' [ 5s8 } 
V experiment when the legs of the inflniment are of une- 
qual bores, I would recommend it to the makers of thefe 
inftruments, to make ufe of tubes that are equal, or at 
" leaft nearly fo, that the error may become next to no- 
' thing, it being a thing very eafy to be done. In this 
manner we can readily determine the greatefl force,which 
the wind has blown with, during the time the inft rument 
' has been expofed to its action. But as it may be fafely 
~ left alone, by fcrewing its fpindle into the proper hand, 
or into the top of a poft, and as the wind never fails 
to turn the mouth of it towards itfelf, it is not neceffary 
for the obferver to continue always by it ; for it may be 
allowed to Hand all night, expofed to the wind, without 
any inconvenience, though it fliould even happen to 
■ rain very heavily. However, recourfe can only be had 
to this method of uling the inftrument on lliore : for at 
fea it muft always be held up in a perpendicular pohtion 
in the hand, whether it be ufed when only half full of 
water, or when quite full; which lafl will be frequently 
' found to be the only pradticable method of afcertaining 
the force of the wind during the night, when it blows fo 
* hard that it is dmpofiible to keep any lights on deck. 
A perfon filling the wind-gage, in a calm place, with 
water, in . order to determine the force of the wind, in 
the way which I have been juft now defcribing, will be 
. apt to imagine, that it cannot give the meafurement cor- 
, re6i ; for he will find fuch a repulfion to arife from the 
f edges of the hole g, as to fuftain a column of water in 
the kneed or bent tube, perhaps half an inch above the 
level; 
