mpire.

SECTION F.

Severus completes Hadrian's Wall--Mile
Castles--Stations--Garrison--Vallum--Rival
theories--Evidence--Remains--Coins--Altars--Mithraism--Inscription to
Julia Domna--"Written Rock" on Gelt--Cilurnum aqueduct.

F. 1.--It is to Severus, therefore, that we owe the final development
of this magnificent rampart, the mere remains of which are impressive
so far beyond all that description or drawing can tell. Only those
who have stood upon the heights by Peel Crag and seen the long line of
fortification crowning ridge after ridge in endless succession as
far as the eye can reach, can realize the sense of the vastness and
majesty of Roman Imperialism thus borne in upon the mind. And if this
is so now that the Wall is a ruin scarcely four feet high, and, but
for its greater breadth, indistinguishable from the ordinary local
field-walls, what must it have been when its solid masonry rose to
a height of over twenty feet; with its twenty-three strong
fortresses[286] for the permanent quarters of the garrison, its
great gate-towers[287] at every mile for the accommodation of the
detachments on duty, and its series of watch-turrets which, at every
three or four hundred yards, placed sentinels within sight and call of
each other along the whole line from sea to sea?

F. 2.--Of all this swarming life no trace now remains. So entirely did
it cease to be that the very names of the stations have left no shadow
of memories on their sites. Luguvallum at the one end, and Pons
Aelii at the other, have revived into importance as Carlisle and
Newcastle,[288] but of the rest few indeed remain save as solitary
ruins on the bare Northumbrian fells tenanted only by the flock and
the curlew. But this very solitude in which their names have perished
has preserved to us the means of recovering them. Thanks to it there
is no part of Britain so rich in Roman remains and Roman inscriptions.
At no fewer than twelve of these "stations" such have been already
found relating to troops whom we know from the 'Notitia' to have been
quartered at given spots _per lineam valli_. A Dacian cohort (for
example) has thus left its mark at Birdoswald, and an Asturian
at Chesters, thereby stamping these sites as respectively the
_Amboglanna_ and _Cilurnum_, whose Dacian and Asturian garrisons the
'Notitia' records. The old walls of Cilurnum, moreover, are still
clothed with a pretty little Pyrenaean creeper, _Erinus Hispanicus_,
which these A