ATLANT. DEEP-SEA EXPED. 1910. VOL. i] PHYSICAL OCEANOGRAPHY AND METEOROLOGY 



79 



bottom-water cominy in this way to the Atlantic Ocean 

 from the north, but we have no observations for studying 

 this question. That there must be a seasonal variation, 

 with a ma.ximum flow of bottom-water alter the winter 

 cooling, is a priori most probable. 



We know practically nothing of the movements of 

 the deep water. Quite generally we may infer that the 

 bottom-water in the greater part of the North Atlantic 

 moves with a component towards the south, but nothing 

 definitely can be stated with regard to velocities. We 

 possess too few trustworthy determinations of salinity (den- 

 sity) from the bottom-water, and lack vertical series of 

 observation from great depths. No dynamic calculations 

 of any value can be made. 



36. The Horizontal Distribution 

 of Temperature in the Troposphere. 



If there were no currents in the sea, isotherms repre- 

 senting the mean annual temperatures would go in an 

 almost east-west direction and be evenly distributed, with 

 some relatively small deviations due to variations along 

 the parallels in the balance oE gain and loss of heat. 

 The large and often irregular horizontal variations actu- 

 ally existing in the North Atlantic and in other oceans, 

 with isotherms going in many directions and with very 

 variable horizontal gradients of temperature, are mainly 

 due to currents. When we disregard the pure wind-cur- 

 rent at the surface, we may state as a general rule, that 

 the greater the horizontal variations of temperature are, 

 the stronger are the currents, and vice versa. We have, 

 then, to consider the variations not only at a certain level, 

 but at higher and deeper levels as well. The validity of 

 the rule is due to the fact that the variations in tempera- 

 ture coincide, on the whole, with variations in density. 



In places where the horizontal variations of tempera- 

 ture are relatively large, therefore, the transport of water 

 is also relatively large, with a quick conveyance of its 

 content of heat, as well as of its other constituents such 

 as salt, plankton organisms etc. In order to illustrate this 

 we may refer to the charts for 100, 400 and 600 metres 

 on p. 96*. They show a crowding of the isotherms especi- 

 ally in a belt which exactly corresponds to the site of the 

 Great Atlantic Current. In regions where the horizontal 

 gradient of temperature at all levels is relatively small. 

 the water moves but slowly. This connection between the 

 distribution of temperature and the currents has already 

 been mentioned above (pp. 21 and 36), and it will be 

 further explained in the chapter on the dynamics of the sea. 



According to a law of "parallel solenoids", we find 

 that the isotherms generally run parallel at the various 



depths, when we make proper allowance for the vertical 

 variations of temperature. This means that temperature 

 charts of different levels resemble each other wheti we 

 disregard the absolute values of temperature and the vari- 

 able number of isotherms. The resemblance is clearly seen 

 in the charts on p. 96 ■■ when we e.xcept the regions where 

 water from the Mediterranean appears and makes the condi- 

 tions so peculiar that the parallelism is blurred. We see, 

 for instance, that the isotherm for 8"" C. at 100 metres 

 has very nearly the same course as the isotherm for 6° C. 

 at 400 metres and for 5 C. at 600 metres. In the deeper 

 strata comparatively few isotherms appear, and the really 

 deep water is nearly homotherm in a horizontal direction 

 over wide stretches. 



The conditions mentioned here are evidently of con- 

 siderable significance for marine zoo-geography. It may 

 be assumed that the biological limits are nearly parallel 

 at different levels above the deep n'ater, though the com- 

 munities of animal organisms may vary from one level 

 to another. 



It has been observed that there is often a certain 

 resemblance between the horizontal distribution of tempe- 

 rature (and salinity) even at the surface of the sea and the 

 topography of the sea-bottom. This is a consequence of 

 the law of "parallel solenoids" just mentioned, in combi- 

 nation with another general law discovered by W. Ekman. 

 This law tells us that a gradient current which flows in 

 a direction where the depth to the bottom decreases, turns 

 cum sole (i. e. to the right in the northern hemisphere 

 and to the left in the southern). The current turns in the 

 opposite direction when the depth to the bottom increases. 

 This holds good even if the depths are very great. The 

 isotherms must, then, exhibit similar bends. This will 

 explain, for instance, the characteristic features of the 

 temperature charts in the vicinity of the Newfoundland 

 Banks. The "Gulf Stream" coming from the south-west 

 meets the slope S. of the Banks where the bottom-depth 

 decreases, and makes a turn to the right. Then the cur- 

 rent passes a locality with increasing depths, and turns to 

 the left. Later on, it turns again to the right, and finally 

 continues across the ocean. Our small-scale charts (p. 96*) 

 are not supposed to show the variations in detail. In fact, 

 these cannot be studied at present, e.xcept in a few places, 

 because the observations from most regions are inadequate. 

 We may, however, point out that the horizontal distribu- 

 tion of temperature must exhibit great local variations in 

 areas where the currents are fairly strong and the depth 

 to the bottom of the sea varies. The uneve/i/iess of the 

 sea-bed affects the distribution of temperature (and sali- 

 nity etc.) upwards through the water to the surface. 



The charts on p. 96* are rather schematic, even for 

 those parts of the eastern North Atlantic where relatively 



