GEOLOGICAL SEASONS. 191 



as expansion would increase it ; so that the equatorial 

 protuberance of the waters is quite imaginary. 2. The 

 directions of the superficial and deep currents are not 

 such as would result from a normal circulation like that 

 in the atmosphere. That is, the out-going currents, mod- 

 ified by terrestrial rotation, and neglecting the influence 

 of continental barriers, should be toward the northeast and 

 southeast (with an eastward direction immediately over 

 the equator) in the upper portion of the film' and the 

 returning currents should be from the northeast and 

 southeast (with a westward direction over the equator) in 

 the lower portion of the film. Observation shows, how- 

 ever, that the upper portion of the watery film is charac- 

 terized by movements coincident with those of the lower 

 portion of the atmospheric film. The two sets of motions 

 cannot, therefore, be traced separately to the same phys- 

 ical cause. 3. The amount of expansion of the inter- 

 tropical waters would not be adequate to cause a tendency 

 to flow toward the poles. As Mr. Croll has shown, the 

 difference between the equatorial and polar temperatures 

 of the waters would disturb the equilibrium by only the 

 trifling amount of four and a half feet. Distribute this 

 between the equator and the poles, and the descent would 

 not be sufficient to overcome the viscidity of the water. 

 4. Should any intertropical protuberance exist as a result 

 of the cause assigned, and should it produce a flow toward 

 the poles, the process would only continue until the pro- 

 tuberance should be removed. No cause can be assigned 

 why the deeper, colder and heavier water should rise into 

 the lighter, and reproduce the protuberance. That is, 

 the water, however rarefied, would reach a state of statical 

 equilibrium, and remain so thereafter. It may fairly be 



