SHORT MEMOIRS ON METEOROLOGICAL SUBJECTS. 453 



As we have seen that the progress of the hurricane may be computed 

 to have been equal to six miles per hour, I have divided the hour into 

 six parts ; each of these parts, therefore, corresponds to a length = 1 

 mile, and the given curve may thus be considered as representing the 

 atmospheric pressure across the entire mass of air composing the hurri- 

 cane. With reference to the curved line thus constructed, I will remark 

 that the curve everywhere lies within the limits of the probable error, 

 since it is expressly stated in Professor Dove's description that the 

 barometer was in such a state of agitation that the mercury, at every 

 violent gust, suddenly fell two lines, but rose immediately afterward to 

 its usual height. 



The first thing that we now perceive, when we cast a glance at the 

 two branches of the given curve, is that these may be considered as 

 symmetrical curves with reference to the vertical line C Z, which passed 

 over St. Thomas at 7'' 52™; and this symmetry therefore authorizes us 

 to consider the line C Z as the axis around which the rotation took 

 place. In view of this, we may assume that the radius of the circum- 

 scribing cylinder of the calm space in the middle of the hurricane was 

 fj. = 0.4 hours = 2.4 miles. And if we next suppose that the rotating 

 mass proper extended over the entire space, in which the pressure 

 decreased proportionately inward toward the center, and especially 

 that the rotating mass reached outward to the point where the pressure 

 of the air was only two lines less than the normal i')ressure (33G'"), we 

 may make the radius of the exterior circumscribing surface of the 

 rotating mass (a -\- H) = 4.4 hours = 26.4 miles ; and, as a consequence 

 thereof, the thickness H of the rotating mass = 24 miles. During the 

 prevalence of the Antigua hurricane, the radius, therefore, of the space 

 where there was no wind was a = 0.1 S. And since the radius of any 

 element of the rotating mass is designated hy r = a •{- x, it will further 

 be easy, according to the constructed curve, to determine the amount 

 of pressure that was exerted at the i^laces in the hurricane which cor- 

 respond to values of x equal successively to 



O.IE', 0.2ir; 0.3^; OAH-, 0.5B; 0.75H; and S. 



By seeking in the figure the corresponding pressures given by the 

 curve for the increasing hurricane, we find the heights 



= 6.0; = 10.3; = 13.0; = 14.8; = IG.O; = 17.5; = 18.4 Paris lines. 



And by looking for these pressures in the curve of tbe abating hurri- 

 cane, we find 



^ = 6.6; = 10.0; = 13.4; = 15.2; = 16.2; = 17.5; = 18.2 Paris lines, 



from which are obtained, as mean numbers, 



z = 6.3; = 10.4; = 13.2; = 15.0; = IG.l; = 17.5; = 18.3 Paris lines. 



If we now consider the formula (4), in which X has the value deter- 

 mined by the formula (3), which value is seen to be a real number, wo 



