.Ow the Climate of the Canary Islands. 127 



much violence as to be even dangerous to vessels. On the 

 supposition that the course of the trade-wind is an ascending 

 one, it might be expected in autumn to cease to blow at heights 

 some days before it ceased on the sea shore, progressively de- 

 scending in height as the northern limit of the wind at the 

 surface of the sea approached the island ; and such is found 

 to be the fact. The S.W. wind, which takes the place of the 

 trade-wind as the latter ceases to blow, is observed in all 

 years to descend progressively from those highest points, on 

 which, as has been remarked, it prevails throughput the year. 

 Its descent is traced by the clouds, which, in October, veil the 

 south side of the Peak of Teneriffe ; sinking slowly, they rest 

 on a ridge of mountains, between Orotava and the southern 

 coast, six thousand feet in height, where they break in dreadful 

 storms. It is a full week, and sometimes even more^ before 

 the S.W. wind is felt on the sea coast, where it prevails for 

 months aftenvards, whilst it rains on the declivity of the moun- 

 tain, and snow is on the Peak. A remarkable fact, viewed in 

 this connexion, is the greater height of the barometer in the 

 summer months, when the opposite currents are prevailing 

 over the island, than in the winter months, when the S.W, 

 alone prevails. The mean of Don Francisco Escolar's register, 

 kept during three years, gives, for the months of winter and 

 summer respectively, as follows, the heights of the column of 

 mercury being reduced for temperature. 



Inches. Lines. 



May, June, July, August . . 28 „ 3.173 

 September to April , . . 28 „ 2.017 



Excess of summer over winter . 1.156 



or rather more than one-tenth of a British inch. 



The island of Grand Canary presents a remarkable pecu-» 

 liarity in the progression of its mean monthly temperature, 

 which is highly worthy the attention of meteorologists. The 

 following table exhibits the result of a register kept at Las 

 Palmas, during ten years, by Dr. Bandini de Gatti. The time 

 of observation was daily at noon ; from whence the results in- 

 serted in the first column are immediately derived ; those in 

 the second column are the approximate mean temperatures, de- 

 rived from the observations at noon, by presuming the same 



