muir, at Tinnis, Lour, by the Tweed opposite Neidpath, and other places; but it would be somewhat rash to assume that the markings are by the hand of man.1 Up till very recently the peculiar structure known as the Broch was hardly known to exist in Scotland, unless to the north of the Caledonian valley, and in the northern and western isles. In those areas the number of brochs was estimated at 370. But south of the Caledonian valley, only three structures of this kind were supposed to exist, one of these on Cockburn Law in Berwickshire. Since then the camp or fort at Torwoodlee, in Selkirkshire, has been found to consist partly of a broch, and thus a connection has been established between the civilisation of the north and south of Scotland, for there can be no doubt that the occupants of these brochs were workers in bronze and iron. They used bone, horn, and stone, made pottery, had even Samian ware. The women spun and weaved, and the people lived on beef, venison, pork, veal, mutton, lamb, and fish. The broch has been defined "as a hollow circular tower of dry-built masonry, from 40 to 70 feet in total diameter, having within the thickness of its walls a series of chambers and galleries, lighted by windows, all looking into the central area. The only aperture to the outside is a doorway through a tunnel-like square-headed passage, with slightly inclined sides, 5 to 6 feet high, and about 3 feet wide, constructed in the thickness of the wall. The latter, which is from 9 to 20 feet thick, is solid for 1 On the natural origin of some of those markings, see a Paper by Professor Duns, Proceed. Soc. Ant., xx. 126: 1886. about 10 feet from the ground, except where it is pierced by the entrance or partially hollowed out for chambers."1 The Torwoodlee Broch is situated within the line of a British fort, of more than average dimensions, on the Crosslee Hill, called also Harrigait Head, at an elevation. of 900 feet above sea-level. The diameter of the fort within the walls from west to east is given as 350 feet, and from north to south, when perfect, as 412 feet. It has portions of mounds or walls on the west nearly 25 feet in thickness, and a ditch on each side of 24 feet in width. Excavation in May 1891 revealed for the first time, within the area of this fort, the foundations of what appears to be a true broch, of the type north of the Caledonian Canal. The wall, about 3 feet in height, is 17 feet 6 inches in thickness, and it encloses an open central area, of which the diameter is 38 feet. The wall of the broch on Cockburn Law in Berwickshire varies from 15 to 20 feet, and the largest diameter of both walls and open space is 92 feet. The diameter of the Torwoodlee Broch through walls and central space is 75 feet. The entrance, of the usual characteristic kind, is well marked, and there are traces of stairs leading to upper chambers in the walls. Charcoal and fragments of pottery, glass, stone, and bronze implements, have been found on digging the floor. The special interest of this broch arises from the fact that it is within a fort on the line of the Catrail, at indeed a point where this line is usually supposed to terminate.2 XXV., 1 Mr Dalrymple Duncan in Trans. Glasgow Archæological Society, No. 1886. See Mr J. Anderson, Scotland in Pagan Times, 174 et seq. 1883. 2 These particulars are given partly from communications to the Scotsman, May 14 and June 13, 1891, of the work of the Galashiels Ramblers' Club; and partly from personal examination of the fort, before and since the discovery of the broch. The question suggests itself as to whether any more of those structures, so rare south of the Caledonian Canal, are to be found in the line of the Catrail, or whether this is an exceptional case, leading rather to the supposition of a northern or foreign tribe having made, for a time at least, a successful capture of the Catrail boundary, and planted at its seeming termination, within the ordinary British fort, a work of defence greatly stronger than the usual ringed fort. If this be so, the popular designation of the Picts Work Ditch applied to the Catrail in Selkirkshire may not be wholly wrong. It must, however, be kept in mind that as yet there has been no thorough, careful, and skilled examination, especially by excavation of the classes of remains here referred to in the valley of the Tweed and the Lowlands generally. In the absence of this we must hold any conclusion regarding them with reserve, and subject to modification. 54 CHAPTER III. ORIGINAL INHABITANTS-NAMES OF PLACES AND ANOTHER important and perhaps more definite source of information and inference regarding the original inhabitants of the Lowlands is to be found in the language of the district. Here we must look both at the vernacular spoken and written, and particularly at the names of places and natural features of the country. These latter are frequently the symbols of races of men, which witness for them when they are gone, and when there is neither memory nor trace of their homes or their graves. Any one who scans the Ordnance map of the valley of the Tweed and the Lowlands generally will readily be impressed with the fact that the great proportion of names there is Teutonic. The dwelling-places of men, the most of the smaller streams or burns, the shaws, the muirs, and the lower hills, bear Teutonic appellations.1 It is, however, by no means an easy matter to assign to each of the different branches of the Teutonic language its share of those names, whether we look to the chief or 1 Cf. Murray's Dialect of the South of Scotland, 16. to the subordinate ramifications of the language represented by that name. And here it should be explained that certain terms connected with this point,-especially Saxon and Anglo-Saxon, are somewhat vague, and open to misconception. When I use Anglo-Saxon as applied to a portion of the Teutonic speech and people, I intend it to apply to the three tribes, Jutes, Angles, and Saxons, who were allied in race, and who used a common language, with, of course, dialectical differences. These tribes certainly came to be supreme over a large portion of the early Celtic area of Britain, and they imposed their language on the sphere of their influence. The Romans and Celts applied the term Saxon to those three tribes individually and collectively, while the invaders apparently called themselves Angles. No Angle, however, according to Mr Freeman, seems ever to have named himself Saxon. The use of the term Saxon, even as applied to the whole three tribes, or to each, is thus historically justifiable; but it would not be expedient. The phrase Anglo-Saxon as applied to the whole three tribes dates from about the tenth century, and is equivalent to the Angli-Saxones or Angul-Seaxe of that period. The portions of Britain usually regarded as colonised more especially by the Angles are-(1) East Anglia, to the north of the East Saxons, or Essex. (2) Northumbria, including Bernicia and Deira. These embraced the district on the east coast of the island, extending from the Humber to the Forth: Deira stretched from the Humber to the Tees, Bernicia from the Tees to the Forth. (3) Mercia, which embraced nearly all central England westwards to the borders of Wales. The Angle |