RECAPITULATION. 381 



ing higher grounds, washed in by land-floods,"* Mr. Tylor 

 virtually adopts the explanation of the phenomena given in 

 this work, since the formation or removal of this gravel neces- 

 sarily involved an alteration of the surface and a deepening 

 of the valley. 



When, finally, the excavation of the valley was completed, 

 the climate had gradually become more like our own, and 

 either from this change, or rather perhaps yielding to the irre- 

 sistible power of man, the great Pachydermata became extinct. 

 Under the altered conditions of level, the river, unable to 

 carry out to sea the finer particles brought down from the 

 higher levels, deposited them in the valley, and thus raised 

 somewhat its general level, checking the velocity of the stream, 

 and producing extensive marshes, in which a thick deposit 

 of peat was gradually formed. We have, unfortunately, no 

 trustworthy means of estimating' the rate of formation of this 

 substance, which indeed varies considerably, according to the 

 conditions of the case ; but on any supposition the production 

 of a mass in some places more than thirty feet in thickness 

 must have required a very considerable period. Yet it is in 

 these beds that we find the remains of the Neolithic or later 

 Stone period. From the tombs at St. Acheul, from the Eoman 

 remains found in the superficial layers of the peat, at about 

 the present level of the river, we know that fifteen hundred 

 years have produced scarcely any change in the configuration 

 of the valley. In the peat, and at a depth of about fifteen 

 feet in the alluvium at Abbeville, are the remains of the Neo- 

 lithic period, which we have ample reason for believing, from 

 the researches in Denmark, Switzerland, and other countries, 

 to be of no slight antiquity. Yet all these are subsequent 

 to the excavation of the valley. What date then are we to 

 ascribe to the men who lived when the Somme was but begin- 

 ning its great task ? No one can properly appreciate the lapse 



* 1. c. p. 105. 



