandia itself’.Footnote 20 The Cimbrian peninsula is Jutland and Scandia therefore equates to Scandinavia. Rivet and Smith stated that Dumna was the Long Island, that is Lewis and Harris, citing its name in Old Irish and Gaelic, Domon, as well as its placing in Ptolemy after Scetis, which they identify with Skye, and before Orkney.Footnote 21 Wolfson preferred the island of Hoy for Dumna, but the important point for this discussion is that he retained it in a northern British context.Footnote 22
In short, in Pliny's list, we have both Scandinavia and an island in north Britain. This does not help us in identifying Berrice. The safest conclusion is to follow Wolfson in accepting that Pliny was confused since he was using two different sources for his list.Footnote 23 Wolfson also noted that Berrice is a Greek name and therefore may derive from Pytheas’ account.Footnote 24 This may imply that the place was not known to later Romans.
The Belgae of Mela is equally perplexing. As Wolfson pointed out, ‘an isolated Thule cannot be adjacent to any coastline, especially one situated west of the Rhine’.Footnote 25 The reference by Mela may simply relate to the belief that Britain lay parallel to the Continent, perhaps stretching as far north-east as the Elbe.Footnote 26
HAEMODAE/ACMODAE
Mela recorded that there were seven Haemodae and that they lay towards Germany; Pliny gave the same number though a slightly different name, Acmodae.Footnote 27 As Pliny mentioned both Acmodae and Thule we might presume that he regarded them as being separate places.Footnote 28 Rivet and Smith argued that Haemodae/Acmodae were distinct from Ebudae which they concluded were the Inner Hebrides, not least as Pliny listed 30 Hebudes.Footnote 29 There are 16 islands in the Shetland archipelago, but seven larger islands. Mela also stated that there were 30 Orkney Islands.Footnote 30 There are in fact 60, but only half of these were inhabited in the 1901 census; this coincidence allows us some confidence in Mela's statements.
Wolfson disagreed with the identification of Haemodae as Shetland by Rivet and Smith, but his argument is prefaced by the statement that the acceptance of Haemodae as Shetland is based upon a reluctance to accept Tacitus’ interpretation of the islands as Thule.Footnote 31 Among Wolfson's objections are that ‘Mela locates his Haemodae … in the Baltic facing Germany’. But he does not. Mela stated that ‘there are 30 Orcades, separated from one another by narrow straits, and seven Haemodae, located towards Germany’. This is where Mela would place the Shetland Isles in view of the misunderstanding of the location of Britain.Footnote 32 Pliny lists ‘40 Orcades, separated from each other by moderate distances, seven Acmodae, 30 Hebudes’, and then he moves to list the islands between Britain and Ireland and then islands off the coast of Germany.Footnote 33 Wolfson argues that Pliny starts with the Orkney Islands and then moves south-west through the Irish Sea, ‘thereby eliminating, by reason of direction, the Acmodae as the Shetland Islands’.Footnote 34 But this is not the case. Pliny goes naturally from Orkney to Shetland and then turns to the west coast. There is no reason to doubt Rivet and Smith's conclusion that the Shetland Isles were Haemodae/Acmodae.
THE THULE OF TACITUS AND PTOLEMY
Ptolemy in his Geography, prepared 50 years after Agricola's expedition and based on the work of Marinos of Tyre, placed Thule above Orkney, locating it at 63o N.Footnote 35 Owing to ancient errors about the size of the degree, this needs to be corrected to 59o 14’.Footnote 36 Tierney pointed out that Marinos’ information on Scotland was extremely poor and that it was doubtful if Agricola's fleet brought back better information about the shape and dimensions of the north of Britain.Footnote 37 The crucial statement that the fleet did bring back was that they had sighted Thule. As a result, Marinos moved Thule from its previous position on the Arctic Circle to a new, southerly, location at 63o (= 59o 14’), which is close to the actual location of the Shetland Isles.Footnote 38 In other words, the independent evidence of Ptolemy for the Shetland Isles being Thule should be discounted as it was based on the assertion of Agricola. However, as Ptolemy did provide the correct latitude for Shetland, this is support for the statement that Agricola's fleet did see the islands.
Thule is one island in Ptolemy's list, whereas Shetland consists of several islands. Rivet explains this ‘by the fact that Agricola's fleet sighted it but did not visit it’.Footnote 39 However, it is perhaps more likely to be the result of Ptolemy transferring Pytheas’ single place to its new location.
As the name of the Shetlands Isles was Haemodae, or the like, why did Agricola not use it if he thought that the fleet had sighted them? Was Tacitus simply trying to suggest that Agricola's forces had reached the furthermost known place on earth?Footnote 40
Tacitus would not have been the only Roman writer to make such an exaggerated claim. Statius in his poetical work Silvae, published about a.d. 95, hailed Vettius Bolanus (governor of Britain from a.d. 69 to 71) for entering Thule on the instructions of the emperor, that is Vespasian.Footnote 41 Elsewhere, Statius averred that Bolanus had reached the Caledonian plains, which is more believable.Footnote 42 Wolfson has noted that the ‘curious similarity, between Bolanus entering Thule on Vespasian's instructions, and Agricola's fleet reaching Thule on Domitian's instructions, raises the question of major distortion’.Footnote 43 Wolfson is taking a step too far here because it can only be an assumption that Domitian issued the order to reach Thule but progress no further. Were the emperor's instructions so detailed? He did not know at this stage of the Roman victory at Mons Graupius, so he could hardly have issued an instruction for action after a battle which he did not know had taken place. The emphasis is surely on the next statement in the Agricola, ‘winter was approaching’, coupled with the statement that Agricola ‘instructed the prefect of the fleet to sail round Britain’.Footnote 44
There is a further puzzlement to the claim of Tacitus. Strabo quoted the ridiculing of Pytheas’ account by Polybius and stated his personal belief that Pytheas was a liar and inventor of information.Footnote 45 Why then, should Tacitus have wished to mention a place which was not believed to exist? We have, however, already noted that others did believe Pytheas, including Geminos of Rhodes and Pliny. We simply do not know which, if any, of these sources was available to Agricola or Tacitus.
DID THE FLEET LAND ON THULE?
dispecta est et Thule, quia hactenus iussum et hiems appetebat Footnote 46
Also a clear sighting was obtained of Thule because that was the limit of their orders and winter was approaching.
‘Wolfson claims that Tacitus’ Thule was Shetland and that Agricola's fleet actually landed there.Footnote 47 He tries to discredit any translation of dispecta est as Thule being observed/looked at/sighted, and to push the meaning of the verb dispicere to ‘examine closely, inspect thoroughly’, hinting at physical action on Thule. This is unacceptable: the verb is a compound of specere meaning ‘look at, see, observe’ and the prefix dis- which adds emphasis; thus dispicere means ‘see clearly, pick out with the eyes’, as the entry in the Oxford Latin Dictionary states. dis- can also mean ‘apart’/‘in different directions’ so dispicere could also mean ‘look in several directions, scan’. Wolfson claims that dispicere ‘is an exact synonym for perspicere, paralleled in all its nuances’. This is not so: it is at times synonymous, but in any case Tacitus did not write perspecta est et Thule. Curiously Wolfson then proceeds to use examples of perspicere rather than dispicere to achieve his translation ‘examine closely, inspect thoroughly’, which immediately brings in the nuance of a physical rather than just an ocular investigation. To reinforce this, he had already introduced the word ‘physical’ by insisting that ‘all compounds of spic/spec must initially embrace the physical concept of looking’. He executes a further sideways move by claiming that ‘dis -spic-spec- has its exact parallel in Greek, ana-scop-scept which means examine well. Nowhere in any lexicon is it translated as sighted’. The Greek verb ana-skopein is not an ‘exact parallel’ or ‘equivalent’ for dispicere, and his triumphant, ‘If this meaning is not found for the Greek equivalent, how can it exist in Latin?’ is illogical. Wolfson's handling of this key passage is tendentious and flawed.
Et simul classis secunda tempestate ac fama Trucculensem portum tenuit …Footnote 48
and at the same time with the support of the weather and their reputation the fleet reached the Trucculensian Harbour …
Wolfson comments, ‘but there appears to be no mention of Shetland where there ought to be [our italics]. The answer lies in the enigmatic and contentious sentence presented thus in the standard editions’.Footnote 49 The only enigma in this first part of the sentence is the identification of the unknown place Portus Trucculensis. The name may have become corrupted. Wolfson sees in it a reference to a harbour on Thule and suggests that Tacitus’ text originally read: ‘et simul classis secunda tempestate ac fama trux Tulensem portum tenuit which may be translated as: And at the same time the fleet, its ruthlessness enhanced by rumour and the favourable weather, reached Shetland Harbour.’Footnote 50 This is palaeographically quite possible because it involves only a very small change from the alternative MS reading portum Trutulensem. But the phrase fama trux is not idiomatic Latin. Trux does mean ‘harsh/savage/ruthless/cruel’, and can be followed by a noun in the ablative. Wolfson quotes two examples from Tacitus with translations: ‘cohortem … tumultu trucem’Footnote 51 — ‘the cohort menacing in its uproar’ and ‘aciem telis et armis trucem’Footnote 52 — ‘a battle-line fearsome with its missiles and weapons’.Footnote 53 However, trux is not found with abstract concepts like fama (‘reputation’). The idea that ruthlessness can be ‘enhanced by … favourable weather’ is at best awkward.
The Agricola 10.4 passage implies that the fleet was instructed just to report a clear sighting or take a careful observation of Thule and nothing more, because of the approach of winter. This immediately renders invalid any attempt to argue that Tacitus meant a landing there.
CONCLUSIONS
The surviving evidence points to Pytheas’ Thule being either Iceland or Norway. The fact that Ptolemy provided the right location for Shetland suggests that it was sighted by Agricola's fleet, but called Thule, perhaps because Ultima Thule was believed to be the most northerly known place. Ptolemy then accepted the identification in the Agricola, transferring the location of Thule from its more northerly correct position to the location of the Shetland Isles. The fact that he described it as one island reinforces the fact that there is no evidence that the fleet landed on ‘Thule’. The emendations of Agricola 10.4 and 38.4 are rejected.
