Of I the Climate of the Canary Islands. 103 



years observations in 78° of latitude ; for which we are indebted 

 to the indefatigable, learned, and bold Greenland navigator 

 William Scoresby of Whitby in Yorkshire. 



Havannah, on the contrary, and Cairo, form, on the table, 

 two curves, which belong to quite different meteorological longi- 

 tudes, running parallel on each side of the Atlantic. The in- 

 fluence of the winter is considerable on both. In Havannah, 

 just under the tropic, the thermometer sinks in winter almost to 

 the freezing point, at the level of the sea, (Humboldt). On the 

 other hand, the summer in Cairo ascends so very considerably 

 above that of the Canaries, which lie still farther south, that it 

 is quite evident the decrease of temperature is not here subject 

 to the same laws. 



When the above mentioned temperatures of the Atlantic 

 zone are calculated by the known formula of Mayer, according 

 to which the temperatures decrease as the quadrant of the sine 

 of latitude, multiplied by an arbitrary co-efficient, it will be found, 

 that the respective results are by no means exact which Hum- 

 boldt has already remarked in his excellent treatise on isothermal 

 lines ; one of the richest sources for the meteorological know- 

 ledge of the earth\s surface. (Mem. d'Arcueil, iii. 481). But it 

 results at once from such calculations, in what latitudes the tem- 

 perature decreases more rapidly, and where more slowly, than 

 is consistent with the fundamental law ; and we are thereby led 

 to investigate and ascertain the interfering and modifying fac- 

 tors. Thus, the mean temperature of the pole, as obtained from 

 a comparison of the observations of Santa Cruz and at Kendal, 

 is 4°.9 of Reaumur, (20°.98 Fahr.) that of the equator 28°.2 

 (95°. 45 Falir.) The first is not so far from the truth as the 

 latter ; for Scoresby has shewn that the medium temperature of 

 the polar sea is several degrees under the freezing point. He, 

 indeed supposes the mean temperature of the 78th degree of lati- 

 tude to be 6°.7of Reaumurj(16°.93 Fahr.); but this is probably too 

 low, and rests on the untenable supposition that the polar curve 

 would have a similar progression with that of Stockholm. But 

 this place is too remote from the sea. It would be more cor- 

 rect to suppose the curve similar to that of the Norwegian coast, 

 and, were it determined by the data, furnished by M. Strom of 



