An attempt to ascertain the absolute date of prehistoric events must of necessity fail, since it is based on the improbable assumption that the physical agents have acted uniformly, and that therefore the results may be used as a natural chronometer. The present rate of the accumulation of _débris_, as at the Victoria Cave of the preceding chapter, or of that of silt in the deltas of rivers, such as the Nile, or the Tinière, may convey a rough idea of the high antiquity of prehistoric deposits; but a slight change either of the climate, or of the rainfall, would invalidate the conclusion. When the greater part of Europe lay buried under forest, when Palestine supported a large population, and when glaciers crowned some of the higher mountains of Africa, such as the Atlas, the European and Egyptian climates were probably moister than at the present time, and the rainfall and the floods greater, and consequently the accumulation of sediment quicker than the observed rate under the present conditions. And in the same way all estimates of the lapse of past time, based upon the excavation of a river valley, or the retrocession of a waterfall, such as Niagara, lie open to the same kind of objection. It is not at all reasonable to suppose that the complex conditions which regulate the present rate of erosion, have been the same during the time the work has been done, and it therefore follows that the work done is a measure of the power employed, and not of the length of time during which it has been in operation. We must, therefore, give up the idea of measuring the past beyond the memory of man, as represented in historical documents, by the historic unit of years. We can merely trace a definite sequence of events, separated one from another by uncertain intervals. And for that series of events which extends from the borders of history back to the remote age where the geologist, descending the stream of time, meets the archæologist, I have adopted the term prehistoric.[85] _The Prehistoric Fauna._