490 



Appendix D. 



They are also relieved of a source of error arising from tlie regular de- 

 cline for each day of the barometer, as evidenced by the observations 

 made during June and July, 1858, in mean latitude 23"" 52' N., mean 

 longitude 119^ 12' E. This downward tendency will be apparent from 

 the following readings for each hour: — for Ih. (a.m.) — - 0.001, 2h. -\- 

 0.005, 3h. + 0.0012, 4h. + 0.015, 5h. + 0.012, 6h. -f 0.006, 7h. — 

 0.02, 8h. — 0.012, 9h. — 0.021, lOh. — 0.02, llh. — 0.018, noon — 

 0.015, Ih. — 0.008, 2h. -|- 0.007, 3h. -f 0.021, 4h. + 0.025, 5h. -f 

 0.023, 6h. -f 0.015, 7h. -|- 0.008, 8h. — 0.001, 9h. — 0.008, lOh. — 

 0.014, llh. — 0.015, 12h. (midnight) — 0.011. These quantities are 

 to be read as implying that when added to or deducted from those sup- 

 plied by actual observations, they result in the quantities alreadj'" as- 

 signed as the corrected averages for the day. The direction as well as 

 strength of the wind are copied from the averages as calculated by the 

 Commodore from the ship's log, the meteorological journals and the daily 

 postings made by the Commodore himself. 



According to the delineation of the path of the cyclone, as prepared 

 from the observations recorded, the following table already, referred to, 

 gives the approximative distance of the ship at stated points from such 

 central path, as compared with that deduced from barometrical observ- 

 ations, allowing for the differences already mentioned. In the case 

 of the wind-pressure, the average is deduced from the mean of successive 

 observations taken every hour, and for tlie most part divided into inter- 

 vals of three hours each. 



17tli August 

 l«th 



19ih 



2nth 



4 A.M. 

 noon. 



iniilniglit. 



6 A.M. 

 9 A.M. 

 6 P.M. 

 3 A.M. 



5 A.M. 



8 A.M. 



noon. 

 3 P.M. 



6 P.M. 



9 PM. 



midnight. 



6 A.M. 



