Reposted from: http://penelope.uchicago.edu/Thayer/E/Gazetteer/Places/Europe/Great_Britain/_Periods/Roman/home.html

Thayer's Note: Stevenson's "translation" of Ptolemy, to which this page belongs, is abysmally bad. It should not be used for any serious purpose. For details and correctives, see  Ptolemy homepage.

Penelope's notes on the map: No information has been added to Ptolemy's text as I have it, so there is almost no topographic data; the courses of the rivers also remain unmapped. The interesting item is of course one of Ptolemy's biggest mistakes: Scotland is rotated clockwise by about 90°. The error occurs at a very specific place, which I've marked by a red line on the map, extending from the Solway Firth to Newcastle. No coincidence at all that this is almost exactly the line of Hadrian's Wall: all of a sudden the Roman geographer loses all his land data and has to rely on pilot's accounts, with choppy seas and strong currents accounting for the bad data. Ptolemy may well have had only a single report for the navigation of the top of Scotland; what else could he do? (For an alternative and vaguer explanation — I'm not in the least convinced by my own — see John Ward's The Roman Era in Britain, p14; for yet another explanation, far more convincing and detailed, see Thomas G. Ikins' Roman Map of Britain. There are, out there, many other diagnoses of this error by Ptolemy and of his other errors, limited only by human ingenuity expressed in shelves full of books and a passel of webpages. None of them is to be trusted, probably.) This map, when examined together with Ptolemy's maps of Belgica and Cimbria (now the Low Countries and Denmark), reveals something important about the sources of error in the Geography. In Book II, Chapter 8 we see the Dutch coast correctly placed; navigation thence along the coast ought to have showed Denmark to be, in terms of absolute coördinates within his own system, where Ptolemy puts northern Scotland, and the positions on the map of Britain would have been provided with cross-checks. What happened instead is that in Book II, Chapter 10 we do indeed see Denmark: similarly tilted by a same large clockwise angle. I haven't yet figured them exactly, but Ptolemy's distances from Scotland to Denmark appear to be roughly the same as ours. He did his cross-checks all right, but both Scotland and Denmark were beyond the pale of Roman dominion. He is relying on sea data. To me this suggests that Ptolemy kept two sets of data, obtained by land and by sea; that he cross-checked each internally; that he favored the land data because he knew they were more accurate; and that for the sea data he knew they weren't good, but at least he made them consistent. But once again, don't believe anything you read about Ptolemy, not even my own stuff — it's all theories and smoke.