HIS GEOLOGICAL RELATIONS. 123 



caverns becomes at once an established certainty. No 

 creature save man lights a fire, no creature save man 

 ever lighted one ; and the testimony of these wood- 

 ashes as to the existence of a cave-dwelliag race is as 

 conclusive as if we had witnessed their grimy coun- 

 tenances lighted up by the fires of which these frag- 

 ments were the latest embers. Or again, suppose we 

 are excavating a canal along some level plain, and 

 pass through first a layer of soil, then a bed of peat, 

 next a layer of shelly marl, and lastly through a 

 stratum of clayey silt, the obvious inference is, we are 

 cutting through the sediments of some ancient lake — 

 the sUt, the shell-marl, and the peat-earth marking its 

 successive stages of fiUing-up and obliteration. Of 

 these stages we have no chronology in years, but were 

 Roman remains found imbedded in the peat, and tree- 

 canoes in the silt beneath, these would prove that 

 some two thousand years ago the place was in the 

 condition of a marsh, and that long before — it may 

 have been thousands of years before — it was the site 

 of a lake over which the early inhabitants of the 

 country paddled their rude canoes. We arrive at the 

 first inference from our knowledge of the time our 

 country was invaded by the Romans, and at the second 

 from our equally certain knowledge of the slow rate 

 at which lake-sUts accumulate and shell-marls are 

 formed. Such is the nature of geological evidence, 



