Appendix D. 487 



tlieory, since the wind in both cases is ahnost the same as would at mid- 

 night of the 19th indicate a central point, falling actually behind that 

 portion of the path of the line of centres already traversed on the 18th. 

 Upon this showing the direction of the wind at 6 p.m. of the 18th may be 

 assumed as that of the centre of the cyclone. In fact, the path of the 

 cyclone at this point lay parallel with the course the ship was holding, 

 whence only trifling variations would be observable in the direction of the 

 wind at those periods. Besides, the cyclone was at that time approach- 

 ing the vertex of its orbit, in doing which it encountered the large and 

 tolerably lofty island of Okinawa-Siraa of the Loo Choo group, which 

 mvist have resulted in a certain expenditure of the force causing the 

 gyratory movement of the cyclone. In analyzing the path of the cyclone, 

 account must also be taken of the winds that prevailed from the 17th 

 August up to midnight, although these are to be considered, with respect 

 to the cyclone proper, only in so far as they were winds that had been 

 altered in direction at the origin of the typhoon in conformity with the 

 laws of cyclones, which by no means imply in all cases a perfect gyra- 

 tion. However, as these winds are varied in direction by the same causes 

 which are in full activity in the case of the cyclones, such variations 

 must follow the same laws, and the lines of centres which present them- 

 selves with reference to these as parts of a circular orbit, naturally lie in 

 the same direction as that of the cyclone at its origin. 



As early as the 13th August a marked alteration in the temperature 

 of the air had been perceptible at Shanghai ; the thermometer fell from 

 between 86^ and 95° Fahr. to between 73°.4 and 78'.8 Fahr. : easterly 

 breezes set in, and the barometer rose in a remarkable manner for that 

 latitude and season. On the 17th the weather was still fine, but the sun 

 set red and fiery behind a dense mass of clouds. 



The morning of the 18th broke with continued fine weather ; but 

 cumulous clouds were massed on the sky, and looked black and threat- 

 ening to the N.E. By 8 a.m. the wind and sea had both risen ma- 

 terially. By 3 P.M. the roll of the sea was from N. by E., the sky became 

 stiU more cloudy, and the clouds began to descend ; banks of clouds in 

 the direction of the central point. At midnight between the 18th and 

 19th altered course to W. by S., in order to run out of the cyclone by 

 reaching its southern edge. 



