whole account implies that the invaders came much further into the province
than the Vallum of Pius, and I imagine he meant the Wall of Hadrian.
Once lost, the ground was never regained. Septimius Severus has left no
traces but two or three denarii—perhaps due to accidental intercourse—at
Cramond. In the third century, under Alexander Severus and Gordian III.,
the Eoman frontier was the southern wall, with the outlying forts of High
Rochester and Risingham in Northumberland, Bewcastle and Netherby in
Cumberland, and these were lost about or after 250 A.D. Only in the
beginning of the fourth century is there some trace of a fresh advance. As
I have said, the history of Birrens after the time of Marcus is a blank, and
the blank is doubtless significant. The number and variety of its lapidary
monuments suggest that it may have been retained later than the northern
positions, but it seems to have been lost early in the third century at the
latest. In the fourth century there are again traces of Romans in it,
glass of Constantinian style, a few fragments of architectural work,1 an
aureus of Chlorus. Hadrian's Wall was unquestionably still occupied at this
time—witness the milestones on Stanegate—and the occupation may have
been extended to Birrens.
These conclusions are perhaps satisfactory so far as they extend. But the
careful critic will doubtless notice that I have used words like " perhaps "
and " probably " very freely, and have hedged here and there with a caution
supposed to be uncommon in a southern writer. The cause is obvious.
Much has lately been done to increase our knowledge of Roman Scotland—
at Birrens, at Ardoch, in the murus caespiticius, at Cappuck—but more
remains. No fort along the Vallum of Pius has yet been adequately
explored, and until that is done we cannot speak positively of many things.
Enough is perhaps known for me to venture on the preceding paragraphs.
I should be rejoiced if it soon became possible for Dr. Macdonald or some
other competent scholar presently to venture on much more.
List of Coins.
The following list is the result of an endeavour to catalogue the discoveries
of Roman coins in Scotland, so far as the descriptions of these coins and
the places of discovery have been adequately recorded. In the compilation
I received very generous help from Dr. James Macdonald, who put at my
disposal a list drawn up by himself, and aided me thus to approach nearer
to completeness than I could have done single-handed. I hope that, thanks
to his help, the result may be full and accurate enough to form a basis for
historical argument. I have made four lists, first and second the coins found
in the occupied places, that is, in or immediately near (i.) the forts along the
vallum and (ii.) other ascertained Roman forts in Scotland ; (iii.) the hoards,
1 See Macdonald and Barbour's Birrens and its Antiquities (Dumfries, 1897),
Plate III. (A), figs. 21, 22, 23. These pieces have not received the attention they