The former alternative does not merit any consideration whatever. The latter, as we shall presently see, involves us in an assumption, for which no defense either aprioristic or factual is available. We can, indeed, distinguish between spatial, and temporal, distribution within the narrow limits of a single locality by using the criterion of superposition; for in regions of outcrop, where one sedimentary rock overlies another, the obvious presumption is that the upper rock was deposited at a later date than the lower rock. But the criterion of superposition is not available for the correlation of strata in localities so distant from each other that no physical evidence of stratigraphic continuity is discernible. Moreover the induction, which projects any local order of stratigraphical sequence into far distant localities on the sole basis of fossil taxonomy, is logically unsound and leads to conclusions at variance with the actual facts. Hence the alleged time-value of index fossils becomes essentially problematic, and affords no basis whatever for scientific certainty. As previously stated, the sequence of strata is visible only in regions of outcrop, and nowhere are we able to see more than mere parts of two or, at most, three systems associated together in a single locality. Moreover, each set of beds is of limited areal extent, and the limits are frequently visible to the eye of the observer. In any case, their visible extent is necessarily limited. It is impossible, therefore, to correlate the strata of one continent with those of another continent by tracing stratigraphic continuity. Hence, in comparing particular horizons of various ages and in distinguishing them from other horizons over large areas, we are obliged to substitute induction for direct observation. Scientific induction, however, is only valid when it rests upon some universal uniformity or invariable sequence of nature. Hence, to be specific, the assumption that the time-scale based on the European classification of fossiliferous strata is applicable to the entire globe as a whole, is based on the further assumption that we are sure of the universality of fossiliferous stratification over the face of the earth, and that, as a matter of fact, fossils are always and everywhere found in the same order of invariable sequence.