(1) It is improbable that (as Müllenhoff asserts) so capable an astronomer as Pytheas should have made a mistake of several hours when he gave the length of the night as two or three hours. There is little intrinsic probability in the conjecture that he had overcast weather all the time he was in the north of Scotland and Orkney, and therefore relied on the approximate statements of the natives, which he did not fully understand, and which when translated into Greek measures of time might produce gross errors. But it is worse when we look at it in connection with Hipparchus’s statements from Pytheas, that in Britain the longest day was of eighteen hours, and nineteen hours in a region (i.e., Shetland) farther north, where the sun at the winter solstice stood less than three cubits above the horizon. Unless he has given the latter region a long extension to the north, he must have made several conflicting statements about the same region. It will be seen that this leads us to a violent and arbitrary alteration of the whole system of information, which is otherwise consistent.
(2) The assertion that Pytheas did not himself say that he had been in the country where the night was two and three hours long, conflicts with the words of Geminus. Cleomedes also tells us that Pytheas is said to have been in Thule.
(3) The definite statements in a majority of the authorities that Thule lay within the Arctic Circle and was the land of the midnight sun, also exclude the Shetland Isles. The astronomer Pytheas cannot have been so far mistaken as to the latitude of these islands.
(4) That it was six days’ sail to Thule from Britain[55] will not suit Shetland, even if we make allowance for the frequently obscure statements as to the day’s journeys that are attributed to Pytheas (e.g., by Strabo).
(5) That Strabo in one place [ii. 114] calls Thule “the northernmost of the British Isles” cannot be used, as Müllenhoff uses it, as a proof of its belonging to these islands and having a Celtic population. There is not a word to this effect. To Strabo, who also placed Ierne (Ireland) out in the sea north of Britain, it must have been natural to call all the islands in that part of the world British. Indeed, he says himself in the same breath that Thule, according to Pytheas, lay within the Arctic Circle. How little weight he attached to the expression British is additionally apparent from another passage [ii. 75], where he says that “Hipparchus, relying on Pytheas, placed these inhabited regions [Shetland] farther north than Britain.”
(6) Pliny [Nat. Hist. iv. 104] mentions among islands north of Britain as “the greatest of all, ‘Berricen,’ which is the starting-place for Thule.” Berricen, which in some MSS. is written “Nerigon,” has been taken for Mainland of Shetland,[56] while others have seen in the form Nerigon the first appearance in literature of the name of Norway (“Noregr”),[57] though with doubtful justification, since this name was hardly in existence at that time. But whether the island be Shetland or Norway, this passage in Pliny puts Thule outside the Scottish islands. And the reference to that country makes it probable that the statements, in part at any rate, are derived from Pytheas.
(7) Finally, it may perhaps be pointed out that Thule is nowhere referred to as a group of islands; the name rather suggests the idea of a continuous land or a single island. To this it may be objected that neither is Orkan referred to as an archipelago in the oldest authorities; but it is uncertain whether in Pytheas, as in Diodorus, Orkan was not used of the northern point of Brettanice, and only later transferred to the islands lying to the north of this. Thule, on the other hand, always appears as a land far out in the ocean, and it is moreover uncertain whether Pytheas ever expressly described it as an island.
Thule is not Iceland
But if none of the statements about Thule answers to Shetland, it becomes a question where we are to look for this country.[58] The Irish monk Dicuil, who wrote about 825 A.D., regarded it as self-evident that Iceland, which had then been discovered by Irish monks, must be Thule, and called it so. After him Adam of Bremen and many others have looked upon Iceland as the Thule of the ancients. The objections to this hypothesis are: first, that Thule was inhabited (cf. Geminus, Strabo, and others, see [pp. 54-55]), while Iceland probably was not at that time. Even in Dicuil’s time only a few monks seem to have lived there (see below on the discovery of Iceland). Nor is it likely that Pytheas should have continued his voyage at haphazard across the ocean, unless he had heard that he would find land in that direction. To this must be added that Iceland lies so far away that the distance of six days’ sail will not suit it at all. Finally, if Pytheas had sailed northward at haphazard from Scotland or from Shetland, the least likely thing to happen was for him to be carried towards Iceland; neither the currents nor the prevailing winds bear in that direction; but, on the other hand, they would carry him towards Norway, and it would be natural for him to make the land there, perhaps just between 63° and 64° N. lat. or thereabouts.
Thule is Norway