230 M. L. FULLER THE SEA MILLS OF CEPHALONIA 



of the latter are not likely to be duplicated in nature. The hypothesis 

 appears to be untenable as an explanation of the circulation at Cephalonia 

 for the following reasons: (1) Air currents of any strength in under- 

 ground passages must be the result of water movements rather than vice 

 versa, so that even if air were present no suction giving rise to an indraft 

 of water could be produced ;f (3) the underground passages below the 

 sea or ground-water level are necessarily completely filled with water, 

 thus precluding the existence of air other than as an included gas in- 

 capable of producing currents; (3) no high velocities are possible in 

 underground passages below the water table, where, as at Cephalonia, 

 only moderate heads are available; and (4) the postulated suction can 

 not, according to the laws of hydrostatics, take place between two liquids 

 like those at Cephalonia, which have, as has been pointed out, substan- 

 tially equal heads. 



Source of water. — If the depth at which the commingling of salt and 

 fresh waters takes place were known, an estimate of the head, and indi- 

 rectly of the source of the fresh waters, could be made. The point of 

 union apparently can not be excessively deep, as the limestone is probably 

 not over a few thousand feet thick. It is probable that the water comes 

 from the island itself, most likely within a few miles of the sea mills, 

 rather than from the mainland, some 30 miles away. 



Point of emergence. — It is a matter of observation that there are no 

 large saline springs near at hand, either on Cephalonia, Zante, or on the 

 adjacent portions of the mainland, making it almost certain that the out- 

 let is at some point beneath the sea. If the outlet were in very shallow 

 water, near shore, it could hardly have escaped recognition. As has been 

 pointed out, however, the weight of the overlying body of sea water will 

 oppose the emergence of the mixed column if the outlet is below sealevel, 

 the pressure to be overcome by the latter being equivalent to 1^/^ feet of 

 its head for every 100 feet of depth of the sea. If, therefore, as is usually 

 the case, the head of the underground water at the coast is low, it will 

 follow that the point of emergence can be at no great depth and probably 

 at no great distance from the sea mills. 



RELATIVE COMPETENCY OF TEMPERATURE AND DILUTION 



Something of the relative competency of temperature and dilution has 

 been indicated in the course of the foregoing description, but it is more 

 forcibly brought out by the following table : 



t Air currents suclj as those flowing in ani out of caves are too feeble to produce any 

 material suction, 



