352 THE PHYSICAL CONDITION OF THE OCEAN. 



in the waters outside the submarine rid yje that forms tlie iudosing walls, 

 separating them from deejier areas beyond, and is, in all cases that 

 have been observed, equal to that on the ridge. From this fact we are 

 enabled to sui)plement our imperfect knowledge of depths, because if 

 in a certain i^art of an ocean we And that the temperature at great 

 depths is higher than we know exists at similar depths in waters api^ar- 

 eiitly connected, we can feel certain that there is a submarine ridge 

 which cuts off the bottom waters from moving along, and that the 

 depth on this ridge is that at which is found the corresponding tem 

 perature in the outer waters. As a corollary we also assume that the 

 movement of water at great depths is confined to an almost impercep- 

 tible movement, for if there was a motion that we could term, in the 

 ordinary accej)tation of the word, a current, it would infallibly surmount 

 a ridge and pour over the other side, carrying its lower temperature 

 with it. A notable instance is the bottom temperature of the North 

 Atlantic. This is nowhere below 35° F., althougli the depths are very 

 great. But in the South Atlantic at a depth of only 2,800 fathoms the 

 bottom temperature is but a little above 32° F., and we are therefore 

 convinced that somewhere between Africa and South America, though 

 soundings do not yet show it, there must be a ridge at a depth of about 

 2,000 fathoms. We also come to the same conclusion with regard to 

 the eastern and western portions of the South Atlantic, wliere similar 

 differences prevail. Again, the few temperatures that have been 

 obtained in the eastern South Pacific show a considerable difference 

 from those in the South x\tlantic, and we are compelled to assume a 

 ridge from the Falkland Islands to the Antarctic continent. 



It is interesting that the investigation into the translation of the 

 great seismic wave caused by the eruption of Krakatoa in 1883 led to a 

 similar and entirely independent conclusion. The Avave caused by the 

 explosion in the Straits of Sunda reached Cape Horn, where by good 

 chance a French meteorological expedition liad erected an automatic 

 tide gauge, but instead of one series of waves being marked on the 

 paper there were two. A little consideration showed that the South 

 Pole having directly interposed between Sunda Straits and Cape Horn, 

 the waves diverted l)y the land about the pole would arrive from both 

 sides. One wave, however, made its appearance seven hours before 

 the other. Study showed that the earliest wave coincided in time with 

 a wave traveling on the Pacific side of the pole, with a velocity due to 

 the known depth, while the later wave must have been retarded in its 

 journey via the South Atlantic. The only possible explanation is that 

 the wave had been imjieded by comparatively shallow water. The evi- 

 dence from bottom temperature was then unknown, and thns does one 

 branch of investigation aid another. 



In the Western Pacific the water is colder, a few bottom tempera- 

 tures of a little over 33° F. having been found in the deep trongh east 

 of the Tonga Islands; but the North Pacific, though the deejier ocean — 



