PART ii. THE CYCLE OF A DAY. 9 



vations comes out on being compared with the similar 

 daily series for the same months, with the same instru- 

 ments, and by the same lady observer, in Lisbon ; during 

 a former year, indeed, but tested also by the French 

 published barometrical and thermometrical records for 

 that city during the same year, and that not accused of 

 being very sensibly different from the Madeiran year 

 (see also Appendix I.). Wherefore we may commence 

 at once on the barometer returns, thus : 



The barometric pressure was always above 30 inches 

 at both places, as theoretically suitable for the latitude 

 and season, but higher in Madeira than in Lisbon by 

 0*09 inch ; which is also agreeable to theory, from 

 Madeira being 6 farther into the trade-wind region. 



Again, the bi-monthly range of the barometer was 

 smaller in Madeira than in Lisbon, or as 0*29 inch 

 against 0*40 inch. But this is also agreeable to geogra- 

 phical position ; and it used at the time to be almost a 

 wonder to us, lately arrived from northern lands, to see 

 how a fall of the mercury, through Something under two- 

 tenths of an inch, sufficed to produce all the phenomena 

 of the worst weather we ever had in Madeira, viz. the 

 breaking in upon the north-east trade-wind of a south- 

 western storm for two or three days, with rain, and then 

 gusty squalls, as the wind gradually veered to the west, 

 then the north-west, and finally fell into the north-eastern 

 quarter once again. 



So far we have taken both months together at each 



