In pursuing the sequence of implements from the Mousterian back to still earlier types, cave-hunting will as a rule provide one step only, though this is of the greatest value. In a few caves, implements of the type made famous by discoveries in alluvial gravels at S. Acheul in France (and designated the Acheulean type) have been found in the deeper levels. Such a cave is that of La Ferrassie (cf. p. 74); another is that of La Chapelle, in which (it will be remembered) the Acheulean implements underlay the human interment. Kent's Hole in Devonshire is even more remarkable. For the lowest strata in this cavern yielded implements of the earliest Chellean form, though this important fact is not commonly recognised. Such caves are of the greatest interest, for they provide direct evidence of the succession of types, within certain limits. But the indefatigable labours of M. Commont[29] of Amiens have finally welded the two series, viz. the cave-implements and the river-drift implements, into continuity, by demonstrating in the alluvial deposits of the river Somme, a succession of types, from the Mousterian backwards to much more primitive forms. These newly-published results have been appropriately supplemented by discoveries in the alluvial strata of the Danube. Combining these results from the river deposits, and for the sake of comparison, adding those from the caves at Ofnet and Sirgenstein, a tabulated statement (Table II) has been drawn up. The two examples of human skeletons from alluvial deposits given in Table A are thus assigned to epochs distinguished by forms of implement more primitive than those found usually in caves; and moreover the more primitive implements are actually shewn to occur in deeper (_i.e._ more ancient) horizons where superposition has been observed. The greater antiquity of the two river-drift men (as contrasted with the cave-men) has been indicated already by the associated animals, and this evidence is now confirmed by the characters of the implements. It may be remarked again that the details of stratigraphical succession have but recently received complete demonstration, mainly through the researches of Messrs Commont, Obermaier[30], and Bayer[30]. The importance of such results is extraordinarily far-reaching, since a means is provided hereby of correlating archaeological with geological evidence to an extent previously unattained. (_d_) It will be noted that this advance has taken little or no account of actual human remains. For in the nature of things, implements will be preserved in river deposits, where skeletons would quickly disintegrate and vanish. TABLE II.