The HIGHLANDS of SCOTLAND. 27 



precife period, we may refer the foundation of many of the 

 towns in the weft of England, which are known to have had a 

 Roman origin, as Lancafter, Manchefter, Warrington, Ribche- 

 fter, Overborough, Colne, &c. *. 



At this time, therefore, A. D. 79, the Britons of the north- 

 weftern parts of England, had acquired a considerable know- 

 ledge of regular architecture. But all to the north of the Ro- 

 man conquefts, we muft prefume was in its original flate of 

 barbarifm. Improvement, however, muft have kept pace with 

 the advances of the Romans into the country ; and it is there- 

 fore not difficult to mark its progrefs. In the year 80, we find 

 Agricola employed in erecting a chain of forts between the 

 friths of Clyde and Forth ; and in 83 f, the laft year of his 

 command, he had penetrated to the foot of the Grampian 

 mountains in the northern parts of Angus. From this time, 

 during the remainder of the reign of Domitian, and through 

 the whole of the reigns of Nerva and of Trajan, a period 

 of above thirty years, the Romans made no progrefs in the 

 ifland. The northern parts of the province were ill defended, 

 and the Caledonians, in that interval, recovered all that part of 

 Scotland which Agricola had gained ; for, in the fecond year 

 of Hadrian, A. D. 120, when that Emperor built his vallum 

 acrofs the ifland, between Solway frith and the mouth of the 

 Tyne, he confidered the Roman Province as extending no fur- 

 ther to the north than that rampart. " Murum per octoginta 

 " millia pafluum primus duxit qui barbaros Romanofque divi- 

 " deret." Vit. Hadr. Hijl. Aug. Script. 



This interval, therefore, of more than thirty years, muft 

 have been a period of remarkable improvement to the favage 

 Caledonians. Maintaining a conftant intercourfe with the Ro- 

 mans, not diftinguifhed by extraordinary hoftilities, and gradu- 



D 2 ally 



* Whitaker's Manchefter, book I. chap. 7. 



f Or 84 ; for the year is not certain. See Horsley, p. 48. 



