158 



NATURE 



[Dec. 24, 1874 



by late mails, it appears that the distribution of temperature in 

 the ocean is occupying the attention of a certain portion of the 

 scientific public, and even giving rise to considerable discussion. 

 The observations made on board this ship, and more especially 

 in the Atlantic, have furnished the greater part of the material 

 on which the various speculations have been founded. It appears 

 to me that one point suggested by these observations has not 

 received sufficient attention from those who have written and 

 spoken on the subject — I mean the effect of the changing seasons 

 on sea-water. Consider the state of the water at and near the 

 surface of the ocean, somewhere not in the tropics. To be more 

 precise, let us suppose that we have taken up our position in the 

 middle of the North Atlantic, somewhere about tlie 30th 

 parallel. This part of the ocean is not vexed with currents, and 

 affords the best possible field for the observation of the pheno- 

 menon in question. The Avhole ocean, enclosed by the 20th 

 and 40lh parallels of north latitude, and the meridians of 30" 

 and 60° west longitude, forms one oceanic lake, not affected by 

 the perturbing influence of currents or ol land ; and where, 

 therefore, the true effect of differences of atmospheric tempera- 

 ture on the waters of the ocean may be most advantageously 

 studied. Let us assume the winter temperature of the surface- 

 water to be 60° F. and the summer temperature to be 70" F. If 

 we start from midwinter, we find that, as summer approaches, 

 the surface water must get gradually warmer, and that the tem- 

 perature of the layers below the surface must decrease at a very 

 rapid rate, until the stratum of winter temperature, or 60° F., is 

 reached ; in the language of the isothermal charts, the isothermal 

 line for degrees between 70" F. {if we suppose that we have 

 arrived at midsummer) and 60° F. open out or increase their dis- 

 tance from each other as the depth increases. Let us now con- 

 sider the conditions after the summer heat has begun to waver. 

 During the whole period of heating, the water, from its increasing 

 temperature, has been always becoming lighter, so that heat 

 communication by convection with the \\ater below has been 

 entirely suspended during the wh^le period. The heating of the 

 surface water has, however, had another effect, besides increasing 

 its volume ; it has, by evaporation, rendered it denser than it 

 was before, at the same temperature. Keeping in view this 

 double effect of the summer heat upon the surface water, let us 

 consider the effect of tlie winter cold upon it. The superficial 

 water having assumed tlie atmospheric temperature of, say, 60° 

 F., will sink through the warmer water below it, until it reaches 

 the stratum of water having the same temperature as itself. 

 Arrived here, however, although it has the same temperature as 

 the surrounding water, the two are no longer in equilibrium, for 

 the water which has come from the surface has a greater density 

 than that below at the same temperature. It will therefore not 

 be arrested at tlie stratum of the same temperature, as would 

 have been the case with fresh water, but it will continue to sink, 

 carrying of course its higher temperature with it, and distributing 

 it among the lower layers of colder water. At the end of the 

 winter, therefore, and just before the summer heating recom- 

 mences, we shall have at the surface a more or less thick stratum 

 of water, having a nearly uniform temperature of 60° F., and 

 below this the temperature decreasing at a considerable, but less 

 rapid rate, than at the termination of the summer heating. If 

 we distinguish between surface -lOatc'r, the temperature of which 

 rises with the atmospheric temperature, following thus, in direc- 

 tion at least, the variation of the seasons, and sub-siii/act water, 

 or the stratum immediately below it, we have for the latter the 

 (at first sight) paradoxical effect of summer cooling and winter 

 heating. The effect of this agency is to diffuse the same heat to 

 a greater depth in the ocean, the greater the yearly range of 

 atmospheric temperature at the surface. This effect is well shown 

 in tlie chart of isothermals, on a vertical section between Madeira 

 and a position in lat. 3° S' N., long. 14° 49' W. The isothermal 

 line for 45° F. rises from a depth of 740 fathoms at Madeira, to 

 240 fathoms at the above-mentioned position.* In equatorial 

 regions there is hardly any variation in the surface-temperature 

 of the sea ; consequently, we find cold water very close to the 

 surface all along the line. On referring to the temperaiure sec- 

 tion between the position lat. 3" 8' N., long. 14° 49' W., and 

 St. Paul's rocks, it will be seen that, with a sui face-temperature 

 of from 75° F. to 79° F., waier at 55' F. is reached at disiances 

 of less than too fathoms from the surface. Midway between the 

 Azores and Bermuda, with a surface-temperature of 70° F., 



* There will, I think, be no violence in assuming an acquaintance with 

 these charts, at least among the scientific pubhc, as they have lately formed 

 the subject of lectures by Dr. Carpenter, and will no doubt have been pub- 

 lished before this reaches England. 



it is only at a depth of 400 fathoms that we reach water of 

 55° F. 



The above theory of vertical diffusion of temperature in the 

 ocean, owing to convection brought about by the yearly range of 

 temperature at the surface, presupposes that (at least in regions 

 where the range is considerable, and where the great vertical 

 diffusion of heat in question is oberved) the slightly concentrated 

 water, descending from the surface as the winter approaches, 

 does not meet water of greater density at the same temperature 

 than its own. Unfortunately the determination of the specific 

 gravity of water below the surface is much less simple than that 

 of the temperature. For although we have an instrument which 

 gives, within any required degree of accuracy, the density of the 

 water at any depth in exactly the same way as the thermometer 

 gives its temperature, the results of the observations are com- 

 posed of three factors, which depend on the temperature, the 

 pressure, and the salinity. By sending down a thermometer 

 along with it we might clear the result for temperature ; by 

 noting the depth we might clear for pressure ; but tlie result so 

 cleared would not'represent the salinity of the water at the depth 

 in question, but the average excess of salinity of the column of 

 water above it, over or under the mean salinity assumed for sea- 

 water, in the calculation of the pressure exercised by a column 

 of it. There remains, therefore, nothing for it but to fetch a 

 sample of water from each depth, and determine its specific 

 gravity on board. As this is an operation which takes up some 

 time, the number of " serial specific gravity" determinations is 

 comparatively small. 



The following are the results of two which were obtained on 

 the voyage between Bermuda and the Azores. The results show 

 the specific gravity at 60° F., that of water at 39° '2 F. being 

 taken as unity. 



I. was taken on June 18, 1S73, in lit. 35° 7' N., long 

 52° 32' W. 



II. was taken on June 24, 1873, in lat. 38° 3' N., long. 

 39" 19' W. 



For comparison I give one equatorial and one South Atlantic 

 "serial specific gravity" determination. 



III. was taken on Aug. 21, 1873, in lat. 3° 8' N., long. 14° 

 49' W. 



IV. was taken on Oct. 3, 1S73, in lat.126' 15' S., long. 32 

 56' W. 



From the figures in the Table it will be seen that in that pirt 

 of the ocean the specific gravity of the water in summer decreases 

 from the surface downwards. As a rule it attains an inferior 

 limit at a depth of from 400 to 500 fathoms, which it preserves 

 to the bottom. In those latitudes, therefore, the stratum of 

 intermixture extends down to 500 fathoms ; and this may be 

 said also to be the depth to which the sun's influence at the sur- 

 face penetrates. The results in column III. show the curious 

 phenomenon of the surface water being specifically Hghter than 

 any water below it, and that under an equatorial sun. The 

 position of this soumiing was peculiar, inasmuch as it was within 

 line of separation between the Guinea and the equatorial cur- 

 rents. All along the equatorial section the water at 50 and too 

 fathoms was found to be specifically heavier than cither at the 

 surface or that at greater depths. All along the equator, how- 

 ever, a current runs with great velocity ; and I have invariably 

 observed that strong surface-currents introduce considerable irre- 

 gularities into the specific gravity of the water near the surface. 

 The effect of the greater specific gravity at 100 fathoms conspires, 

 of course, within the small yearly range of temperature, in pre- 



