It is conceivable that, after the development and extension of a continental ice-sheet, it might be wholly wasted away. The maximum extension of such an ice-sheet would mark the culmination of a glacial epoch. If subsequently another ice-sheet of considerable dimensions were accumulated, its development and extension would constitute a second glacial epoch. These successive ice-sheets might be so related to each other in time, in position, and in the sequence of geological events, as to be regarded as separate epochs of the same glacial period.[14] On the other hand they might be so widely separated from each other in time, in position, and in the sequence of geological events, as to make their reference to separate glacial periods more appropriate. In any case their separation would be sufficiently marked to necessitate their reference to separate ice epochs. So far we believe there would be no disagreement. [14] The terms period and epoch are here used in the sense in which they have been used most commonly in the literature of glacial geology in the United States. If, instead of entirely disappearing, the first ice-sheet suffered great reduction of volume and area, and if this reduction were followed by a second great expansion of the ice, might the time of such expansion be regarded as a second glacial epoch of the common glacial period? To this question, too, as thus stated, we apprehend there would be but one answer, and that affirmative. It seems certain that the edge of the continental ice-sheet was subject to more or less extensive oscillations, as are the ends of glaciers and the edges of ice-sheets to-day. How much of an oscillation is necessary, and under what attendant conditions must it take place, in order that the recession of the ice-edge shall mark an interglacial and its re-advance a distinct glacial epoch? When the question takes this specific form, and when inquiry is made concerning the quantitative value of the different elements entering into the problem, we reach the battled ground. It is the battled ground, partly because it is the ground of misunderstanding. It is the ground of misunderstanding, partly because glacialists are not agreed as to the meaning of certain terms in common use by them. Four elements seem to enter into the idea of an ice epoch as distinct from other ice epochs. These are (1) the distance to which the ice retreated between successive advances; (2) the duration of the retreat, or the time which elapsed between successive ice extensions; (3) the temperature of the region freed from ice during the time between maxima of advance; and (4) the intervention between successive advances, of changes interrupting the continuity of geological processes.