The quantitative distinction is definitely recognisable. The cantons, at least as they present themselves among the Celts and the Germans, are throughout tribes more than townships; this very essential element was peculiar to all Celtic territories, and was often covered over rather than obliterated even by the subsequent Romanising. Mediolanum and Brixia were indebted for their wide bounds and their lasting power essentially to the fact that they were, properly speaking, nothing but the cantons of the Insubres and the Cenomani. The facts, that the territory of the town of Vienna embraced Dauphiné and Western Savoy, and that the equally old and almost equally considerable townships of Cularo (Grenoble) and Genava (Geneva) were down to late imperial times in point of law villages of the colony of Vienna, are likewise to be explained from the circumstance that this was the later name of the tribe of the Allobroges. In most of the Celtic cantons one township so thoroughly preponderates that it is one and the same thing whether we name the Remi or Durocortorum, the Bituriges or Burdigala; but the converse also occurs, as _e.g._ among the Vocontii Vasio (Vaison) and Lucus, among the Carnutes Autricum (Chartres) and Cenabum (Orleans) balance each other; and it is more than questionable whether the privileges which, according to Italic and Greek organisation, attached as a matter of course to the ring-wall in contrast to the open field, stood _de jure_, or even merely _de facto_, on a similar footing among the Celts. The counterpart to this canton in the Graeco-Italic system was much less the town than the tribe; we have to liken the Carnutes to the Boeotians, Autricum and Cenabum to Tanagra and Thespiae. The specialty of the position of the Celts under the Roman rule as compared with other nations--the Iberians, for example, and the Hellenes--turns on this, that these larger unions continued to subsist as communities in the former case, while in the latter those constituent elements, of which they were composed, formed the communities. Older diversities of national development belonging to the pre-Roman epoch may have co-operated in the matter; it may possibly have been more easily practicable to take from the Boeotians the joint diet of their towns than to break up the Helvetii into three or four districts; political unions maintain their ground even after subjugation under a central power, in cases where their dissolution would bring about disorganisation. Yet what was done in Gaul by Augustus or, if it be preferred, by Caesar, was brought about not by the force of circumstances, but chiefly by the free resolution of the government, as it alone was in keeping with the forbearance otherwise exercised towards the Celts. For there was, in fact, in the pre-Roman time and even at the time of Caesar's conquest a far greater number of cantons than we find later; in particular, it is remarkable that the