222 OBSERVATIONS ON THE MEAN TEMPEKATURE 



even of the tropical regions. Rising in the parallel of 60 ''j 

 and sweeping over the whole continent of North America, 

 through an extent of 2600 miles, they retain their gelid in- 

 fluence even in the latitude of 18". Can we suppose, then, that 

 such winds as these cease all at once to impart their cooling- 

 energies to more southern zones. Acting with such power in 

 the parallel of 18°, will they not refresh the opposite shores of 

 the Pacific Ocean, and temper even the burning heats of the 

 Equinoctial Line? Whatever law of progression we may 

 adopt to represent the decreasing influence of these northern 

 currents, in their passage towards the Line, there is none 

 which allows them the influence described by Humboldt on 

 the coast of Tabasco, that will not extend that influence to the 

 Equator itself. 



Although the preceding views, respecting the distribution 

 of heat within the Polar Circle, make the temperature much 

 lower than had previously been supposed, yet when taken in 

 conjunction with the results of the expedition under Captain 

 Pauky, they rather encourage the hopes which have been so 

 reasonably entertained, of reaching the Pole itself. 



Upon the supposition that there are two Isothermal Centres 

 in 80° of latitude, and that their temperature is —3} of Fahren- 

 lieit, the mean temperature of the Pole of the Globe will be about 

 11", incomparably warmer than the regions in wliich Captain 

 Parry spent the winter. If an expedition, therefore, were to 

 set out for Spitzbergen, and remain there for one or more sea- 

 sons, till an opening should be found through the icy barrier 

 which stretches from that island to the Greenland coast, there 

 is every reason to believe that it would ultimately succeed. 



If the Pole is placed in an open sea, the difficulty of reach- 

 ing it entirely ceases ; and if it forms part of a frozen conti- 

 nent, those intrepid individuals who sustained the rigorous 



cold 



