Chase.] JJ^ [May 19. 



very great, and the currents tend to a gyratory or cyclonic motion, 

 •which appears to be modified by glaciers or local poles of great cold. 

 7. The cyclonic tendency is most marked in the southern hemi- 

 sphere, where the ocean waters experience the least interruption to 

 their normal motions. Mr. Ferrel's computation gives "28° 30' for 

 the polar distance of the parallel where the surface of the fluid, or 

 the stratum of equal pressure, meets the surface of the earth."* It 

 will be readily seen that at that distance the south pole is entirely 

 surrounded by water, and it seems, therefore, highly probable, that 

 if there were no land in the frigid zones, there would be no definite 

 magnetic poles, but a simple polar belt towards which the magnetic 

 currents would flow in parallel spirals, f except when deflected by 

 land radiation, or tidal or other gravitation currents. 



II. Dip. 



8. The lines of equal dip are arranged in approximate parallels, 

 around the two (principal) magnetic poles. 



9. In consequence of this parallelism, they are convex towards 

 the north in the Pacific Ocean, and towards the south in the Atlan- 

 tic Ocean. 



10. The magnetic parallels also approximate to the isothermal 

 parallels, both in direction and in position, but with some important 

 departures. 



11. In South America, the magnetic equator is depressed nearly 

 30° south of the isothermal equator; it is, however, nearly equi- 

 distant from the (principal) north and south magnetic poles. 



1'2. The magnetic parallels near the magnetic poles, are more con- 

 vex than the isothermal lines, but they present some interesting in- 

 stances of parallelism to the ocean currents, which are indisputably 

 gravitation currents. 



13. This parallelism is specially observable in the regions of the 

 equatorial currents, the Gulf Stream, and the North Pacific and 

 Japan currents. 



14. If a series of waves were propagated through the air, from 

 the magnetic poles to the equator, with slight deflections by the con- 

 tinental contours and ocean currents, they could be represented with 

 great accuracy by the magnetic parallels. 



All of the foregoing njodifications, as well as the theoretical phe- 

 nomena previously referred to, have their exact counterparts in the 



* Loc. citat. 



f This inference accords with Barlow's conclusion that every place has 

 its particular polarizing axis. 



