THE SIGNIFICANCE OF IS AND UWCH IN WELSH COMMOTE AND CANTREF NAMES A very marked feature of the names for medieval Welsh units of administration like commote and cantref is their frequent division into is and uwch (med. Welsh uch). The line of demarcation is always a natural phenomenon, most often a river, but also sometimes a mountain or forest. The exact significance of is and uwch in this context has not, as far as I know, been adequately discussed. The usual meaning of the two forms is, of course, 'lower' and 'higher', the comparative degrees of the adjectives isel 'low' and uchel 'high'. As prepositions they mean 'below' and 'above'. But why above and below? The most obvious interpretation is a purely geographical below? The most obvious is a purely one, and I originally thought in terms of the courses of the rivers fciuers which provided the lines of division. Is an uwch might conceivably no h refer to the two banks, left and right. But no such correlation could bonus be established. Is and uwch are used indiscriminately of the lands on the left and right banks. And in any case, they are not applied to rivers solely. A second possibility was orientation, i.e. whether is and uwch corresponded to points of the compass, north, east, not south, or west, as the case might be. But again there were no signs compass of correlation. Recently, however, a third explanation has presented itself, o The division of Gaul into Cisalpine (Lombardy) and Transalpine (Gallia Cisalpina and Gallia Transalpina) provided the clue. Cis means 'on this side of, as does its companion citra. Trans is 'the I u^c^ other side of. Cisalpine Gaul was certainly so named from the -noun point of view of Rome, the central authority, and means 'Gaul on our side of the Alps'. Why then could not is and uwch mean some- thing similar: is = cis, citra, and uwch = trans, ultra? Here, I think, we have been misled by the use of Latin sub, subtus, and supra for some of the medieval Welsh divisions into is and uwch. They are mistranslations, and it would have been much more accurate and better from our point of view if the terms cis, citra, trans, ultra had been used, and as we shall see, this was actually the case in a few instances. The working hypothesis that I wish to suggest is quite simply this: that the terms is and uwch in this context are not purely geographical, but are rather historico-geographical, and that they