218 PHILOSOPHY OF STORMS. 



Facts collected by Mr. Redfield, taken from Sillimarfs Jour- 

 nal, vol. XX. 



158. u The earliest supposed trace of this hurricane 

 which has been obtained, is from off Turk's Island, in the 

 West Indies, where it appeared on the 1st of September, 

 two days previous to its reaching our coast. It was felt 

 there severely, but at what hour in the day we are not in- 

 formed. 



The next account we have is from lat. 23 43', where the 

 storm was severe, September 1st, from S. E. to S. W. 

 Whether these two accounts are considered as identifying 

 the storm, or otherwise, will not, at this time, be deemed 

 material. 



Our next report is from lat. 32 30', Ion. 77 west from 

 Greenwich, on the night of September 2d ; a hurricane for 

 three hours. 



At 3, A. M., on the 3d of September, a severe gale was 

 experienced thirty miles outside of the American coast, off 

 Wilmington, N. C. 



At Wilmington there was no gale. 



At Ocracock bar, N. C., at day light on the morning of 

 the 3d, a severe gale from E. S. E. 



At Edenton, N. C., the gale was at N. E. 



Off Roanoke, on the morning of the 3d, a dreadful gale 

 at E., then S. W. and N. W. 



A vessel from Charleston, S. C., two days previous to ar- 

 riving in the Chesapeake, experienced the gale at 4, A. M., 

 on the 3d, from S. E. to W. S. W. 



A vessel from Bermuda experienced the gale from the 

 westward, on the inner edge of the Gulf stream. 



Another vessel, from Charleston, did not experience the 

 gale. 



In lat. 37 30', on the inner edge of the Gulf stream, gale 

 from the westward, with squalls. 



