40 F. L. Ekman, 



According to Pouillet's researches on the amount of heat, radia- 

 ted from the sun '), compared with Meech's work '»On the relative inten- 

 sity of the heat and lii^ht of the sum), there must daily fall throughout the 

 whole of the torrid zone on an average 75.3 ^) kilo-units of heat on 

 every square decimetre of surface. Let us suppose that this quantity of 

 heat, deducting only one fourth for atmospheric absorption, i. e. 56.5 kilo- 

 units on the square decimetre, is taken up by the sea-water, and that 

 the quantity of heat contemporaneously communicated to the Arctic zone 

 is ^ 0; we have thus, with respect to the quantity of heat absorbed, 

 placed the calculation in a footing far too favourable to the former re- 

 gion. The loss of heat through evaporation and radiation must be im- 

 mensely greater in the Torrid than in the Arctic zone. But let us sup- 

 pose them equal, or, in other words, take no account of these losses of 

 heat in calculating the difference of level between the two zones which 

 the phœnomena of heat would produce. We may in this case be sure, 

 that we shall not find the difference of level, caused by heat, too little, 

 even though we should neglect the influence, which the atmosphere may 

 have on the sea's temperature, as also that arising from the conversion 

 of motion into heat; as regards this latter we have moreover as jet no 

 reason to suppose, that any greater difference of temperature between 

 different regions of the ocean would be caused by it. The maximum 

 difference of level, which the effect of temperature could produce in a 

 da}^ between the sea-surface in the Torrid and the Arctic zones, we 

 should therefore find by calculating the dilatation of the water, which an 

 accession of 56.5 kilo-units of heat on every square decimetre of the 

 surface would produce. If we for this purpose calculate the dilatation 

 of a column of water 56.4 decimetres high, when the temperature is rai- 

 sed from 24" Centg. to 25*^, we make the calculation in a manner too 

 favourable to the greatness of the dilatation. But the dilatation is ne- 

 vertheless only iVio millimètres. 



However imperfect the observations of evaporation may be, we 

 may nevertheless with certainty assume, that the change of level, pro- 

 duced by it in a day within the Torrid zone, is considerably more than 



') According to this determination every square centimetre of the earth's sur- 

 face wouW receive 231000 gramme-units of heat from the sun yearly, if there were 

 no heat-absorbing atmosphere and the heat, communicated to the earth, were uni- 

 formly distributed on the surface. 



-) ()n a narrow belt just at the equator itself the cipher is a trille larger and 

 goes up to 77.3. 



