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



85 



1911 during the "Deutschland" Expedition [Brennecke, 

 1921]. The temperatures observed in the upper 400 metres 

 are recorded in the following table, which gives also the 

 mean variations (in brackets). 



It is probable that the effective vertical convection in 

 these low latitudes does not go deeper than 80—90 metres 

 and in places not even as deep as this. The highest 

 numerical values of the mean variation are found here 

 at a greater depth than farther north. 



In coastal waters and in areas where Arctic water 

 covers the Atlantic water, the vertical variations of salinity, 

 and consequently the stability, may be so great that the 

 winter convection is limited to a relatively thin layer. In 

 such areas the vertical variations of temperature may be 

 very great in summer. Our stations on the Newfoundland 

 Bank show that the vertical variations of salinity are small 

 in the surface layer down to 10—25 metres, and very great 

 below this layer. We find the following mean variations 

 of temperature at the "Michael Sars" stations at the New- 

 foundland Bank and the Flemish Cap: 



with an inversion of temperature somewhere below 50 

 metres. 



The vertical gradient of temperature is, of course, 

 greater at some levels and may be much greater than 

 the mean variation. For instance, at Stat. 72, 30 metres 



10^" is about 



dz 



100. 



At all these stations there is a maximum vertical 

 variation of temperature between 25 and 50 metres, 



At Stats. 73—76 incl. the minimum temperature is 

 negative; the lowest value found is — 1-65° C. (at Stat. 

 74, 75 metres). The low temperatures are not a sure 

 indication of the effect of the winter cooling at the sur- 

 face in this region, but may quite as well be ascribed to 

 horizontal movements of the water layers (the Labrador 

 Current). The water masses between 25 metres and the 

 level of minimum temperature are far from being homo- 

 haline and do not show any sign of such mixing as the 

 convection in winter should create. — 



The author has calculated the difference of tempera- 

 ture between 200 and 400, 400 and 600 metres and so 

 forth' for intervals of 200 metres down to 1200 metres, for 

 all stations available from the North Atlantic. The results 

 may be summarized in the following way (numbers in 

 brackets indicating the mean variation according to the 

 definition given above): 



The vertical .variations of temperature are relatively 

 small at all depths below 200 metres in the Arctic or Sub- 

 arctic water in the northwestern area, S. of Greenland. 

 In the Great Atlantic Current ("The Gulf Stream") the 

 vertical variations may be very great. Apart from these 

 regions we find, for depths between 200 and 400 metres: 

 minima of temperature variation (0 to — 0-5) in the sea 

 off the Bay of Biscay and the British Isles, and in the ocean 

 E., N. and W. of Bermuda; maxima (— 2 to — 3) near the 

 equator, in the ocean N. and E. of the Antilles and also 

 in some isolated areas S. of 25° N. For 400 to 600 

 metres we have: a minimum (0 to — 0-5) in the eastern 

 North Atlantic off north-western Morocco and northwards; 

 a maximum (about — 2) N. and E. of the Antilles. For 

 600 to 800 metres we have: a minimum from Morocco 

 and Madeira northwards, sometimes with an inversion of 

 temperature (the mean variation positive, maximum about 

 4- 0-2) due to water from the Mediterranean; another 

 minimum is found near the equator; and a maximum 

 (about — 2) in the sea around Bermuda. For 800 to 1000 

 metres we have: a minimum in the eastern North Atlantic 

 N. of Morocco and Madeira (sometimes with inversion) 

 and in the sea S. of 18° N.; a maximum (about — 2) W. 

 of Bermuda. For 1000 to 1200 metres we have: a mini- 

 mum in the tropics (0 to — 0-5) and in the easternmost 

 part of the ocean as far north as about 43° N. (+ 0-3 to 

 - 0-5); and a maximum ( 1 to — 1-5) near Bermuda. 



