1873.. WARD — LAKE-DISTKICT GEACIATION. 433
Fig. 5. — Contour-map, showing the form of the land at 2000 feet.
VII. Axioms to be bokne in mind in dbawinq Conclusions.
In drawing conclusions from all that has now been brought for-
ward, I would wish to bear in mind the following axioms : —
1. That a great series of glacial scratches, all pointing in the
same direction, evidences the onward movement of a sheet of land-
ice, and is not likely to be due to drifting and floating ice.
2. That the height to which such uniform scratches extend upon
the summits of mountains is an index of the amount of hill-area
covered by the ice-sheet ; and that the height to which they extend
upon the sides of valleys is an index of the least thickness of the ice
in such valleys.
3. That the onward motion in one constant direction of a great
thickness of land-ice would tend to push and carry forward portions
of the underlying and surrounding rocks in the same direction.
4. That floating ice might bear boulders in directions in which
land-ice could not have transported them.
5. This being the case, it is evident that boulders from very dif-
ferent regions may be commingled, one set being pushed forward
by land-ice from one direction, and another set being floated, at a
different time, from quite another direction.
6. Since changes in geography affect ocean-currents, boulders
may be floated at one stage of submergence in one direction, and at
another stage of submergence in some other direction, thus, again,
causing boulders from distinct regions to be commingled at the same
spot.
7. That where a body of land-ice is confined within the narrow-
est limits, there, if anywhere, will it be most likely to be pushed
over the enclosing walls on one side or the other.