Jormed by Ancient Glaciers. 349 
the air near the lateral edges of the glacier, caused by radia- 
tion from the neighbouring rocks, occasions a more copious 
melting of the ice and snow in those parts, and consequently 
a greater quantity of water is introduced into what remains 
unmelted. This water being refrozen at night, produces a 
greater expansion near the edges than in the central portions 
of the glacier; and this excess accelerates the motion of the 
sides downwards. It must be recollected that a glacier is not 
one solid piece of ice, but is broken up and intersected by 
many chasms or fissures of greater or less width and depth, 
which allow an independent motion of its different parts. 
This greater expansion has a tendency also to divert the de- 
scending stones from a rectilinear course, and to carry them 
toward the point of least resistance, in other words, nearer 
to the sides; so that a large portion of them, instead of reach- 
ing the foot of the glacier, are deposited in longitudinal or 
irregularly curved lines on the inclined slopes of the contigu- 
ous rocks, their peculiar form being modified by local cireum- 
stances. It is also evident that the greatest accumulations of 
these lateral shelves, or morains, will be found near and upon 
the most prominent slopes of rock, especially on the side next 
to the head of the glacier; because these projections not only 
arrest the stones in their downward course, but by their 
agency in causing more radiation, melting and freezing, at- 
tract, if 1 may so say, a greater quantity of surface wreck. 
Again, as the opposite sides of a glacier at any given point 
have a general coincidence of level, these morains will often 
be found to correspond in horizontal position; though, for 
obvious reasons, not so preciselyas the opposite shores of a 
bay or lake. They must also be formed solely of fragments 
_ from the higher surrounding rocks, not rounded into pebbles, 
but more or less angular, or mixed with clay or earth, in pro- 
portion to the nature and hardness of the material. Neither 
sea nor freshwater shells will be found among them. Other 
morains are formed at the foot of the glacier, and often pre- 
sent very different appearances; but as it is to the lateral 
ones that most of the terraces on the Galashiels hills are to be 
referred, I shall not pursue the explanation further. 
I think a careful perusal of the details given in my former 
paper will show that a large portion of these terraces, or 
rather shelves, correspond with the morains whose origin I 
have just been tracing. Their broken and interrupted cha- 
racter on the Kastern Eildon and on Williamlaw ; their irre- 
gular width and rude horizontality of surface, combined with 
a general coincidence of level; their angular stones and the 
total absence of gravel, sand or shells, are precisely such as 
