(Newstead) Roxburghshire, Scotland.
The Roman fort, which took its name from the
triple peaks of the nearby Eildon Hills, lies on the E
side of the village of Newstead, guarding the point at
which the Roman road from York to the Forth crossed
Excavations in 1905-10 and 1947 disclosed that there
were four superimposed forts. The earliest, built by
Agricola ca. A.D. 80, covered 4.14 ha and was defended
by a clay rampart, 6.9 m thick, and double ditches. Its
outline was unusual, the rampart alignment being staggered at each of the four gates. Only one internal building, a timber-framed stable, is known, but it is possible that the garrison comprised two alae quingenariae. This
fort was replaced in the late 80s by another, 5.7 ha in extent. The new rampart was 13.5 m thick and the single
ditch over 4.8 m wide; the scale of the defenses reflect
the fact that Trimontium was now the most important
Roman military position N of the Cheviots. The barracks
were of wattle-and-daub set on stone sill-walls, and their
size suggests that they were designed for legionaries.
About A.D. 100 this phase terminated in disaster, and an
interval of some 40 years elapsed before the third fort
was built during the Antonine reoccupation of the Lowlands. Enclosing an area of 5.9 ha, the early Antonine rampart was faced with a stone wall, in front of which there were two ditches.
The buildings of this period are known in some detail.
The praetentura contained 12 barracks, while the main
buildings comprised a principia; two horrea, one on
either side of the principia; and a praetorium with private bathing establishment. The retentura was cut off
from the rest of the fort by a dividing wall with a single
central gateway flanked by towers: this feature, unique
in Britain, is best explained by the fact that the garrison
was a dual one, a legionary vexillation occupying the
praetentura, while the retentura housed a regiment of
auxiliary cavalry. The fourth fort, of late Antonine
date, succeeded the third after a brief interruption. Although the existing curtain wall was retained, the rampart backing was increased in width and a third ditch added. Inside the fort the dividing wall was demolished
and new barracks and stables erected for an ala milliaria.
The date of the final abandonment of the site is uncertain, but occupation may have extended into the 3d c.
Outside the fort were annexes of different periods on
all sides except the N; the W annex contained a courtyard building, perhaps a mansio, and a bath house. A
number of temporary camps and a small roadside post
of Flavian date have also been found in the area while
a signal station, of which traces are still visible, has been
identified at the N end of the Eildons. The excavations
of 1905-10 produced rich finds, now in the National
Museum of Antiquities of Scotland. Most of these came
from pits in the S annex. Especially noteworthy are the
arms and armor of both legionaries and auxiliary troops,
tools and other items of daily use, and large quantities
J. Curle, A Roman Frontier Post and
(1911); Proc. Soc. Ant. Scotland
1-38; 86 (1952) 202-5; Inventory of the Ancient & Historical Monuments of Roxburghshire
2 (1956) 312-20; Britannia
3 (1972) 53-54.
K. A. STEER