CONTRIBUTIONS TO METEOROLOGY. 



2.i 



According- to Table V, the average direction of tiie wind tor tlie tiiree months here considered is 

 two degrees nortli of east. According to Table VI it is 4.] degrees nt)rtli of east. The average 



Table V.—Obnermtions of the wind from Mauri/'n Pilot Churln of the Atlantic Ocean. 



ATJGUSX. 



SEPTEMBER. 



OCTOBEK. 



course of the storms mentioned in Table III, while moving westward, was 26° north of west; 

 that is, they came from a point 26° south of east, which differs 28° from the average course of the 

 wind by Table V, and ditters 'o()° by Table VI. It is clear, then, that the West India cyclones do 

 not follow the average direction of the wind for the region in which they occur. Tables Y and 

 VI, however, show that winds from the SE. and ESB. are very common, and the observations 

 quoted in article 'S.i show that SE. winds very generally succeed a West India cyclone. These 

 facts seem to indicate that the direction of a cyclone's progress is not determined by the direction 

 of the i>reva]eut wind for that season of the year so much as by the direction of the principal wind 

 which prevails at the time of the cyclone. 



36. I next undertook an investigation of the cyclones origi?)ating in the region south of the 

 continent of Asia. Table VII contains various particulars respecting those cyclones of this region 

 whose paths have been best determined. It includes all those which were most carefuUv investi- 



