July 9, 1896] 



XA TURE 



235 



llie depth of the wave jialh. ShoviUl this be so, the next sug- 

 ^C'tion is, that as a wave proceeds dt^wnwards refraction may 

 take place, and that a focal concenlration of energy may be 

 found at the antipodes of a seismic centre. 



From these remarks it is clear that amongst the most important 

 work with which the seismologist has now before him, is to 

 measure the speed at which the preliminary vibrations of an 

 earthquake are transmitted, and because this is high, the defini- 

 tion on the recording surface must be clear, the rate at which 

 this is moved must be such that time intervals may be measured 

 to within 10 seconds, and the observing stations, if they are 

 limited in number, should be widely separated. 



In the choice of stations, at all of which there must be the 

 means of keeping fairly accurate time, the plan originally 

 -suggested by the author was to choose these relatively to districts 

 where large earthquakes are frequent. The districts selected 

 were the South American coast, Japan, and the Philippines, 

 Himalaya and Central Asia. By a system of trials it was found 

 that fifteen stations could be chosen, nearly all of which happen 

 to be in the United States or British colonics, about ten of 

 which would form a seiies appro.\iuialely 2000 miles and 

 multiples of 2000 miles distant from any of the three districts. 

 With a series of this description, data of a fairly complete 

 nature respecting the rate at which motion may be transmitted 

 round and, possibly, through our earth at varying depth should 

 be obtainable. Any addition to this series would naturally 

 render our information more certain, and add to the value of 

 records obtained from centres other than those specified. 



The cost of installation at each observatory would be approxi- 

 mately £,y>. 



The proposal here made is similar to one published by the 

 writer in January 1S95, and does not materially dift'er from the 

 one put forward by that distinguished investigator, the late Dr. 

 E. von Rebeur-l'aschwitz, and now being so warndy advocated 

 by Dr. G. Gerland of Strassburg (see pp. 135, 1^6). 



J. ^r. 



THE SPECIFIC GRAVirV OF THE WATERS 

 OF THE SEA. 



7 \ continuation of his paper on oceanic circulation, in the con- 

 cluding volume of the Challenger Reports, which chiefly 

 dealt with the distribution of teni|)erature. Dr. Buchan has 

 published in the Traiisactioiis of the Royal Society of Edinburgh 

 a series of maps showing, so far as the present state of knowledge 

 permits, the specific gravity of the waters of the great oceans at 

 various de|)ths ; and accompanying the maps is an extended dis- 

 cussion of some of the points treated in the previous memoir. 



In the paper just published. Dr. Buchan has departed from the 

 mode of representing salinities and specific gravities employed in 

 the Challenger Report, and instead of charting the actual values, 

 gives the departures above and below an average assumed to be 

 a mean for all the oceans. It is difficult to see that anything is 

 gained by this method ; and it has the undoubted disadvantage 

 that any future change in the assumed means will involve the 

 reconstruction of all the maps. Even at the surfiice there are 

 considerable portions of the sea of which we can only guess at 

 the -mean temperature and salinity, and the general average 

 given by Dr. Buchan may therefore undergo modification, not- 

 withstanding the attempts to apply a process of integration. 

 Below the surface, the general average is simply the mean of 

 existing observations ; and while an inspection of the map shows 

 that these are by no means perfect, the fact that there is only a 

 single line of observations in the North Pacific, one in the 

 Southern Ocean, none in the .\llantic north of 40', and none in 

 the Indian Ocean, indicates that the general a%erages must be mere 

 approximations. Another unsatisfactory efl'ect of the adoption of 

 this method is due to the fact that values above and below the 

 general average are thrown into strong contrast by being printed 

 in difterent colours on the maps, thereby frequently exaggerat- 

 ing their apparent difi'erence. In the case of the North Pacific, 

 for example. Dr. Buchan lays great stress on the low specific 

 gravity of the waters of this ocean at all depths. Undoubtedly 

 the observations show that they are lighter than the .\tlantic by 

 a quantity amounting below the surface to about ooooS; but the 

 fact remains that a change of, say, '0003 in the mean for the 

 globe at 300 fathoms would throw the whole of the North 

 Pacific above the average, while the observations within that 

 area themselves show inconsistencies amounting to double thai 



NO. 1393, VOL. 54] 



quantity. We draw special attention to this poirt, because it 

 seems to lie at _the root of a certain weakness in the line of 

 argument taken up by Dr. Buchan, leading to a confusion of 

 what we inay call the statical and dynamical problems of ocean 

 circulation, somewhat analogous to that involved in Ferrel's 

 theory of cyclones. In drawing up any general scheme of the 

 movements of oce.anic waters, it is necessary to keep clearly in 

 mind certain "conditions of continuity"; if the surface salinity 

 is anywhere reduced by copious rainfall, it must somewhere else 

 be correspondingly increased by evaporation ; if reduced by 

 melting ol field ice, a corresponding quantity of salt must have 

 been added to the deeper waters where the ice was formed ; 

 if up-welling is produced by an off-shore wind, the same force 

 must be competent to cause a down-draught somewhere else. 

 These considerations seem to suggest several .simplifications in 

 the scheme of circulation proposed by Dr. Buchan. 



Comparing the Atlantic and Pacific Oceans, we find in the 

 former a limited area subject to atmospheric systems of con- 

 siderable intensity, the air over a considerable proportion of the 

 surface being relatively dry. Over the Atlantic there is accord- 

 ingly a relatively great amount of evaporation, producing high 

 surface salinity, and the water carried ofi' is distributed over an 

 area nearly four times as large as in the case of the Pacific, 

 allowing of its gradual return. In the Pacific, on the other 

 hand, the winds are not so strong, the rate of evaporation is 

 slower, and the redistribution of the moisture more local. It 

 may therefore be possible to account for a considerable part of ■ 

 the low salinity of the Pacific without assuming that the high 

 rainfall of the East Indian region produces effects so markedly 

 in excess of those due to the immense discharges of fresh water 

 into the Atlantic or the Indian Ocean. Dr. Schott's observa- 

 tions in East Indian waters support this view, indicating that 

 the great Ireshening due to heavy rainfall is here, as elsewhere, 

 largely restricted to the immediate neighbourhood of the land. 



As an example of the converse of the foregoing, we may take 

 the case of the undercurrent flowing from the Mediterranean to 

 the Atlantic through the Stiaits of Gibraltar. From the observa- 

 tions of temperature and salinity Dr. Buchan regards it as placed 

 "beyond dispute " that the warming effect of this outflowing 

 water " becomes strikingly apparent at about 500 fathoms," and 

 " beyond this depth its influence is felt over nearly the whole 

 breadth of the Atlantic to at least about 1000 fathoms." Now 

 at the Straits of Gibraltar the depth is something under 200 

 fathoms, and the extreme width at the surface a little greater 

 than the Straits of Dover ; and it is known that the loss by 

 evaporation from the surface of the Mediterranean is not nearly 

 compensated for by the fresh water additions from the rainfall 

 and the rivers which empty themselves into its basin. The 

 amount of water issuing into the Atlantic must therefore be 

 greatly less than the amount entering the Mediterranean, and a 

 comparison of the volumes and temperatures of the two bodies 

 of water shows that it is almost impossible to give the outflow 

 from the Mediterranean credit for such widespread action. 



The two cases we have quoted, perhaps the strongest of several 

 suggested by Dr. Buchan's papers, seem to support the results 

 of a number of recent investigations, indicating that the efltct of 

 differences of specific gravity, while by no means a negligible 

 quantity, is in general small compared to the dynamical effects 

 due to the momentum of the surface currents, even at great 

 depths. 



From this point of view, we at once obtain help from the 

 researches of Pettersson and Kriimmel, noting specially their 

 results as to the tendency of surface currents to induce reaction 

 currents under them, and to divide on changing dir.ction, and 

 bearing in mind the deflecting influence of the earth's rotation at 

 all depths. In the Atlantic, the water driven northward alon"- 

 the American coast is blocked by the land, and is partly drained 

 oft' by the easterly drift currents, partly sent downwards in a 

 column separated into two parts, at least in certain seasons, by 

 a bulging out of the cold Labrador current. Crossing towards 

 the west coast of Europe, the easterly drift divides, a part 

 escaping northward under regulation of the polar streams from 

 the east of Greenland and Iceland, and a part banking up 

 against the French and Spanish coasts and the north-west of 

 Africa, as is shown by the "bottle charts" of the Prince of 

 Monaco and M. Hautreux. The shape of the coast prevents 

 all the water escaping laterally, and a part descends, carrying with 

 it the elilux from the .Mediterranean. 



In the Pacific the elTect is similar, subject to the difference 

 that while the circulation is less energetic, it is also less inter- 



