242 GOODCHILD : GLACIAL PHENOMENA OF UPPER RIBBLESDALE. 
been learnt by only a few. Obviously that lesson is, that the direction 
indicated by glacial strize is not necessarily the same as that followed 
by the upper part of the ice. The upper part of the ice is known to 
have moved, in many cases, at right angles to the course taken by 
its lower strata. The movements of the two parts may even have 
been in diametric app site directions. Again, land ice flows in 
the dir of least resistance, so that its line of movement, as 
orbes showed jel ago, as often as not lies from the base at 
one point to the surface at another point farther from its source.* 
This cause, combined with the effects of surface ablation, tends to 
elevate stones or other bodies caught up by the sole of the ice, nearer 
and nearer to the surface, in opposition to the force of gravity.t 
Another factor to be taken into account in dealing with boulder 
distribution is, thatan impulse in a given direction imparted to ice 
at one point is transmitted along its upper strata for miles beyond its 
initial point, and may even be communicated, to some extent, to 
strata far below the surface. Connected with this is the larger 
years. But so far as the study of British glacial phenomena 1s 
concerned none of these explanations make it clear how it came 
about that the sole of the ice-sheet moved so steadily, and with such 
even motion, over large areas, as to enable it to plough out continuous 
glacial furrows, fifteen, twenty, or more feet in length. Yet that is 
one of the most important facts with which a glacialist has to deal. 
My own explanation} is a new one. I am convinced that under 
conditions of very low temperature a large mass of ice occupying 
a deep valley, will flow outward more rapidly at the do¢tom than at 
the surface. The reasoning is based on the fact that at all degrees 
below the freezing point ice contracts under a fall of temperature and 
expands under a rise, more than any substance known. ‘Therefore, 
when but little affected by heat-waves, the upper part of a thick mass 
of ice undergoes contraction to an extent that may, in certain cases, 
be hardly more than compensated by the downflow of ice from the 
rent source. On the other hand, the cold waves (temperature 
undulations below Zero C.) do not lower the temperature of the 
e Dr. Drygalski, Zeit. Ges. fiir Erdkunde zu —— Band XXVII, No. t 
i ee photographic illustrations of this phenomeno 
+ In the Geol. Mag. for November, seth I gave this as the explanation of the 
occurrence of the shells on Moel Tryfaen, 
t Geol. Mag., 1891, p. 19. * Motion of Land Ice.’ ee 
a : Naturalist,” 
