US ON THE •' BELGIC DITCHES, former one.^ It "^as a ditcli with tico mounds, and these gradually became lower as I traced it to the eastward, a mile or two beyond Tilshead. If this ditch be a continuation of the former one, I cannot satisfactorily account for its change of character. I could find no remains of this Belgic boundary — if we may venture to give it such a title — north of Beacon HilL Even " the unmutilated remains of a bank and a ditch/' on AVick-down, turned out to be merely a deep ditch with a low mound on each side of it. But south of the hill, the Ames- burj bounds presented appearances which strongl}^ resembled those of an ancient earth-work, and we may be allowed to conjecture that they were once connected with the " Devils Ditch," east of Andover, and ^ith the boundary-line, a frao-ment of which still remains to the south of ^alburv. O t. According to these speculations, the second Belgic boundary must have included the valleys of South Wiltshire, and then have swept round, so as to separate the downs of Hampshire from the woodlands which encircled Scots Poor. The hypo- thesis does not seem an unreasonable one, and I know of no other which can satisfactorily account either for the boundary- line to the north of Heytesbury, or for the lines which are found in the neighbourhood of Walbur}'^ and Andover." It will be seen that the writer differs from Stukeley in considerino- the first and second of his ditches as forming parts of one continuous boundary ; and in denying alto- gether to the ditch which runs immediately north of Old Sarum, the character of a Belgic earthwork.^ Were this last ^ When these mounds approach the whenever they speak of the Belgic Pro- " Lonfr Barrow," which lies about a mile vince, treat it as a whole, from Tilshead, they turn at right angles, It may be observed, that there are and after having half enclosed the mound, some ditches near Chisborough, which pui-sue their former course. Our best have not beer inserted in the map. There chance of explaining anomalies like these, can be little doubt that four distinct lines would be a really critical edition of the of boundary passed near that fortress, " Gromatici veteres." and to have noticed the remains of all these ' It may, perhaps, be said, that the boundaries would have answered no good lines near Walbury and Andover might purpose, and would have made the plan have been the boundaries of a Belgic much too complicated, settlement, whose capital was Winchester; ^ The period at which, and the purpose and which was miited to its western for which, this earth- work was constructed, neighbour before British geography was were sufficiently discussed at Salisbury, known to the Romans. But there is Those who feel an interest in the matter reason to believe that the State of the may see what are the writer's views on Southern Belgoe was not merely a poli- this subject, by consulting the paper be tical, but an ethnographical unity. The has written for the Sahsbury Volume on other Belgic districts, though politically the " Early English .Settlements ia South united, are always spoken of as peopled by Britain." different races ; but the classical writei-s,
