1872.] 



' Shearwater ' Scientific Researches. 



575 



flood ; the excess being proportional to the amount of fresh water which 

 the upper part of the river brings down. And thus it happens that a 

 floating body thrown into such a river is at last carried out to sea, though 

 it may have been brought back by the tide twenty, fifty, or a hundred 

 times, each time stopping at a point a little further down than before. 

 Now there can be no doubt whatever, that, putting the action of Wind out 

 of the question, a vessel which enters the Western end of the Strait would 

 be gradually carried into the Mediterranean by the predominant easterly 

 movement of the upper current ; though its general easterly progress Would 

 be interrupted by a succession of returns to the westward, or, when there 

 might be no actual return, by periods of rest. And the evidence appears 

 to me just as conclusive, that if a body could be so weighted as to remain 

 freely suspended in the Mediterranean stratum off Gibraltar, and its move- 

 ments could be watched from above, we should find it in like manner 

 gradually working its way towards the opposite end of the Strait, and at 

 last clearing the "ridge" to descend along the Atlantic slope beyond. 

 This it could not do unless the Tidal action were supplemented by an out- 

 flow produced by some other agency ; and such an agency exists in the 

 excess of lateral pressure in the Mediterranean column above that of the 

 Atlantic column, which is proportional to the excess of the Specific Gravity 

 of Mediterranean water. — In the strong continuous Under-currents whose 

 existence in the Dardanelles and in the Baltic Sound has now been con- 

 clusively established, we see what larger differences in Specific Gravity, 

 uncomplicated by Tidal action, can effect. (See Appendix II. ; also 

 §134.) 



60. In one respect it appears to me that the Specific Gravity observa- 

 tions,— by which it can be certainly determined whether the water at any 

 given depth is Atlantic, Mediterranean, or a mixture of the two, — are 

 more satisfactory than those made by the Current-drag. For the actual 

 movement of the drag can only give the real rate of movement of the 

 stratum in which it is hanging, when that rate is the same as that of 

 the surface-stratum in which the suspending buoy is floating ; in which 

 case there will be no pull of the buoy upon the drag, or of the drag 

 upon the buoy (see Appendix II.). If the under-cwxxzxit be flowing faster 

 than the upper, the drag will be kept back by the buoy ; whilst if the sur- 

 /ace-current be flowing faster than the deeper stratum, the drag will be 

 pulled onwards by the buoy. If, again, the two strata be moving in opposite 

 directions, the reciprocal influence of the drag and the buoy will make the 

 observed rate still more different from the real rate, and perhaps will even 

 render both stationary. Hence the actual rate of the under-current can only 

 be estimated imm the observed rate of movement of the suspending buoy; and 

 the rule adopted by Capt. Nares in making such estimates was to halve the 

 difference between the observed rate of the suspending-buoy and that of the 

 surface-current, and to addt\\\s half to the former when the rate of the under- 

 current was the greater, deducting it when the rate of the surface-current was 

 the greater. Thus when the surface-current was running at 1 mile an hour, 



