As for the roads they all radiated from London, as indeed they do still in large measure. One passed direct from London to Colchester and thence, viâ Stratford St. Mary and Long Stratton and Scole, to Caistor-by-Norwich. Such names as Stratford and Stratton, unless shown to be of modern origin, are strong evidence of Roman occupation, and at Scole, where the road crosses the Waveney and enters Norfolk, have been found some Roman remains and, perhaps, traces of a paved ford. That is the road on which we are now travelling. Caistor-by-Norwich, where we should not have seen much if we had halted--that is the worst of these Roman remains--is in all human probability _Venta Icenorum_, concerning the situation of which debate used to be carried on vehemently. What we might have seen is a rectangular enclosure of earthen mounds covering massive walls, having bonding tiles and flint facing to a concrete core, the walls themselves being visible on the north and west, and a great fosse surrounding the whole. Its area is about thirty-four acres, and there were towers at each corner. A careful analysis of the evidence leads to the sure conclusion that this was a small country town and not a great military fortress. This particular road crossed the Ipswich river a few miles to the north-west of Ipswich, and a branch from it ran by way of Goodenham to Peasenhall. Thence it can be traced due east to Yoxford, where it ends, so far as our certain knowledge goes. From Peasenhall another direct road can be traced as far as the Waveney, near Weybread, and no further. Other roads there are of uncertain Roman origin, but the most important of them was the Peddar's (or Pedlar's) Way, which can be traced with certainty from Barningham, in Suffolk, to Fring, about seven miles from Brancaster, and perhaps even to Holme, which is nearer, and is, indeed, one of the supports for the theory concerning the nature and origin of Brancaster, but the modern roads seldom follow its course. A Roman road was supposed to run from Caistor viâ Downham Market, and across the Cambridgeshire Fens to Peterborough, but its existence is hardly proved in Norfolk, and its origin is hardly clear to demonstration in Cambridgeshire. These are all the Roman roads which need concern us, and the references to Roman roads in guide-books and on Ordnance and other maps may be disregarded. This is written not at all by way of disparaging the ordinary guide-books, some of which are monuments of learning and industry, and by no means in any mood of conscious superiority. There is no credit at all in knowing that which Mr. Haverfield has made easy, and, until he co-ordinated the facts and sifted the evidence, it was practically impossible for anybody but a specialist to know the truth. He is a specialist of the true scientific temperament, eager to acquire knowledge, cautious in inference, and it is to be feared that he and his like knock a good deal of romance out of travel in England.