90 W. Ferrel — Relative Tempemt teres of tin ILm /'spheres. 



new discussion of the subject, availing himself of all the more- 

 recent observations made in the lower latitudes of the southern 

 hemisphere, especially of those made bv the Venus expedition? 

 to the islands of Kerguelen, Auckland and St. Paul. . With 

 the aid of these additional observations lie obtained as an ex- 

 pression of the mean temperature of the southern hemisphere, 

 in centigrade degrees. 



T = 26°-0 + 6°-94 sin <p— 42*28 sin 3 <p 

 in winch <f is the latitude. 



By the integration of the temperature, given by this expres- 

 sion^ over the whole surface of the hemisphere, the average is 

 found to be 15°*4. From a comparison of this with the result 

 obtained by the writer* for the northern hemisphere, 15°'3, 

 Dr. Hann concludes that the mean temperatures of the two 

 hemispheres are probably exactly the same. 



Using exclusively only ocean stations of observation, he ob- 

 tained, instead of the expression above, the following: 



Prom tins expression the mean temperature of the southern 

 hemisphere was found to be 15°"2, very nearly the same as. 

 found from the other 



The comparison of the temperatures, given by the first of 

 the preceding formulas for the several latitudes, with those 

 obtained by Forbes for the northern hemisphere, is as follows: 



Latitude, 40° 45° 50° 55° 60° 



Difference, —0-6 + 0*1 + 0'7 +1*0 +1*6 



It is seen from these differences that, at about the parallel of 

 45°, the southern hemisphere begins to be warmer than the 

 northern, and as Dr. Hann now considers the temperatures at 

 the latitudes 40° to 50° to be pretty well determined, he thinks 

 it leaves the result with scarcely a doubt. It seems, therefore, 

 to be now well establ ..;, the temperatures of 



the lower latitudes of the southern hemisphere are less th;m 

 those of the same latitudes of the northern hemisphere, yet it- 

 is the reverse in the higher latitudes, and that there is no sen- 

 sible difference between the mean temperatures of the two 

 hemispheres. 



The reason of this becomes clear when we consider that if 

 the surface of the earth were all land, the difference between 

 the temperatures of the equatorial and polar regions would be 

 much greater than if the whole surface were water. In the 

 former case there would be no transfer of heat by means of 

 f Meteorological Researches, Part I; TJ. S. Coast Survey Report, 1875. 



