116 REPORT— 1867. 



five years, and their number therefore was 7300, of which 1076 were for 

 calms, and 98 for variables, leaving 6126 for the direction of the wind. The 

 distribution of this latter number for the foiu- quarters of the horizon was as 

 follows : from north to east (not including east) 683, from east to south (not 

 including south) 47-40, from south to west (not including west) 158, and 

 from west to north (not includiag north) 515, — showing that the number of 

 times the wind came from the points between east and south was nearly four 

 times as great as the number of times it came from the remaining thi-ee 

 quarters together. From east to S.E. inclusive, the number of observations 

 was 4286, which is more than two-thirds of the total number of observations. 

 This shows the great preponderance of the trade-wind, which prevails 

 throughout the whole year, but is strongest and steadiest from May to No- 

 vember, and more especially in June, July, and August. 



The wind veers almost always with the sun, or from S.E. through east to 

 north, IST.W., &c., decreasing in force as it veers. It often remains steady at 

 E.S.E. for a week or ten days. After it passes N.E., calms and variables 

 with light north-westerly and westerly breezes, and close sultry weather pre- 

 vail for two or three days. The trade-wind then reappears at S.S.E. A similar 

 revolution sometimes takes place in the course of a day. The wind seldom 

 veers in the opposite direction ; but it always does so during the passage of a 

 revolving storm on the east side of the island. 



VII. EORCE OF THE WlND. 



Diurnal Variation. — As Osier's anemometer is not affected by light breezes, 

 the force of the wind at the observation hours has usually been estimated. 

 Table XXVII. gives the mean estimated force derived from the six-hourly 

 observations. The results are but approximations; for, independently of 

 other sources of error, the site of the Observatory is not favourable for de- 

 termining the true force and direction of the wind, even with the most ap- 

 proved instruments. StiLL, the observations indicate that the force varies 

 directly as the temperature, the greatest pressure occurring at the waiTuest 

 hours. 



The last column in Table XXXVIII. gives the mean estimated force for 

 each hour derived from the term-day observations. Here Ukevnse, notwith- 

 standing the fewness of the observations, we see a general agreement between 

 the variations of the force of the wind and the temperature. 



Annual Variation. — Table XXVIII. gives the mean estimated force. Ja- 

 nuary, February, and March are the months in which the wind is strongest 

 at the observation hours, and next to them June, Jul}', and August. In the 

 former months huj-ricancs occur, and in the latter the S.E. trade-wind blows 

 in full force. From February to May the -wind decreases ; in June, July, and 

 August it is high ; from August to November it decreases, and from Novem- 

 ber to February it increases. There is thus an indication of a double pro- 

 gression. 



Mean Monthly Maximum Force. — Table XXIX. shows the mean maximum 

 force of the wind for each month, as recorded by Osier's anemometer, with- 

 out regard to the hour of the day. We find that, notwithstanding the severe 

 hurricanes which occasionally occur in the summer months, the mean maxi- 

 mum force of the wind is greater for June, July, and August than for any 

 other three months. We perceive also that this Table, like the former, points 

 to a double maximum and minimum. From November to February the wind 

 increases with an tacreasiag temperature, and from February to April it de- 



