614 ESrCIDENTAL EFFECTS OF HUMAN ACTION". 



ure, because the water filtered tlirougli the porous bottom of the 

 pans, leaving little salt behind. But this was a small evil, com- 

 pared with other destructive consequences that followed. When 

 the sea was driven into the cave by violent west or northwest 

 winds, it shot 2ijet d^eau through the well to the height of sixty- 

 feet, the spray of which was scattered far and wide over the neigh- 

 boring gardens and blasted the crops. The well was now, closed 

 with stones, but the next winter's storms hurled them out again, 

 and spread the salt spray over the grounds in the vicinity as be- 

 fore. Repeated attempts were made to stop the orifice, but at the 

 time of Parthey's visit the sea had thrice burst through, and it 

 was feared that the evil was without remedy.* 



I have mentioned the great extent of the heaps of oyster and 

 other shells left by the American Indians on the Atlantic coast 

 of the United States. Some of the Danish kitchen-middens, which 

 closely resemble them, are a thousand feet long, from one hun- 

 dred and fifty to two hundred wide, and from six to ten high. 

 These piles have an importance as geological witnesses, independ- 

 ent of their bearing upon human history. "Wlierever the coast 

 line appears, from other evidence, to have remained unchanged 

 in outline and elevation since they were accumulated, they are 

 found near the sea, and not more than about ten feet above its 

 level. In some cases they are at a considerable distance from the 

 beach, and in these instances, so far as yet examined, there are 

 proofs that the coast has advanced in consequence of upheaval or 

 of fluviatile or marine deposit. Where they are altogether want- 

 ing, the coast seems to have sunk or been washed away by the sea. 

 The constancy of these observations justifies geologists in arguing 

 where other evidence is wanting, the advance of land or sea re- 

 spectively, or the elevation or depression of the former, from the 

 position or the absence of these heaps alone. 



Every traveller in Italy is familiar with Monte Testaccio, the 

 mountain of potsherds, at Rome ; f but this deposit, large as it is, 



* Wanderungen durch Sieilien und die Levante, i., p. 406. 



f Until recently this hillock was supposed to consist of sherds of household 

 pottery broken in using, but it now appears to be ascertained that it is com- 

 posed chiefly of fragments of earthenware broken in transportation from the 

 place of manufacture to the emporium on the Tiber where such articles were 

 landed. 



