Book II, Chapter 3

No information has been added to Ptolemy's text as I have it, so there is almost no topographic data. The color-coding of town names follows the tribes as given in the text: notice especially the Phoenicians, typically only occupying coastal towns.

The courses of most rivers remain unmapped, but for the Ana Ptolemy gives two intermediate points, and a bend is implied near Canaca; and the case of the Baetis, where he gives the mouth and "the sources", is a demonstration of one of the geographer's failures: he almost never tells us what rivers his towns are on, even when he maps both. The result here is that the best course one can draw for the Baetis is a straight line from the sources to the mouth. As it happens, Corduba actually lies on the river, but the other places that appear to above do not; and some places actually on it don't appear to be. It would have been an easy matter to have told us for example that Hispalis was on the Baetis.

Something else. If you are observant, you noticed a bare patch at the S tip of Spain: no towns, nothing. Ptolemy is right, there's nothing there, and nothing should be: much of it is in fact open sea. The reason is clear, and, to me at least, fascinating.

From Baelon (the ruins of which are about 2 km E of today's fishing town of Camarinal) to Calpe (the rock of Gibraltar) is actually about 20 minutes of longitude — not 1°15. The difference is clearly due to pilot data thru the Straits, where even today strong currents, choppy seas and high traffic decrease ship speed: the coastline thus appears longer. If we replot Carteia and Calpe 20 minutes E of Baelon and redraw the coast from there, presto! the bare patch disappears. In fact, I had to pull Barla in as well. The bare patch would disappear completely if we pulled the coast to Malaca west to its proper place relative to Calpe, although that distance is much more nearly correct.