36 PROCEEDINGS OF THE AMERICAN ACADEMY. 



sea at most ports corresponding with the change of seasons. The prin- 

 cipal motion of the water of the Atlantic Ocean affecting its level is 

 the motion by which it is supposed to complete a gyration in about 

 three years. In -the paper already referred to it was shown that the 

 force arising from this gyration would cause the middle of the gyrating 

 mass of the water to stand about five feet higher than the exterior part 

 on the coast of Europe and America. Now as the greatest difference 

 of temperature in the ocean between the equator and the poles must 

 be in the latter part of winter, a little later than the time of the greatest 

 difference in the atmosphere between the equator and the poles, the 

 greatest gyratory motion of the water of the Atlantic Ocean, on account 

 of the inertia of the water, must happen still a little later, say in April, 

 and then the surface of the water must stand highest in the centre of the 

 gyrating part, and lowest at the exterior part, and consequently at the 

 ports of Brest and Key West. On the contrary, in October, when the 

 gyratory motion is the least, the surface must fall a little in the middle 

 and rise a little at its exterior part, and consequently stand at its maxi- 

 mum height at the ports of Brest and Key West. The position of the 

 gyrating mass also changes with the seasons, being farthest north in the 

 fall, and nearest the equator in the spring, as must necessarily be the 

 case, and as the vibrating motion of the northern part of the Gulf 

 Stream indicates. The surface of the gyrating water being a little 

 convex, this circumstance must also affect the mean level slightly at 

 som6 ports. 



The difference between the maximum and minimum mean height 

 of the sea at Brest is about six inches, which, when corrected for the 

 effect of the winds and other causes, would probably be a little more. 

 The difference at Key West, corrected in like manner, would probably 

 be about the same. A decrease, therefore, of less than one half in the 

 gyratory velocity of the ocean from April to October would be suffi- 

 cient to cause a variation of that amount in the mean level of the sea 

 at Brest and Key West ; and as the argument of the variation of the 

 gyratory motion of the Atlantic Ocean, as we have seen, must very 

 nearly or quite correspond with the argument of the variation of mean 

 level unaccounted for by other causes, we have reason to think that the 

 variation of the gyratory motion of the Atlantic Ocean is the cause of 

 this part of the change of mean level. 



