654 Mr W. HOPKINS, ON THE EXTERNAL TEMPERATURE OF THE EARTH, 



radiates directly from the Sun to the Earth's surface in the above latitudes respectively. If 

 we denote hQ by H, and adopt a notation analogous to the above, we shall have 



H a - H 2 = 10°,S4 (C), 



H° - H 3 = 17°,2, 



H -H t = 20<M. 



„ *«-*t 9,41 __ 



Hence — — - = = ,87, 



10,84 



16,8 

 17,2 



22°,9 

 H -H 4 20,1 = ' ' 



22. It will be observed that in the data from which these results have been deduced, I 

 have omitted the value of s — S lt the difference between the values of s at the equator and 

 in the latitude of 30° N. It leads, as above remarked, to an anomalous result. In fact we have 



s - «, = - 1°,18, 



which would give the particular effect of the solar heat, expressed in s, greater in the latitude 

 of 30° than at the equator. The anomaly of s, being greater than s is due to the circumstance, 

 that the effect of currents and other local causes has been only partially eliminated in the 

 lower latitudes by taking for u the mean value for a whole parallel. In fact the influence of 

 these causes is such that the parallel of maximum annual temperature, instead of coinciding 

 with the equator, lies in about 10° N. lat. The anomaly would be in some degree corrected 

 by taking the mean of the values of u for 30° N. and 30° S. latitude ; but the difference 

 between s and s l would still be much smaller than that which, I conceive, must necessarily 

 result from the difference of position with reference to the Sun, of the equator, and either 

 parallel of 30° of latitude. On referring to the values above given of the temperatures (m„) of 

 each parallel, it will be observed that the difference of these values for the equator and the 

 latitude of 30° is only 5°,5, while the difference for the latter parallel and that of 48°. 50' 

 amounts to 14°,65, though the difference of latitude in the latter case little exceeds half that in 

 the former. This is due, I conceive, to the peculiar position of the region immediately about 

 the equator as the region of maximum temperature. If the surface of the Earth were perfectly 

 uniform, and there were no atmospheric currents producing a horizontal transference of heat 

 from one parallel of latitude to another, the influence of the Sun expressed in s would be some 

 determinate function of the latitude, equally with u . But the values which u , and therefore 

 also s, would have on this hypothesis in any assigned latitude, would manifestly be changed by 

 the difference between the heat brought to the proposed region by horizontal transference, and 

 that carried away from it in a similar manner, in a north and south direction. Now this com- 

 pensation afforded by the heat which is brought, for that which is carried away must generally 

 exist in a more or less perfect degree in those latitudes which are intermediate to the equatorial 

 and polar regions, while in those regions themselves no such compensation can exist ; and this 



