sturian exiles must have brought with them as a memorial
of their far-off home.

F. 3.--Many such small but vivid touches of the past meet those who
visit the Wall. At "King Arthur's Well," for example, near Thirlwall,
the tiny chives growing in the crevices of the rock are presumably
descendants of those acclimatized there by Roman gastronomy. At
Borcovicus ("House-steads") the wheel-ruts still score the pavement;
at Cilurnum the hypocaust of the bath is still blackened with smoke,
and at various points the decay of Roman prestige is testified to by
the walling up of one half or the other in the wide double gates which
originally facilitated the sorties of the garrisons.

F. 4.--The same decay is probably the key to the problem of the
"Vallum," that standing crux to all archaeological students of the
Wall. Along the whole line this mysterious earthwork keeps company
with the Wall on the south, sometimes in close contact, sometimes
nearly a mile distant. It has been diversely explained as an earlier
British work, as put up by the Romans to cover the fatigue-parties
engaged in building the Wall, and as a later erection intended to
defend the garrison against attacks from the rear. Each of these views
has been keenly debated; the last having the support of the late Dr.
Bruce, the highest of all authorities on the mural antiquities. And
excavations, even the very latest, have produced results which are
claimed by each of the rival theories.[289]

F. 5.--Quite possibly all are in measure true. The "Vallum" as we now
see it is obviously meant for defence against a southern foe. But the
spade has given abundant evidence that the rampart has been altered,
and that, in many places at least, it at one time faced northwards.
Though not an entirely satisfactory solution of the problem, the
following sequence of events would seem, on the whole, best to explain
the phenomena with which we are confronted. Originally a British
earthwork[290] defending the Brigantes against the cattle-lifting
raids of their restless northern neighbours, the "Vallum" was
adapted[291] for like purposes by the Romans, and that more than once.
After being thus utilized, first, perhaps, by Agricola, and afterwards
by Hadrian (for the protection of his working-parties engaged in
quarrying stone for the outer fortifications), it became useless when
the Wall was finally completed,[292] and remained a mere unfortified
mound so long as the Roman power in Sout