A HISTORY OF DERBYSHIRE 



of more moment. It would represent their first definite enrolment in the 

 Empire, as men liable to taxation and to conscription. 1 It would imply 

 the systematic organization of this district in the Roman system. 



It remains to determine the Romano-British name of Brough. 

 Fortunately the task is easy. A milestone found at Buxton in 1862 

 states that that place was 10 (or perhaps 11) miles ANAVIONE 

 (p. 226). This might be interpreted either a Navione, 'from Navio,' or 

 Anavione ' (from) Anavio.' But the Foligno incription just cited shows 

 that the latter is preferable. If we proceed to inquire where Anavio was, 

 we shall find only one Roman site distant i o or 1 1 miles from Buxton 

 and connected with it by a road. Brough, therefore, is Anavio, and the 

 Anavionensian Britons are the hillmen of High Peak. With this result 

 we may connect two items in the Anonymus Ravennas. He mentions 

 a river Anava next to a river Dorvantium, and a place Navione * or 

 Nanione next to Aquis. If we take Dorvantium to be the Derwent, 

 Aquae to be Buxton, and Navione to be a mis-spelling of Anavione, 8 

 everything falls into its place. We may even go on to think that the 

 river at Anavio was called Anava, and that the name still survives in Noe. 



4. MELANDRA 



Melandra is a Roman fort in the parish of Glossop and the town- 

 ship of Gamesley, near to the Cheshire border and the Dinting station of 

 the Great Central Railway. Its position is no less significant than that 

 of Brough. Near Glossop, the gorge of Longdendale, meeting lesser 

 valleys, begins to widen towards the Cheshire and Lancashire lowlands. 

 Here is the easiest entrance to the north Derbyshire hills. Here, too, is 

 the nearest point to those hills that is readily accessible from the western 

 plains. The ancient soldier, wishing to plant a fort within striking 

 distance of High Peak and yet within safe reach of western communica- 

 tions, would find his fittest site near Glossop. 



The actual position selected for the fort agrees well with Roman 

 custom (fig. 1 6) . Not far from Glossop a low spur of hill runs out from 

 the south side of the valley into Longdendale, and on the end of this spur 

 stands the fort. The ground falls away from it on all sides except for 

 a narrow neck to the southward, which quickly changes to a rise. But 

 on the other three sides a slope of varying steepness descends a hundred 

 feet to the river Etherow and the Glossop Brook, which meet a little way 

 north of the fort, and to a little ravine which descends to the Glossop 

 Brook. It is a well-protected site, and it commands a wide outlook 

 over Longdendale and the great hills around it. And at the same time 



1 Corf. Insc. Lat. xi. 52i3 = Dessau 1338. Watkin, who first saw that this inscription referred to 

 our district, misinterpreted it both by taking censitor as the officer of a cohort, and by eliciting a cohort 

 with the impossible title of Brittonet a Navione (Archteologjcal Journal, xli. 255 : Derb. Arch.Journ. vii. 83). 



8 Parthey and Finder gave Nanione as the reading of the best MSS. But Prof. J. S. Phillimore, 

 who examined the chief MS. for me, the Vatican Urbinas 961, tells me that it reads Nauione. In 

 either case the divergence from Anavione is slight. 



8 Omissions of initial letters are not the least common among the many errors of the Ravenna lists. 



2IO 



