Chap, ii.] Jfater-fpouts. 473 



flood of water fell upon him with fuch violence as al- 

 moft to beat him down, and die fpout immediately 

 paffed off with a roaring noife into the lea. The 

 water from the ipout, he remarked, was perfectly 

 frefli. 



Dr. Perkins adds feveral other inftances on the 

 teilimony of mariners, who all affirmed, that tltey faw 

 the water dejcend from the cloud through the water- 

 fpout into the fca, contrary to the opinion of Mr. 

 Oliver, that it always afcends. 



A whirlwind, therefore, in the opinion of Dr. Per- 

 kins, cannot be the caufe of a water-fpout; nor 

 can both of thefe phenomena proceed from the fame 

 caufe. A whirlwind, he fuppofes to be produced by 

 the afcentof the heated or rarefied air into or through 

 the colder regions of the atmofphere above. Now, 

 Dr. Arbuthnot fays, that the rarefaction of the hotteft 

 day renders the air but one tenth lighter than it is in 

 the coldcft. 



This roaring noife alfo, as remarked by Captain 

 Melling, does not agree with the theory of the afcent 

 of water in the fpout, as it is not very clear why 

 fuch a noife fnould accompany the fimple afcent of 

 water *. 



To determine the matter, it is to be wiihed, that 

 future obfervers would be careful to remark, ift. The 

 incipient ftate of a \vater-fpout, and in particular whe- 

 ther any cloud is feen hovering over the part in which it 

 commences j and 2dly, whether the conical part feems 

 gradually to defcend from the body of the cloud. 



A tornado feems to partake much of the nature of 

 the two preceding phenomena, but is more violent 

 in its effects. It commences very fuddenly, feveral 



* Philad. Tranf. vol. ii. 



clouds 



