Clearly to understand the origin of the drift, and the method by which it attained its present distribution, it may be well to consider some elementary facts and principles concerning climate and its effects, even at the risk of repeating what is already familiar. _Snow fields and ice sheets._--The temperature and the snowfall of a region may stand in such a relation to each other that the summer's heat may barely suffice to melt the winter's snow. If under these circumstances the annual temperature were to be reduced, or the fall of snow increased, the summer's heat would fail to melt all the winter's snow, and some portion of it would endure through the summer, and through successive summers, constituting a perennial snow-field. Were this process once inaugurated, the depth of the snow would increase from year to year. The area of the snow-field would be extended at the same time, since the snow-field would so far reduce the surrounding temperature as to increase the proportion of the annual precipitation which fell as snow. In the course of time, and under favorable conditions, the area of the snow-field would attain great dimensions, and the depth of the snow would become very great. As in the case of existing snow fields the lower part of the snow mass would eventually be converted into ice. Several factors would conspire to this end. 1. The pressure of the overlying snow would tend to compress the lower portion, and snow rendered sufficiently compact by compression would be regarded as ice. 2. Water arising from the melting of the surface snow by the sun's heat, would percolate through the superficial layers of snow, and, freezing below, take the form of ice. 3. On standing, even without pressure or partial melting, snow appears to undergo changes of crystallization which render it more compact. In these and perhaps other ways, a snow-field becomes an ice-field, the snow being restricted to its surface. Eventually the increase in the depth of the snow and ice in a snow-field will give rise to new phenomena. Let a snow and ice field be assumed in which the depth of snow and ice is greatest at the center, with diminution toward its edges. The field of snow, if resting on a level base, would have some such cross-section as that represented in the diagram, Fig. 27. When the thickness of the ice has become considerable, it is evident that the pressure upon its lower and marginal parts will be great. We are wont to think of ice as a brittle solid. If in its place there were some plastic substance which would yield to pressure, the weight of the ice would cause the marginal parts to extend themselves in all directions by a sort of flowing motion. [Illustration: Fig. 27.--Diagrammatic cross-section of a field of ice and snow (c) resting on a level base A-B.]