138 Baron Humboldt on the Mean Temperature 



that over 5-6ths of the circumference of the globe, the equato- 

 rial aqueous zone, instead of presenting a mean temperature of 

 84°.5, has probably not one of 83°.3. Mr Atkinson himself 

 admits, p. 171, that the union of the aqueous and continental 

 parts tends to diminish the mean temperature of the equator. 

 But in confining himself to the continental plains of south 

 America, this philosopher adopts for the equatorial zone from 

 1° to 3° south, and upon different theoretical suppositions, 

 84°.56, or 87°.8. He founds this conclusion on the fact, that 

 at Cumana, in lat. 10°.27, the mean temperature is 81°.68, and 

 that, by the law of the increase of heat from the pole to the 

 equator, an increase which depends on the square of the cosine 

 of the latitude,* the mean temperature ought to be at least 

 above 84°.56. Mr Atkinson finds a confirmation of this re- 

 sult, by reducing to the level of the equatorial seas several 

 temperatures which I had observed on the declivity of the 

 Cordilleras, to a height of 500 toises ; and in employing cor- 

 rections, which he believes to be due to the latitude, and to the 

 progressive diminution of heat in a vertical plane, he does not 

 dissemble how much a part of these corrections is rendered 

 uncertain by the position of places in vast plains, or in narrow 

 vallies,— Mem. Astr. Soc. Vol. ii. p. 149, 158, 171, 172, 182, 

 and 183. 



In studying in all its generality the problem of the distri- 

 bution of heat on the surface of the globe, and in freeing it 

 of the accessory consideration of localities, (for example of the 

 effects of the configuration, the colour and the geographic re- 

 lation of the soil ; of those of the predominance of certain 

 winds, of the proximity of seas, of the frequency of clouds 

 and fogs, and, of the nocturnal radiation towards a sky more or 

 less serene,) we shall find that the mean temperature of a station 

 depends on the different ways in which the influence of the 

 meridian altitude of the sun manifests itself. This altitude 

 determines at once the duration of the semidiurnal arcs, the 

 length and the transparency of the portion of the atmosphere 

 which the rays traverse before reaching the horizon ; the 



* The law of increase approaches much more nearly to that of the simple 

 cosines of the latitude— Ed. 



