years to accumulate. Of the name the spot bore in military records, we are entirely ignorant. The all too fragmentary inscriptions which came to light furnished no clue,. And yet those marching northwards along the great road must always have looked to the triple Eildon as a landmark, and it may well be that they knew the fort as Trimontium. The name occurs in Ptolemy's Geography[1] as that of one of the four towns of the Selgovae situated in the south of Scotland. The form in which it appears in the original text (Τριμοντιον) is merely a Greek transliteration of a Latin word meaning 'the place of the three hills,'—precisely such a designation as would naturally have been applied, especially by those approaching from the south, to a settlement lying under the shadow of the Eildons.

There are several obvious analogies. 'Tripontium' mentioned in the Antonine Itinerary,[2] where it is placed in the territory of the Coritani, can only signify 'the place of the three bridges.'[3] Again, the Brittones 'Triputienses,' who had been transplanted to Germany, appear to have received their distinguishing appellation because their new home was near some 'place of the three wells,' in the neighbourhood of the modern Vielbrunn,[4] in the, country between the Rhine and the Main. Lastly, in the reign of Antoninus Pius, the town of Philippopolis in Thrace actually bore the name of Trimontium, and on a unique specimen of one of the coins of the city, preserved in the Fitzwilliam Museum, Cambridge, the river-gods are represented with three mountain peaks rising behind them. No doubt the country of the Selgovae is generally believed to have lain further to the west, and for this reason Horsley and others have been induced to locate the Trimontium of Ptolemy near Birrenswark in Annandale.[5] Roy, however, as we saw above, preferred to look for it at the foot of the Eildons, this too in express defiance of his 'authority,' the spurious itinerary of Richard of Cirencester.[6] And it is safe to say that neither Birrenswark nor any other group of hills in the south of Scotland suits nearly so well.

1 Bk. ii. c. iii.

2 P. 477.

3 Haverfield, Victoria County History, vol. i. p. 231.

4 E. Fabricius, Ein Limesproblem, p. 17.

5 Horsley, Britannia Romana, p. 377; Maitland, Hist. of Scotland, i. p. 142; Chalmers, Caledonia, ed. 1887, vol. i. pp. 120f.

6 Roy, The Military Antiquities, p. 116.