^{page 380 note a} It may perhaps be imagined that we might content ourselves with the maps which are given in the MSS. of Ptolemy, which are presumably reproductions of those drawn under the author's superintendence by a certain Agathodæmon. By taking these maps as the basis of our discussion, however, we should be liable to be misled by details introduced by the draughtsman, for which Ptolemy affords no authority. However valuable these maps may be as aids to the criticism of Ptolemy's text, they cannot be made a substitute for the text itself.

^{page 380 note b} A recent exemplification of this fallacious method is supplied by the Papers by Mr. Gordon Hills in the Journal of the British Archaeological Association for 1878 and 1881, which have gained a sort of authoritative rank far in excess of their real merits, through the deference paid to them in Prof. Rhys' work on Celtic Britain.

^{page 380 note c} The first copy I made of the map was drawn from the text of Ptolemy as given in the Monumenta Historica Britannica, and the conical projection was employed instead of the rectilinear one described above. There is, however, no very striking difference between the two delineations.

^{page 383 note a} The nearest approach to an anticipation seems to be the view of General Roy, that a portion of the map followed by Ptolemy had been accidentally torn off and joined again in the wrong place. This explanation, however, is both far less probable and far less adequate than the one given above.

^{page 383 note b} The use of this name by Ptolemy shows that our island is so called as being the greater of the British islands, and not, as has been sometimes supposed, in contradistinction to Brittany.

^{page 388 note a} The first portion of the word Parisi cannot be parth, but it may perhaps be some synonymous derivative from the same root. Compare the Parisii of Gaul.

^{page 391 note a} I was at one time inclined to identify the Alaunus with the Stour, on the ground that one of the tributaries of that river still bears the name of Allen. This, however, is one of the commonest of our river names, and there is no improbability in supposing that it may have belonged to both rivers. They are farther apart than are the Axe and the Exe.

^{page 392 note a} The ancient form of the word traeth would be tractos.

^{page 392 note b} This supposition enables us to account for the extraordinary fact that the town of Aberystwith, the name of which means “the mouth of the Ystwyth,” does not stand on this river at all, but on the north bank of the Rheidol, while the Ystwyth discharges itself a mile further to the south. If the coast-line formerly stood a few miles west of its present position, the Ystwyth would receive the Rheidol as a tributary. The town at the mouth of the Ystwyth, being gradually driven back by the encroachments of the sea, would retain its original name even when it had ceased to be appropriate.