64 



Some jfeaturcs of IRoinan jforts in Britain. 



The excavation, during the years 1894 — 8, of several forts 

 on the Wall of Hadrian (one result of whicli lias been Mr. 

 Bosanquet's admirable plan of Borcovicium), the comple- 

 tion in 1901 of the work at Gellygaer, and the interesting 

 investigations now in progress on the Wall of Antonine 

 under the auspices of the Society of Antiquaries of Scot- 

 land, have turned the attention of archseologists during the 

 last few years to the subject of the particular form of 

 defence known as the castellum, which seems to have been 

 used by the Romans for the purpose of watching the tribes 

 of the hill country, or holding the lines of fortifications 

 that marked for the time being the limits of the empire. 



Manchester, as it happens, is not unfavourably situated 

 for this particular study. There may still be seen in the 

 neighbourhood of Knott MilP the remains of the fort 

 which has given its name to the city, and which a writer 

 who visited Manchester about 1540 described - as " almost 

 ii. flyte shottes without the towne." The plan of Man- 

 cunium is now lost beyond recovery, but about twelve 

 miles to the east lay the sister fort now known as 

 Melandra, which is shown by the inscriptions^ on four 



1. Roeder : Roman Manchester, p. 11. Watkin : Homan Lancashire, 

 p. 104. An excellent specimen of the core of one of the walls is pre- 

 served in situ under one of the Railway arches. 



2. Hearne's Leiand, vol. v., p. 94 (edit. 1769-70). 



3. C.I.L., vii., Nos. 178, 213, 214. A fourth is figured m Mem. Lit. 

 Phil. Soc. Manch., vol. v., plate vii.,opp. p. 534, which does not appear 

 in the Corp. Ins. Lat., vol. vii. The explanation seems to be that the 

 Editor of the Corpus, as he states on p. 56, only consulted these memoirs 

 as far back as 1805. Vol. v. is dated several years earlier. The pattern 

 of the border on this stone is similar to that of the Melandra stone. 



