W1NN ON STORMS. 109 



alls is constantly succeeded by hard southerly, or south-west winds, 

 attended with hazy weather and small rain. I think I am warranted 

 from experience to say constantly, for, in twenty-three instances 

 that have occurred since I first made the observation, it has inva- 

 riably obtained. I beg leave, however, to recommend it to notice, 

 as a matter, which when confirmed by further observations, and 

 generally known, may be of more consequence than at first appears. 

 To shew that it may, give me leave to recite the circumstance 

 which first occasioned my taking notice of it. Sailing down the 

 English channel a few days before the autumnal equinox, we had 

 a remarkably bright and vivid Aurora the whole night. In 

 shore, the wind was fluctuating, between N.N.W. and N.W., and 

 farther out W.N.W. Desirous of benefiting by the land wind, and 

 also of taking advantage of an earlier ebb tide, I dispensed with 

 the good old marine adage, never to approach too near a weather 

 shore, lest it should prove a lee shore ; and, by short tacks, clung 

 close along the English coast. Next day the wind veered to the 

 S.W., and soon after to the S. S.W., and sometimes to the S. We 

 were then in that dangerous bay between Portland and the Start 

 point, and carried a pressing sail, with hopes of reaching Torbay 

 before dark 3 but night fell upon us with thick haze and small rain, 

 insomuch, that we could not have seen the land the distance of a 

 ship's length. The gale was now increased to a storm ; in this di- 

 lemma nothing remained but to endeavour to keep off the shore till 

 the wind should change. Luckily the vessel was a stout one, and 

 well rigged. 



Reflecting some time after on the circumstance of this storm, 

 and the phenomena that preceded it, I determined to pay parti- 

 cular attention to future Auroras, and the weather that should succeed 

 them ; and, as I have above observed, in twenty-three instances 

 have found them uniform, except in degree : the gale generally com- 

 mencing between twenty-four and thirty hours after the appearance 

 of the Aurora. More time and observation will probably discover, 

 whether the strength of the succeeding gale is proportionate to the 

 splendour and vivacity of the Aurora, and the distance of time be- 

 tween them. I suspect that the more brilliant and active the first 

 is, the sooner will the latter occur, and will be more violent, but of 

 shorter duration than when the light is languid and dull. Perhaps, too, 

 the colour of the Aurora may be some guide in forming a judgment of 

 the coming gale. That which preceded the storm I have mentioned 

 was exceedingly splendid. The tempest succeeded it in less than 



