452 D. MACKINTOSH ON THE ERRATIC BLOCKS OF 
IX. Exprianation oF Map, Prater XXII. 
As it is impossible to ascertain the precise routes taken by boulders, 
in a map it is perhaps least presumptuous to draw straight or slightly 
curved lines from their sources to their terminations. As most of 
the Kirkcudbrightshire granite blocks would appear to have been dis- 
persed from Criffel mountain or the neighbcurhood, to prevent com- 
plication I have represented that mountain as the centre of dispersion. 
The barrier offered by the westerly extension of the Cumberland 
mountains renders it necessary to assume a curve in the route taken 
by these boulders. As I have already described the Shapfell-granite 
dispersion (see Geol. Mag. for Aug. 1870; and Quart. Journ. Geol. 
Soc. for Aug. 1873), and as it has been made the subject of papers 
by other authors, I have only mentioned it in the map. For the 
same reason the stream of large limestone boulders found along the 
east coast of Morecambe Bay is merely mentioned. The boulders of 
Silurian grit, felstone, &c. which went south from the mountain- 
front of Westmoreland are merely mentioned, as it is not certain 
that many of them found their way further south than Morecambe 
Bay. The “greenstone” boulders are not inserted in the map. 
As the Arenig boulders which are believed to have wandered 
as far south as the neighbourhood of Bromsgrove would appear to 
have gone first in an easterly direction, it is obviously necessary that 
a curved route should be assigned to them in the middle part of their 
course. There are probably many boulders which I have not yet seen, 
and the positions of which, therefore, are not shown in the map. 
An attempt to map the positions and courses of boulders is justi- 
fied by the fact that most of them have been found more or less 
imbedded in clay or gravel, at all angles, often standing on end—in 
other words, in the positions in which they were left by the ice 
which carried them during the Glacial period. 
