on the Climate of the Canary Islands. 93' 



from May 1808 to August 1810, will meet with a favourable 

 reception. However much they still leave undetermined, such 

 as they are, they fill up an important gap in our knowledge 

 of the distribution of temperature, and, it may well be affirmed, 

 cannot be dispensed with in the formation of a scientific meteor- 

 ology. 



I have reckoned the observations of Don Francisco, which he 

 has had the goodness to communicate to me, by decades. Taking 

 the mean of these, I reduced the whole to a tabular form *. 



Don Francisco had provided himself with good EngHsh in- 

 struments, which were placed in the shade, in an open gallery, 

 remote from reflection. In this respect, his observations are ful- 

 ly deserving of credit. The time of observation was at sunrise, 

 and at midday, or somewhat later. Hence, it might be thought, 

 that, in this way, the extreme heat would never be observed ; 

 that the medium would consequently be too low. This impres- 

 sion would be strengthened by the remarkable fact, that the tem- 

 perature at midday exceeds that at sunrise, on an average, 

 only by 1.16 of Reaumur, (2.61 Fahr.) But Monsieur Thi- 

 bant de Chanvallon, (Voyage a la Martinique, 1763), has long ago 

 shewn, that, in islands in w^arm climates, the extreme heat never 

 falls later than one o'clock, or, what very rarely happens, half 

 past one, but is frequently found to be soon after eleven, and 

 very generally at midday ; the elevation of the temperature, by 

 the culmination of the sun, being probably prevented by the 

 sear-breeze, which has now reached its greatest strength. Now, 

 however much the small difference of temperature at the hours 

 of observation might cause an error or unfavourable posi- 

 tion of the instruments to be suspected, this singular phenome- 

 non is, nevertheless, fully confirmed by the observations of He- 

 berden, 60 years earlier, at Funchal in Madeira. For, besides 

 the mean monthly temperature, he gives also the extremes in 

 each month ; and the mean difference of these, in the course of 

 four years, amounts to not more than 2.91 of Reaumur, {6.55 

 Fahr.) It is therefore very probable, that the difference of the 

 medium temperature may only be half so great. There is no 

 plain at Santa Cruz any more than at Funchal, the mountains 



• This will be found in the Scientific Intelligence of this Number. 



