PHONETIC DESCRIPTION OF SCOTTISH LANGUAGE AND DIALECTS with it what has been termed “school English.” The Gaelic language has lost ground so rapidly that its effective eastern boundary would now lie much to the west of the old line, its place being taken for the most part by St.Eng.; on the other hand Lowland Sc. still reaches the line. Indeed, at some points (Campbeltown, Grantown) the nearest Sc. dialect may be said to have crossed the line and invaded Gaelic territory; while a yet more noticeable invasion, of a phonologic order, has been the spread in the Western Highlands of the tone system of sw.Sc. (Glasgow). § 8. Any Scots-Gaelic border line drawn on the map to-day must be regarded not only as a generalisation, but as considerably more of a generalisation than the corresponding line which it was possible to draw as recently as forty-five years ago. Two important criteria were formerly available for determining whether a district was Highland or Lowland in tongue. The first of these was the language, Gaelic or English, used in church. Particulars upon this point contained in the Statistical Accounts throw not a little light upon the linguistic situation in various parishes along the Highland border at the end of the 18th and in the early 19th centuries; and until the first decade of the 20th cent. the language used in church was a true index of the linguistic affinity of a district. The decline in church attendance has since deprived this criterion of its value. There remains another, the Census figures of Gaelic speakers, available in totals and percentages for each parish. These form an interesting record so far as they go, but they require to be handled with caution if a true picture of the state of the language is aimed at; because, while the Census figures may give the number of people in a parish who can converse in Gaelic, it is certain that any qualitative test applied over the border area would show that much of the language represented by these figures is deficient in vocabulary and faulty in form, the decline in the number of Gaelic speakers between one Census enumeration and another being accompanied by a loss of quality probably equal to the loss of quantity. The Census figures, none the less, represent a definite claim on the part of individuals to an ability to speak Gaelic, and have their statistical value as such. § 9. The first mapping of a Highland-Lowland linguistic boundary was Sir James Murray's, described in his Dialect of the Southern Counties of Scotland (1873). Murray's line is stated to be based upon information received from clergymen and others, and it possesses permanent value as a record because the data available at that time justified the drawing of a more definite line than is now possible. His line “passes along the east coast of Arran, cuts off the north of Bute, passes behind Dunoon to Loch Long, enters Dumbarton at Gorton, hence through Glen Douglas to Loch Lomond; it enters Stirling north of Rowardennan, crosses to Aberfoyle and to Callander, passes through Glenartney to Comrie, crosses Glenalmond south of Amulree, follows Strath Braan through Birnam Wood to Dunkeld. It enters Aberdeenshire by Mount Blair, passes to a point four miles east of Braemar, and hence on to two miles east of Crathie and Balmoral. It then proceeds north-north-west to go to Strathdon, where it turns north-west and enters Banff six miles north-east of Tomintoul. It skirts the Livet on the west to the boundary of Elgin. It crosses the Spey two miles south of Inveravon, traverses the Knock of Brae Moray, and hence north-west to Nairn, crossing the Findhorn at right angles and going on to Ardclach, and hence to the Moray Firth, three miles west of Nairn. It crosses the Firth to Cromarty, dips again into the sea, to emerge at Clyth Ness, Caithness. It proceeds overland to Harpsdale, through Halkirk to the river Forss, which it follows to the sea. . . .” See Introduction to A. Warrack's A Scots Dialect Dictionary (Chambers, 1911). Scottish Limit on the West. § 10. The revision of this line for the Scottish National Dictionary is based (1) on the Census Returns for the parishes on Murray's line, and to the west of it, and (2) on the reports from these localities sent in by school teachers who have been good enough to answer the following questions: (a) Has English or Lowland Scots taken the place of Gaelic? (b) If Scots, which dialect? (c) Is Gaelic the speech of the school playground? (d) Is Gaelic a medium for school instruction in the infant room? The course of this revised line, as will be seen, takes it along the south of Argyllsh. in such fashion as to keep the urban communities on the coast of the Firth of Clyde within the Lowland