Let _a_ represent the temperature at the commencement of earth-history and _b_ that necessary for glaciation, and _bc_ the lapse of time between then and now. The curved line indicates the gradual fall in temperature due to diminution of the amount of heat, while the zigzag line represents the oscillations due to secular climatic changes. If the Cambrian period x occurred comparatively soon after the commencement of earth-history as shown in fig. _A_, no glaciation could be produced, even during periods when secular changes caused colder conditions than the mean, whereas if the Cambrian period occurred at a time very remote from the commencement of earth-history as shown in _B_, glacial conditions could be produced then as now, for the mean temperature, as shown by the distance of the curve from the line _bc_, would be practically as it now is. The studies of the last few decades have brought into prominence the occurrence of glacial periods in remote times, probably in early Palæozoic times; and as far as the mean temperature of the earth's surface is concerned, it would appear, from the knowledge in our possession, that matters were not very different in those early times from what they now are. [Illustration: Fig. 25.] Some further remarks will be made in subsequent paragraphs concerning the period of the earth's history at which the geologist is first furnished with definite records, but in the meantime it may be observed that the geologist will do well, when working amongst the strata, to consider that the more active operation of agents, even in times of which he has definite knowledge, may have produced effects which he should be prepared to discover, as their discovery would be of considerable importance, and that he should not be content to infer that because it has been proved that agents operating with the same intensity as that which they have at present, _may_ have produced all the effects which he can actually observe, they therefore necessarily _did_ produce them.