It is a condition absolutely essential to the gravitation theory that the surface of the ocean should be highest in equatorial regions and slope downwards to either pole. Were water absolutely frictionless, an incline, however small, would be sufficient to produce a surface-flow from the equator to the poles, but to induce such an effect some slope there must be, or gravitation could exercise no power in drawing the surface-water polewards.

The researches of the Challenger Expedition bring to light the striking and important fact that the general surface of the North Atlantic in order to produce equilibrium must stand at a higher level than at the equator. In other words the surface of the Atlantic is lowest at the equator, and rises with a gentle slope to well-nigh the latitude of England. If this be the case, then it is mechanically impossible that, as far as the North Atlantic is concerned, there can be any such general movement as Dr. Carpenter believes. Gravitation can no more cause the surface-water of the Atlantic to flow towards the arctic regions than it can compel the waters of the Gulf of Mexico up the Mississippi into the Missouri. The impossibility is equally great in both cases.

In order to prove what has been stated, let us take a section of the mid-Atlantic, north and south, across the equator; and, to give the gravitation theory every advantage, let us select that particular section adopted by Dr. Carpenter as the one of all others most favourable to his theory, viz., Section marked No. VIII. in his memoir lately read before the Royal Geographical Society.[98]

The fact that the polar cold water comes so near the surface at the equator is regarded by Dr. Carpenter as evidence in favour of the gravitation theory. On first looking at Dr. Carpenter’s section it forcibly struck me that if it was accurately drawn, the ocean to be in equilibrium would require to stand at a higher level in the North Atlantic than at the equator. In order, therefore, to determine whether this is the case or not I asked the hydrographer of the Admiralty to favour me with the temperature soundings indicated in the section, a favour which was most obligingly granted. The following are the temperature soundings at the three stations A, B, and C. The temperature of C are the mean of six soundings taken along near the equator:—

Depth in Fathoms.Lat. 37° 54′ N.
Long. 41° 44′ W.		Lat. 23° 10′ N.
Long. 38° 42′ W.		Mean of six temperature soundings near equator.
Temperature.Temperature.Depth in Fathoms.Temperature.

°

°

°

Surface.

70·0

72·0

Surface.

77·9

100

63·5

67·0

10

77·2

200

60·6

57·6

20

77·1

300

60·0

52·5

30

76·9

400

54·8

47·7

40

71·7

500

46·7

43·7

50

64·0

600

41·6

41·7

60

60·4

700

40·6

40·6

70

59·4

800

38·1

39·4

80

58·0

900

37·8

39·2

90

58·0

1000

37·9

38·3

100

55·6

1100

37·1

38·0

150

51·0

1200

37·1

37·6

200

46·6

1300

37·2

36·7

300

42·2

1400

37·1

36·9

400

40·3

1500

..

36·7

500

38·9

2700

35·2

..

600

39·2

2720

..

35·4

700

39·0

800

39·1

900

38·2

1000

36·9

1100

37·6

1200

36·7

1300

35·8

1400

36·4

1500

36·1

Bottom.

34·7

On computing the extent to which the three columns A, B, and C are each expanded by heat according to Muncke’s table of the expansion of sea water for every degree Fahrenheit, I found that column B, in order to be in equilibrium with C (the equatorial column), would require to have its surface standing fully 2 feet 6 inches above the level of column C, and column A fully 3 feet 6 inches above that column. In short, it is evident that there must be a gradual rise from the equator to latitude 38° N. of 3½ feet. Any one can verify the accuracy of these results by making the necessary computations for himself.[99]

PLATE III

W. & A. K. Johnston, Edinbr. and London.

SECTION OF THE ATLANTIC nearly North and South, between LAT. 38° N. & LAT. 38° S.