232 M. L. FULLER THE SEA MILLS OF CEPHALONIA 



tain special conditions without which it can not act. In the case of the 

 first, either an unsymmetrical passage of a form specially adjusted to 

 the development of a higher temperature in the outlet than in the inlet 

 arm, or an iinequal distribution of heat, is necessary, while in the second 

 the heads of the fresh and salt columns must be essentially balanced. 

 From the underground water standpoint neither supposition is inher- 

 ently improbable. Temperature can hardly fail to act if the water 

 reaches to any considerable depth, while dilution may be a factor at any 

 depth. The writer believes that such dilution is inevitable, and that 

 while the heat hypothesis is by no means improbable, that of dilution 

 is to be regarded as a possible alternative. If dilution occurs at all, it 

 will of necessity be several times as effective as heat. 



