50 Transactions. 



work, such as the native races in the East invariably employ in 

 theii' fortifications. But however we suppose the vallum to 

 have been furnished with such means of defence, we cannot in 

 any way account for the existence of the southern agg-er, and 

 especially of its considerable distance from the fosse. 



We are told, however, that the difficulty is just the same if 

 we suppose the vallum to have been erected as a defence against 

 the north. How, it is said, are we then to account for the exist- 

 ence of the north agger ? Mr Neilson feels this difficulty, and is 

 so much impressed by it that he adopts the theory, not at all 

 original, that the vallum was at first intended as a temporary 

 defence against the north. But after the wall was built — which 

 he supposes must have been very shortly after the formation of 

 the vallum — perhaps in ten years' time, or thereabouts — the earth- 

 works were turned into a defence against the tribes to the south, 

 who may have shown themselves in the interim disposed to be 

 troublesome. To render it fit for that purpose, he supposes the 

 north agger to have been then thrown up ; and points to the worn 

 condition of the south agger as suggesting that though it was 

 permitted to remain, it was not kept up, and so has become more 

 nearly obliterated than the rest of the work. 



That is an ingenious theory, because it seems to get rid 

 of the more considerable difficulties. Still, it has some weak 

 ]3oints. It assumes the necessity of guarding the Roman position 

 from the southern Britons all through the time of the Eoman 

 occupation — during most of which, as we know, the country to 

 the south of the wall was so thoroughly Eomanised that it 

 possessed towns and villages, and to all appearance a peaceable 

 and dependent people. It also involves the theory that the 

 Romans, though they took pains to turn the vallum into a rear 

 defence, did not show the usual thoroughness of their work by 

 destroying the south agger, which had become worse than useless, 

 or even of using the earth, as they might have done, to construct 

 the north agger. 



If now we consider the other theory — that the vallum was 

 never anything else than a mere primitive defence against the 

 north, erected in earlier times than the stone wall, and allowed ta 

 remain, simply because it was there and did no harm, we find 

 many points in our favour. It seems to me that Mr Neilson, by 

 his minute and ingenious investigation of the marginal mound, 



