That Thule is often referred to as an island by later authors is of little weight. In the first place we do not know whether Pytheas himself so described it; according to all the geographical ideas of the ancients about the north a land in the ocean farther north than the British Isles must necessarily have been an island, even if Pytheas did not say so. In the next place, if a traveller sails northwards, as he did, from one island to another, and then steers a course over the sea from Shetland and arrives at a country still farther north, it would be unlikely that he should believe himself back again on the continent. Besides, Pytheas made another voyage eastwards along the north coast of Germany, past the mouth of the Elbe, and then he had the sea always to the north of him in the direction of his Thule. In order to discover that this land was connected with the continent, he would have had to sail right up into the Gulf of Bothnia. It would therefore have been illogical of Pytheas if he had not conceived Thule as a great island, as in fact it was spoken of later. It is mentioned indeed as the greatest of all islands. When the Romans first heard of Sweden or Scandinavia (Skåne) in the Baltic, they likewise called it an island, and so it was long thought to be.
According to what has been advanced above we must then believe that Pytheas had already received information in northern Brettanice or in the Scottish islands about Thule or Norway across the sea. But from this it follows that in his time, or more than a thousand years before the beginning of the Viking age, there must have been communication by sea between North Britain and Norway. It may seem that this is putting back the Norsemen's navigation of the high seas to a very remote period; but as we shall see in a later chapter on the voyages of the Norsemen, there are good reasons for thinking that their seafaring is of very ancient date.
Pytheas may have sailed from Shetland with a south-westerly wind and a favourable current towards the north-east, and have arrived off the coast of Norway in the Romsdal or Nordmöre district, where the longest day of the year was of twenty-one hours, and where there is a free outlook over the sea to the north, so that the barbarians may well have shown him where the sun went to rest. From here he may then have sailed northwards along the coast of Helgeland, perhaps far enough to enable him to see the midnight sun, somewhere north of Dönna or Bodö; this depends upon how early in the summer he reached there. On midsummer night he would have been able to see a little of the midnight sun even at about 65-1/2° N. lat.; or south of Vega.[61]