8 
near Forden, the dyke passes nearly due north between the 
road to Montgomery and the abrupt boss of volcanic trap 
which looks at a short distance like a ruined castle, and 
which has been encircled by a very broad and deep moat. 
There can be no doubt but that this was a point of observa- 
tion, and as it is but some 20 yards on the English side 
of the dyke, it was most probably one of the positions 
permanently occupied by the English followers of the 
Mercian king. From this point the dyke gradually swerves 
to the east from the road between Montgomery and 
Buttington and makes directly over the low slopes of the 
hills, in some places being nearly ploughed down, and in 
others, and especially in the small valleys, being of con- 
siderable height and resembling a railway embankment, 
until it reaches the higher ground of Fron. Thence it runs 
through Pentre and gradually approaches the road, and 
finally dies away in the alluvium of the Severn, nearly a 
quarter of a mile to the south of Buttington Church. The 
commanding camp to the south of this portion of the line is 
Caer Digol, or the Beacon Ring, on the top of the Long 
Mountain. The morass, which in Offa’s time must have 
extended between the Main Ditch and the Severn, prevented 
the necessity of any bank being made between Buttington 
and the Cefn. Where, however, the open country demands 
a defence to the north of Cefn, an embankment makes 
sti'aight for the greenstone ridge of the Garreg, and is very 
plainly seen close to the farm-house of that name, near the 
Trewern Gate. Here we lost our clue, and it is very likely 
that the steep ridges of Moel y Golfa, and the marvellously 
strong camps of the Breiddan and Middleton Hills, formed 
a sufficiently strong barrier without any dyke being raised. 
We picked it up again, however, on the western or Welsh 
side of the Severn, from which it runs, as shown in the 
ordnance map, due north to the Four Crosses, where it joins 
the Oswestry road, and where it is cut across by the new 
