HORKICANES. 227 



The mercurial column then begins to descend, rapidly at first, and then 

 more slowly, till the centre of the Hurricane has passed over, when it 

 begins gradually to rise, and the reverse of the commencement ensues ; it 

 attains a higher level, and then as suddenly falls to the mean height. Thia 

 is if the whole of the meteor pass over it, and the centre be crossed. 



Upon a little consideration, it will be evident that Hie form of the surface 

 of the revolving storm, or the section of the vortex, is described by the 

 variations in the barometric column. It by no means follows that, prac- 

 tically, this will always be found : a ship may only skirt the exterior of 

 the storm, and, consequently, the mercury will only rise, or oscillate, 

 according to the relative position of the Hurricane and the ship ; but it 

 may be taken as an indication, when the barometer begins slowly to rise, 

 after being depressed, that the greatest danger has passed over, or that the 

 ship is steering away from it. Therefore, should there be any sudden 

 change in the barometer, either rising or falling, its indications should 

 never be neglected, especially during the period, and in the regions, subject 

 to these storms. The barometer sometimes sinks two inches during the 

 progress of a Hurricane.* 



One great advantage in the Aneroid Barometer is, that its variations 

 occur simultaneously with their causes. In the Mercurial Barometer the 

 friction of the mercury on the tube, and other reasons, concur to make the 

 column rise or fall at some time after the change has occurred. In this the 

 Aneroid Barometer possesses great advantage, and it has another very great 

 claim to notice — that it clearly shows very minute changes, which the 

 oscillation or pumping motion of the mercury in very bad weather will not 

 allow to be estimated. 



Another warning of the coming Storm is a long low swell from the direc- 

 tion in which it is approaching. In the open ocean, when the waves have 

 free action in deep water, the unbroken swell of the ocean travels with 

 immense velocity, rising at times to about 400 miles an hour. Any extra- 

 ordinary swell or rollers, more or less confused, may be taken as the 

 evidence of some distant Storm. Another early sign is when light feathery 

 cirrus clouds are seen radiating from a point on the horizon where a 

 whitish arc indicates the bearing of the centre, though it may still be 

 hundreds of miles away. Unmistakable signs are a falling barometer; 

 halos about the sun and moon ; increasing swell ; hot, moist weather, 

 with light variable winds ; rolled and tufted forms of clouds, with deep 

 lurid red and violet tints at dawn and sunset ; a heavy mountainous cloud- 

 bank, of a dark or leaden appearance, on the distant horizon, with darting 

 forks and threads of pale lightning.f This cloud-bank may be seen, perhaps, 



* At midnight, December 22iid, 1892, the corrected reading of the barometer on board 

 the steamer Werkendam, was 27*75, near the centre of a Hurricane, in lat. 39" 41' N., 

 long, 30° 41' W., one of the lowest readings recorded in the North Atlantic Ocean (see 

 page 111). 



t Although it is true that the prognostics of a common coming storm are, in general, 

 sufficiently plain to be understood by a spectator, from the angry appearance of the 

 firmament, yet it is also true that there is no particular indication in any one quarter 

 of the horizon sufficiently marked, like the space occupied by the Black Squall panoply 

 of the Caribbean Sea, so that an acute seaman may say, "thence will the blast come." 

 On the contrary, the clouds gather together (we speak from experience) in dense masses, 



