If there have been differences of opinion concerning the nature of ice epochs, as distinct from each other and from ice periods, there has been a failure to adequately apprehend the nature, the extent, and the meaning of the real criteria on which the final recognition of separate ice epochs, if such there were, must be based. Such criteria are several in number. They are of unequal value. In some instances a single one of them might be quite sufficient to establish the fact of two ice epochs. In other cases, single criteria which might not be in themselves demonstrative, have great corroborative weight, when found in association with others. In all cases, much discretion must be used in the interpretation of these criteria. They may be enumerated under several specific heads. (1) _Forest Beds._ Beds of vegetal deposits or old soils are frequently found between layers of glacial drift. This is one of the criteria most commonly cited, because it is of common occurrence and easy of recognition. The advocates of the unity of the glacial period maintain that such beds of organic matter might become interbedded with morainic debris during minor oscillations of the ice's edge. The phenomena of existing glaciers make it evident that forest beds or soils might be enclosed by the deposits of an oscillating ice edge. By repeated oscillations of the ice's edge during the general retreat of the ice, such vegetal beds might become interstratified with glacial drift more or less frequently over all the area once covered by the ice, and from which it has now disappeared. The mere presence of vegetable material between beds of drift is therefore no proof of distinct ice epochs. This does not destroy the value of the vegetal beds as a criterion for the recognition of distinct ice epochs, but it makes caution necessary in its application. It does not follow that, since _some_ inter-drift forest-beds do not prove interglacial epochs, _none_ do. The question is not how forest-beds might originate, but how existing forest-beds did originate. Where the plant-remains found in the relations indicated are so well preserved as to make identification of the species possible, we have a means of determining, with some degree of accuracy, the climatic conditions which must have obtained at the place where the plants grew during the time of their life. If these interbedded plant-remains are of such a character as to indicate a temperate climate, we can not suppose that they grew at the immediate edge of the ice, and therefore that they were buried beneath its oscillating margin. To be specific, if the inter-drift plant remains in any given locality of the area once covered by ice are such as to indicate a climate _as warm as the present in the same locality_, the ice must have receded so far to the northward that its re-advance might, in our judgment, appropriately be regarded as a separate ice epoch.