204 On the Courses of Hurricanes. 



advance of the main storm ; thus presenting the wind, for a few 

 hours, in a direction not accordant with that exhibited by the 

 main body of the storm. It may also be added, that in the 

 most violent of these storms, it is at least probable, if not cer- 

 tain, that the course of the surface wind is spirally inward, ap- 

 proximating gradually towards the center of the storm.- 



3. At stations within the tropics, the changes of wind during 

 the passage of the hurricane, are sometimes known to exceed 

 those which pertain to the passage of a regular circuit of wind '^ 

 these changes sometimes running through the entire circuit of 

 the compass, and even more. Again, they have been known 

 to shift hack and forvjard, in alternate and fitful changes, when 

 near the crisis of the storm. These phenomena, so far from dis- 

 proving the rotative character of these gales, only prove some- 

 thing more, and afford at least probable evidence, in support of 

 one or both of the following positions^ viz. 1. That high land, 

 and other obstructions, often produce sudden and fitful gusts and 

 changes in these violent winds. 2. That, in accordance with 

 our observations of minor vortices, the axis of rotation is often 

 impelled, excentrically, around a smaller circuit, in the interior 

 of thci advancing storm. 



4. In the northern intertropical latitudes, the recession or de- 

 parture of the southeastern limb of the storm, appears to be fol- 

 lowed, not unfrequently, by strong squalls or gusts from south- 

 east, this being the true course of the general trade wind that 

 determines the track of the storm. These gusts or squalls, if 

 taken for the regular action of the hurricane, may occasion erro- 

 neous deductions in regard to the course of the storm. 



5. In the latitudes near the exterior limits of the trade winds, 

 the change which here occurs in the course of the storm, pro- 

 duces apparent irregularities or anomalies in. the series of changes 

 presented by the wind. Owing to this cause I was misled to some 

 small extent in my estimate of the path of the first August hurri- 

 cane of 1830, as delineated on my first published chart, with an 

 irregular deflection of the curve on the coast of South Carolina, 

 which was predicated, in part, on the wind setting in at northeast 

 at Charleston, and veering to southeast as the storm became more 

 severe. 



6. At stations apparently within the regular track of the storm, 

 there will sometimes be an absence of violent wind ] or, the vio- 



