[23] For the literature on these see Gardthausen, _Augustus_, II, p. 763 f. Nothing definite is known about them. If very significant for the “conquest” of Germany, why was their construction deferred to the period of Domitius, years after Drusus and Tiberius had been engaged in carrying on the most extensive campaigns? The very fact that these early incursions into Germany had been repeatedly made without the erection of any elaborate network of solidly constructed roads, is the clearest evidence that no permanent occupation of the country was intended. For the purposes of the occasional demonstration mere “war-paths”, supplemented here and there with some light, temporary construction were entirely adequate. It is a striking fact that of permanent road construction not a trace has been found in Germany, not in the lower Lippe valley, where, if anywhere, the highways of armies must have been solidly constructed if Germany was to be held as a province, nor even before the very gates of the camp at Haltern (cf. Koepp, _Die Römer in Deutschland_, 2nd ed., p. 136). Yet along the _limes_ roads were regularly constructed, and were an essential part of the system of defense. Tiberius seems to have begun a _limes_ in the _silva Caesia_, but not to have completed it (Tacitus, _Ann._, I, 50: “limitemque a Tiberio coeptum”). It was obviously a slight undertaking.
[24] Tacitus, _Ann._, II, 7. These were probably roads (Delbrück, _Gesch. d. Kriegskunst_, 2nd ed., p. 128 ff.). The use of “novis” indicates that such structures had been erected earlier. Their flimsy nature is to be inferred from the fact that the work had to be repeated in a few years, and the construction of Germanicus was doubtless no more lasting (see the preceding note).
[25] As for example Agrippa’s system of roads for Gaul. Yet Gaul needed them far less than Germany, for it was a relatively civilized country with means of rapid communication. Caesar seems to have been embarrassed but little in his campaigns by poor roads, in sharp contrast with the conditions prevailing in Germany.
[26] Dio (56, 18, 2) states that “their (i. e. Roman) soldiers were beginning to winter there and were founding cities”, but just what these “cities” were, he neglects to say, and they appear nowhere else either in his narrative, or in that of any other ancient writer; yet the destruction of such incipient “cities” after the defeat of Varus is just the sort of event that could not possibly have been passed over in silence by all our sources. When Dio comes to the appropriate section in his later narrative (22, 2ᵃ = Zonaras) where these should be mentioned, he speaks of nothing but “forts” (ἐρύματα). It is perfectly clear that his sources knew nothing about real “cities”, and that from his knowledge of the way in which settlements grow up about any army post however small, he is indulging in a little exaggeration in telling of the foundation of “cities” so as to give the desired background for his picture of a complete reversal of conditions in Germany.