Chap. XVII.] THE WINDS AND EVAPORATION. 193 



the coasts ; ' but I think that the arguments in Book 

 IV. show that, though in an ocean completely cover- 

 ing the globe the tides would not cause any current, 

 nevertheless the breaking-up of them by the coast- 

 lines may necessitate the formation of currents follow- 

 ing the general course of the tide-waves, and, therefore, 

 revolving in each ocean, westwards in the equatorial 

 regions and eastwards in each of the temperate zones. 

 But the question whether the breaking-up of the 

 tides by the coast-lines does or does not lead to the 

 formation of currents through the central parts of the 

 ocean may perhaps be more easily determined by a 

 practical investigation, for the purpose of ascertaining 

 whether or not they are alternately accelerated and 

 retarded at intervals corresponding with the succes- 

 sive passages of the moon across any meridian, than 

 by theoretical disquisitions on the subject. A cur- 

 rent-creating action of the tides in the manner here 

 indicated may, it seems to me, be much more effective 

 than that of the winds, whose action, to some extent, 

 is certain, whereas that resulting from .differences of 

 specific gravity need not necessarily be brought into 

 play at all where more powerful and rapidly acting 

 forces cause a circulation. 



