50 MK. GEIKIE ON UPHEAVAL OF CHAP. in. 



tain their relative ages, unless they had been found vertically 

 above each other. The varying depths of an estuary, its 

 banks of silt and sand, the set of its currents, and the in 

 fluence of its tides in scouring out alluvium from some parts 

 of its bottom and redepositing it in others, are circumstances 

 which require to be taken into account in all such calculations. 

 Mere coincidence of depth from the present surface of the 

 ground, which is tolerably uniform in level, by no means 

 necessarily proves contemporaneous deposition. Nor would 

 such an inference follow even from the occurrence of the 

 remains in distant parts of the very same stratum. A canoe 

 might be capsized and sent to the bottom just beneath low- 

 water mark ; another might experience a similar fate on the 

 following day, but in the middle of the channel. Both 

 would become silted up on the floor of the estuary ; but as 

 that floor would be perhaps twenty feet deeper in the centre 

 than towards the margin of the river, the one canoe might 

 actually be twenty feet deeper in the alluvium than the other ; 

 and on the upheaval of the alluvial deposits, if we were to 

 argue merely from the depth at which the remains were 

 embedded, we should pronounce the canoe found at the one 

 locality to be immensely older than the other, seeing that the 

 fine mud of the estuary is deposited very slowly and that it 

 must therefore have taken a long period to form so great a 

 thickness as twenty feet. Again, the tides and currents of 

 the estuary, by changing their direction, might sweep away 

 a considerable mass of alluvium from the bottom, laying bare 

 a canoe that may have foundered many centuries before. 

 After the lapse of so long an interval, another vessel might go 

 to the bottom in the same locality, and be there covered up 

 with the older one, on the same general plane. These two 

 vessels, found in such a position, would naturally be classed 

 together as of the same age, and yet it is demonstrable that a 

 very loDg period may have elapsed between the date of the 



