220 OBSERVATIONS ON THE MEAN TEMPERATURE 



In the reasoning from which Humboldt estimates the mean 

 temperature of the Equator, he appears to me to have taken 

 for granted a very material fact. Having found a coincidence 

 between the mean temperature of equinoctial America and 

 equinoctial Asia, he concludes that the mean temperature of 

 the Equator is 81°. 5, and is uniform in every point of that 

 great circle ; but as these are the very regions under the line 

 where the temperature should be the same, in consequence of 

 being similarly situated with regard to Canada and Siberia, no 

 conclusion can be drawn until a similar temperature has been 

 found on the African coasts of Benin and Loango. The heat 

 under the Equator being thus supposed to be uniform, Hum- 

 boldt felt himself entitled to conclude, that the colds of Ca- 

 nada and Siberia did not extend their influence to the equato- 

 rial plains *, and that between the tropics, the isothermal lines 

 are parallel to the equinoctial. 



The theory which I have explained above, requires a diffe- 

 rent distribution of heat at the Equator. The maximum mean 

 temperature of that circle should be 82 ".8 in Africa, in or- 

 der to give 81 ".5 as the equinoctial temperature in America and 

 Asia; and the difference of these values, or 1°.3, must be re- 

 garded as a measure of the influence which the colds of Canada 

 and Siberia extend to the equatorial plains. Nor is this a mere 

 theoretical result. I consider it as fairly deducible from facts 

 furnished by Humboldt himself; and this distinguished travel- 

 ler seems to have drawn from these facts the same conclusion, 

 before he had deduced the uniformity of the equatorial tem- 

 perature 



Edinburgh Philosophical Journal, vol. iii. p. 263. 



