242 



DEPARTMENT OF THE NAVAL .SERVICE 



layer. The divergence of the isosteric surfaces (vide fig. 25) is one of the surest) 

 indications of the existence of such a deep water current. 



Fig. 2.5. — Screwing movement in a deep water 

 current. 



Finally, some experiments made with a rotating water tank remain to be 

 described. This tank was 30 cm. long, 10 cm. broad, and 10 cm. high. A surface 

 layer of fresh water, 4 cm. deep, was poured in, and a bottom layer of salt water, 

 also 4 cm. deep, introduced beneath the first. Before the tank was set in motion, 

 a vertical current of air was applied to the surface of the water; the result will be 

 seen from fig. 26. Under the influence of this current of air, the surface layer> 

 diminished in thickness, the bottom layer increasing, evidently as a result of fric- 

 tion between the surface of the water and the air, which poured out radially to all 

 sides. The tank was then caused to rotate about a vertical axis through its centre, 

 when the situation shown in fig. 27 was observed ; i.e., the exact opposite of that shown 

 in fig. 26. The cause of this accumulation of the surface water under the air cur- 

 rent is evidently this: the rotation of the surface water is somewhat retarded by 

 the radially directed air, which renders its velocity less than that of the tank, 

 whereas the bottom layer of water rotates at the same velocity as the tank itself. 

 Consequently the centrifugal force would be greater in the bottom layer than in the 

 surface water, and the bottom water would therefore be driven out to the ends of 

 the tank, while the surface water massed in its centre. 



w/////yy///y///^'i'/^////y^//'y/'^/4V ^ / ' ^r'yyy^^yy^r^yyyy./y^yyy^^^. 



Fig. 2H. — Experiment with vertical air bla.st against surface of water in layers. 

 (Tank at rest.) 



liastly, by means of a series of obliqi;ely placed tubes, a cyclonic circulation 

 of air was applied to the surface of the water, the effect of this being to induce a 

 rotation of the surface water at a higher velocity than that of the tank. This 

 brought about the same distribution of the water mass as shown in fig. 26, evidently 

 here owing to the fact that the centrifugal force was in this case greater in the sur- 

 face layer than in the bottom water. 



In the sea, therefore, an anticyclone should occasion an accumulation of the 

 surface water beneath its centre, a good example of which is afforded by the Sargasso 

 sea. A cyclone, on the other hand, would drive the surface water outward? to all 

 sides, and draw up the bottom water beneath its centre. 



