But when we pass to the great solitary dykes, that run so far and so continuously across the country, we encounter much thicker masses of igneous rock. Most of the measurements of these dykes have been made at the surface, and the variations noted in their breadth occur along their horizontal extension. The Cleveland dyke, which is the longest in Britain, varies from 15 feet to more than 100 feet, with perhaps an average width of between 70 and 90 feet.[182] Some of the great dykes that cross Scotland are of larger dimensions. Most of them, however, like that of Cleveland, are liable to considerable variations in breadth when followed along their length. The dyke which runs from the eastern coast across the Cheviot Hills and Teviotdale to the head of the Ale Water, is in some places only 10 feet broad, but at its widest parts is probably about 100 feet. The Eskdale and Moffat dyke is in parts of its course 180 feet wide, but elsewhere it diminishes to not more than 40 feet. These variations are repeated at irregular intervals, so that the dyke alternately widens and contracts as its course is traced across the hills. Some of the dykes further to the north and west attain yet more gigantic proportions. That which crosses Cantyre opposite Ardlamont Point has been measured by Mr. J. B. Hill, of the Geological Survey, who finds it to be from 150 to 180 feet broad on the shore of Loch Fyne, and to swell out beyond the west side of Loch Tarbert to a breadth of 240 to 270 feet. A dyke near Strathmiglo, in Fife, is about 400 feet wide. The broadest dyke known to me is one which I traced near Beith, in Ayrshire, traversing the Carboniferous Limestone. Its maximum width is 640 feet. [Footnote 182: At Cockfield, where it has long been quarried, it varies from 15 to 66 feet; at Armathwatie, in the vale of the Eden, it is about 54 feet (Mr. Teall, _Quart. Journ. Geol. Soc._ xl. p. 211).] Unfortunately, it is much less easy to get evidence of the width of dykes at different levels in their vertical extension. Yet this is obviously an important point in the theoretical discussion of their origin. Two means are available of obtaining information on the subject--(_a_) from mining operations, and (_b_) from observations at precipices and between hill-crests and valley-bottoms. (_a_) In the Central Scottish coal-field and in that of Ayrshire, some large dykes have been cut through at depths of two or three hundred feet beneath the surface. But there does not appear to be any well-ascertained variation between their width so far below ground and at the surface. In not a few cases, indeed, dykes are met with in the lower workings of the coal-pits which do not reach the surface or even the workings in the higher coals. Such upward terminations of dykes will be afterwards considered, and it will be shown that towards its upper limit a dyke may rapidly diminish in width.