the sections rightly it must be borne in mind that the vertical scale is 2,000 times greater than the horizontal.
Many of the conditions we have already mentioned are clearly apparent in the sections: the small variations between the surface and a depth of about 100
metres at each station; the decrease of temperature and salinity as the depth increases; the high values both of temperature and salinity in the western part as
compared with the eastern. We see from the sections how nearly the isotherms and isohalins follow each other. Thus, where the temperature is 12º C., the water almost invariably has a salinity very near 35 per mille. This water at 12º C., with a salinity of 35 per mille, is found in the western part of the area (in the Brazil Current) at a depth of 500 to 600 metres, but in the eastern part (in the Benguela Current) no deeper than 200 to 250 metres (109 to 136 fathoms).
We see further in both sections, and especially in the southern one, that the isotherms and isohalins often have an undulating course, since the conditions at one station may be different from those at the neighbouring stations. To point to one or two examples: at Station 19 the water a few hundred metres down was comparatively warm; it was, for instance, 12º C. at about 470 metres (256
fathoms) at this station; while the same temperature was found at about 340
metres (185 fathoms) at both the neighbouring stations, 18 and 20. At Station 2 it was relatively cold, as cold as it was a few hundred metres deeper down at Stations 1 and 3.
These undulating curves of the isotherms and isohalins are familiar to us in the Norwegian Sea, where they have been shown in most sections taken in recent years. They may be explained in more than one way. They may be due to actual waves, which are transmitted through the central waters of the sea. Many things go to show that such waves may actually occur far below the surface, in which case they must attain great dimensions; they must, indeed, be more than 100
metres high at times, and yet -- fortunately -- they are not felt on the surface. In the Norwegian Sea we have frequently found these wave-like rises and falls. Or the curves may be due to differences in the rapidity and direction of the currents.
Here the earth's rotation comes into play, since, as mentioned above, it causes zones of water to be depressed on one side and raised on the other; and the degree of force with which this takes place is dependent on the rapidity of the current and on the geographical latitude. The effect is slight in the tropics, but great in high latitudes. This, so far as it goes, agrees with the
[Fig. 11 and captions]
fact that the curves of the isotherms and isohalins are more marked in the more southerly of our two sections than in the more northerly one, which lies 10 or 15
degrees nearer the Equator.
But the probability is that the curves are due to the formation of eddies in the
currents. In an eddy the light and warm water will be depressed to greater depths if the eddy goes contrary to the hands of a clock and is situated in the southern hemisphere. We appear to have such an eddy around Station 19, for example.
Around Station 2 an eddy appears to be going the other way; that is, the same way as the hands of a clock. On the chart of currents we have indicated some of these eddies from the observations of the distribution of salinity and temperature made by the Fram Expedition.
While this, then, is the probable explanation of the irregularities shown by the lines of the sections, it is not impossible that they may be due to other conditions, such as, for instance, the submarine waves alluded to above. Another possibility is that they may be a consequence of variations in the rapidity of the current, produced, for instance, by wind. The periodical variations caused by the tides will hardly be an adequate explanation of what happens here, although during Murray and Hjort's Atlantic Expedition in the Michael Sars (in 1910), and recently during Nansen's voyage to the Arctic Ocean in the Veslemöy (in 1912), the existence of tidal currents in the open ocean was proved. It may be hoped that the further examination of the Fram material will make these matters clearer.
But however this may be, it is interesting to establish the fact that in so great and deep an ocean as the South Atlantic very considerable variations of this kind may occur between points which lie near together and in the same current.
As we have already mentioned in passing, the observations show that the same temperatures and salinities as are found at the surface are continued downward almost unchanged to a depth of between 75 and 150 metres; on an average it is about 100 metres. This is a typical winter condition, and is due to the vertical circulation already mentioned, which is caused by the surface water being cooled in winter, thus becoming heavier than the water below, so that it must sink and give place to lighter water which rises. In this way the upper zones of water become mixed, and acquire almost equal temperatures and salinities. It thus appears that the vertical currents reached a depth of about 100 metres in July, 1911, in the central part of the South Atlantic. This cooling of the water is a gain to the air, and what happens is that not only the surface gives off warmth to the air, but also the sub-surface waters, to as great a depth as is reached by the vertical circulation. This makes it a question of enormous values.
This state of things is clearly apparent in the sections, where the isotherms and isohalins run vertically for some way below the surface. It is also clearly seen when we draw the curves of distribution of salinity and temperature at the
different stations, as we have done in the two diagrams for Stations 32 and 60
(Fig. 9). The temperatures had fallen several degrees at the surface at the time the Fram's investigations were made. And if we are to judge from the general appearance of the station curves, and from the form they usually assume in summer in these regions, we shall arrive at the conclusion that the whole volume of water from the surface down to a depth of 100 metres must be cooled on an average about 2º C.
As already pointed out, a simple calculation gives the following: if a cubic metre of water is cooled 1º C., and the whole quantity of warmth thus taken from the water is given to the air, it will be sufficient to warm more than 3,000 cubic metres of air 1º C. A few figures will give an impression of what this means. The region lying between lats. 15º and 35º S. and between South America and Africa
-- roughly speaking, the region investigated by the Fram Expedition -- has an area of 13,000,000 square kilometres. We may now assume that this part of the ocean gave off so much warmth to the air that a zone of water 100 metres in depth was thereby cooled on an average 2º C. This zone of water weighs about 1.5 trillion kilogrammes, and the quantity of warmth given off thus corresponds to about 2.5 trillion great calories.
It has been calculated that the whole atmosphere of the earth weighs 5.27 trillion kilogrammes, and it will require something over 1 trillion great calories to warm the whole of this mass of air 1ºC. From this it follows that the quantity of warmth which, according to our calculation, is given off to the air from that part of the South Atlantic lying between lats. 15º and 35º S., will be sufficient to warm the whole atmosphere of the earth about 2º C., and this is only a comparatively small part of the ocean. These figures give one a powerful impression of the important part played by the sea in relation to the air. The sea stores up warmth when it absorbs the rays of the sun; it gives off warmth again when the cold season comes. We may compare it with earthenware stoves, which continue to warm our rooms long after the fire in them has gone out. In a similar way the sea keeps the earth warm long after summer has gone and the sun's rays have lost their power.
Now it is a familiar fact that the average temperature of the air for the whole year is a little lower than that of the sea; in winter it is, as a rule, considerably lower. The sea endeavours to raise the temperature of the air; therefore, the warmer the sea is, the higher the temperature of the air will rise. It is not surprising, then, that after several years' investigations in the Norwegian Sea we
have found that the winter in Northern Europe is milder than usual when the water of the Norwegian Sea contains more than the average amount of warmth.
This is perfectly natural. But we ought now to be able to go a step farther and say beforehand whether the winter air will be warmer or colder than the normal after determining the amount of warmth in the sea.
It has thus been shown that the amount of warmth in that part of the ocean which we call the Norwegian Sea varies from year to year. It was shown by the Atlantic Expedition of the Michael Sars in 1910 that the central part of the North Atlantic was considerably colder in 1910 than in 1873, when the Challenger Expedition made investigations there; but the temperatures in 1910
[Fig. 13]
Fig. 13. -- Temperatures at one of the "Fram's" and one of the "Challenger's"
Stations, to the South of the South Equatorial Current were about the same as those of 1876, when the Challenger was on her way back to England.
We can now make similar comparisons as regards the South Atlantic. In 1876
the Challenger took a number of stations in about the same region as was investigated by the Fram. The Challenger's Station 339 at the end of March, 1876, lies near the point where the Fram's Station 44 was taken at the beginning of August, 1911. Both these stations lay in about lat. 17.5º S., approximately half-way between Africa and South America -- that is, in the region where a relatively slack current runs westward, to the south of the South Equatorial Current. We can note the difference in Fig. 13, which shows the distribution of temperature at the two stations. The Challenger's station was taken during the autumn and the Fram's during the winter. It was therefore over 3º C. warmer at the surface in March, 1876, than in August, 1911. The curve for the Challenger station shows the usual distribution of temperature immediately below the surface in summer; the temperature falls constantly from the surface downward.
At the Fram's station we see the typical winter conditions; we there find the same temperature from the surface to a depth of 100 metres, on account of cooling and vertical circulation. In summer, at the beginning of the year 1911, the temperature curve for the Fram's station would have taken about the same form as the other curve; but it would have shown higher temperatures, as it does in the deeper zones, from 100 metres down to about 500 metres. For we see that in these zones it was throughout 1º C. or so warmer in 1911 than in 1876; that is to say, there was a much greater store of warmth in this part of the ocean in 1911
than in 1876. May not the result of this have been that the air in this region, and also in the east of South America and the west of Africa, was warmer during the winter of 1911 than during that of 1876? We have not sufficient data to be able to say with certainty whether this difference in the amount of warmth in the two years applied generally to the whole ocean, or only to that part which surrounds the position of the station; but if it was general, we ought probably to be able to find a corresponding difference in the climate of the neighbouring regions.
Between 500 and 800 metres (272 and 486 fathoms) the temperatures were exactly the same in both years, and at 900 and 1,000 metres (490 and 545
fathoms) there was only a difference of two or three tenths of a degree. In these deeper parts of the ocean the conditions are probably very similar; we have there no variations worth mentioning, because the warming of the surface and sub-surface waters by the sun has no effect there, unless, indeed, the currents at these depths may vary so
[Fig. 14]
Fig. 14. -- Temperatures at one of the "Fram's" and one of the "Valdivia's"
