of Heat over Islands. Ji43 



and other properties,- greatly modifies the distribution of the 

 warmth which its particles receive from the sun. The surfaces 

 of the seas and oceans are traversed by cm-rents, whose tempe- 

 ratures influence the climate of any land which they approach, 

 and whose directions depend on varied and complicated condi- 

 tions. The resulting effect upon the temperature of the land 

 must correspond in some measure to such complications. 



But even if the ocean were free from currents, it might still 

 influence the climate of the land if it possessed a different mean 

 temperature. That it does actually possess a higher mean tem- 

 perature at its surface than the mean temperature of the air over 

 the land, seems to have been definitely proved within a very 

 recent period. The facts adduced by Admiral Duperrey, by 

 which he was led to infer that the mean temperature of the 

 liquid coating which surrounds three-fourths of our planet is 

 higher than that of even the lower strata of its gaseous envelope, 

 continue to receive additional confirmation in proportion to the 

 number, sagacity, and activity of observers. Lieutenant Maurjf, 

 whose labours have so greatly enlarged our knowledge of the 

 physical conditions of the ocean, has especially contributed to 

 establish the truth of the above conclusion. From whatever 

 cause this superiority of oceanic temperature may arise, it im- 

 parts additional interest to the problem of the influence of the 

 sea on the climate of the land. 



This question has been already treated by Humboldt in his 

 essay on the causes of the inflexions of isothermal lines*; but 

 here I propose to examine it in a more general manner, and by 

 following an order precisely the reverse of that which he has 

 adopted. 



In this way I have succeeded in establishing a general law 

 relating to the distribution of isothermal lines, which does not 

 appear to have been previously noticed, and which comprehends 

 as particular cases such as had been already observed. 



Let us conceive an island situated in either hemisphere of the 

 globe, and let it be completely surrounded with water possessing 

 nearly the same temperature all around the coast. The tempe- 

 rature of any place on the island will depend upon constant and 

 fluctuating causes. The former are the temperature of space, 

 and the extremely small but steady flux of heat from the interior 

 of the earth tlirough its outer crust. The latter are the heat it 

 directly gains from the sun, what it loses by radiation, what it 

 receives from warm and loses from cold currents of the atmo- 

 sphere, what it obtains by the condensation of moisture and 

 gives back by evaporation. 



The four last sources of gain and loss are manifestly connected 

 * Fragments /Isititiques, vol. ii. p. 397. 

 il 2 



