Mtthd ofmeafuring a SU^s U^ny at Seat J^f 



which exlfts between it and the fliip's motion in direft progrefs. To prevent this, a fliort 

 tube is to be fitted to the extremity of the immerfed tube^ in fuch a manner as to be capa- 

 ble of moving freely in a vertical, in an horizontal, or in any intermediate arch (which 

 may be effected by a joint confifting of a portion of a fpherical fliell containing a fimilar 

 fhell) to be furniihed with vanes extending .behind the joint, which are to be placed in 

 fuch a manner, that the ftream adting on them may point the orifice fo as to meet it per- 

 pendicularly; but a (hip at fea is often defle£ted from her path into a curvilinear track by 

 the figure of the waves ; and fince the ftream is caufed by the relative motion of this 

 water to the {hip's motion, which may always be confidered as in the direction of the 

 keel, it muft meet the orifice of the inftrument in the direcflion of the keel, (fee note H) 

 that is perpendicularly, and therefore the water delivered through .the fpouting aperture 

 muft reprefent the undulating curve through which the fliip moves, inftead of its abfcifs, 

 or true diftance, (fee note C) which might fometimes differ confidcrably from it ; for 

 which reafon the following method of correfting it becomes neceflary. Let a barometer 

 made narrow in the middle to check the fudden motions of the mercury, be placed againft 

 a bulk-head, or tranfverfe partition of the fiiip, and fo loaded and hung that when the ftiip 

 heels it may fwing in an arch defcribed on the furface of this partition, fo as not to partake 

 of this lateral inclination (which is not difficult to accomplifti, the motion of heeling being 

 neither very fudden nor fluftuating) but in fuch a manner as to receive the intire etFeft of 

 the pitching motion which it is intended to reprefent ; then if the curve which the fhip 

 defcribes be fuppofed to be divided into indefinitely fmall equal right lines, any one of 

 them will be to the portion of horizontal diftance which it overhangs, as radius to the co- 

 Cneof the angle which it forms with the horizon ; but the length of the column of mer- 

 cury in the inclined pofitlon of the barometer, is to its length in an ere£t pofition as radius 

 to the cofine of the fame angle ; therefore the length of the column of mercury at any time 

 will be to its perpendicular height, as the reftilinear portion of the path which the (hip 

 fliall be then defcribing to true diftance ; and fince thefe reflilinear portions are equal, and 

 pafled through in equal time*, the mean of them obtained by the narrowed tube will repre- 

 fent them all, therefore the length of the column of mercury marked on the fcale of the 

 barometer will always be to its perpendicular height as the fum of thefe reftilinear fpaces ; 

 that is, as the curved path of the ftiip reprefented by the water difcharged by the inftrument, 

 is to the fum of the horizontal fpaces overhung by them ; that is the true horizontal diitance, 

 from which data the inftrumental diftance may be eafily correfted at any time. But it is to 

 be obferved, that the height of the mercury in a perpendicular pofition is variable, on ac- 

 count of the difference in the weight of the atmofphere at different times ; therefore when 

 it is required, the barometer muft be taken from its place, and held carefully by the hand 

 for a few minutes in an upright pofition, until the mercury ceafes to defcend, when it will 

 fliew the true perpendicular height necelTary to this experiment, and at the fame time the 

 barometrical ftate of the atmofphere (fo ufeful at fea) more accurately than any fufpended 

 barometer. 

 Vol. v.— May i8oj. H _ The 



