ROMAN BROUGH : = ANAVIO. 185 



was really the name of the station. This is Mr. Haverfield's 

 opinion. Mr. Thompson Watkin was almost right in his 

 plausible suggestion of Navio, but the stone found at Foligno, 

 in Italy,* with its reference to Brittonum Anavion(cmium), 

 leaves no doubt as to the correct reading, t 



II.— FIRST OUTLINE OF THE FORT. 



The excavations made for this Society during three weeks of 

 August, 1903, were of an exploratory character, designed to 

 determine the area which it was desirable to excavate and to 

 answer other preliminary enquiries. By following the superficial 

 indications, masonry was soon found, both the stone wall of the 

 large building in the centre (Plate II., 1) and the stout founda- 

 tions to the rampart turning the western corner (Plate II., 2). 



The work was recognizably Roman, apart from the small 

 objects — broken pottery, small coins, and the like — which were 

 found in the digging. The masonry was not of the solid charac- 

 ter familiar in the greater engineering works of the Romans, 

 but there were present, nevertheless, those characteristics, both 

 in general design and in some details, which are known in other 

 works of the second or third century. The facing stones, for 

 example, were of the usual pattern, wedge-like, with the narrower 

 end built into the wall. In the case of the outer wall, six feet 

 thick, the middle part in the thickness of the wall between the 

 outer and inner faces was found filled with boulders and rough 

 stones. The faces themselves, when they could be traced, were 

 well aligned, the stones being hammer-dressed with good surface. 



These explorations being only preliminary the wall was not 

 followed all around on this occasion, but was picked up at 

 intervals. The same might be said of the interior building. 

 The tentative plan (Plate III.) illustrates the result, showing by a 

 scored line the portions of walls which were actually traced, 

 and by a dotted line the positions which, by analogy, they may 

 be expected to occupy. In some cases, for which there is no 



* Ephemeris Epigraphica vii., 1102, and vol. vii., p. 84, of \\\\s Journal. 

 t See Mr. Haverfield's note at the end of this article. 



