258 Mr. Nixon's Theory of the Spirit-Level. 



on either vertical side of the vessel concentric to the axis of" 

 rotation. 



Vertical lines passing through points situated near each 

 other, differ so slightly in parallelism, that the vessel might 

 have been moved forty horizontal feet in the interval of mark- 

 ing the two zenith-points, without introducing an error in the 

 zenith-distance oft/ equal to half a second. If the expansion 

 of the liquid should exceed that of the glass of the vessel, the 

 increase of temperature which augments its volume v/ill ele- 

 vate its horizontal surface, and cause a sensible diminution in 

 its area and the length of the line L I. Decrease of tempera- 

 ture will therefore augment the hoi'izontal surface, and elon- 

 gate L I. In either case the zenith-point v/ill be invaluably at 

 the point of bisection of the arch situated above the surface 

 of the liquid, without regard to its extent. 



When the temperature of every part of the vessel is not the 

 same (which may be the result of handling it, or breathing on 

 it), the circular figure is destroyed, and the length as well as 

 probablj' the figure of the surface of the liquid undergo alter- 

 ations. The graduations of the distorted rim are therefore 

 rendered unserviceable from their inequality ; and should the 

 partial temperature affect the vessel where in contact with the 

 surface of the liquid ; in such case the true zenith-point cannot 

 be found otherwise than by drawing a line through C pei'- 

 pendicular to L I. Generally the surface of the liquid will ap- 

 pear to advance towards that part of the rim bulged out by 

 the partial increase of temperature. 



In the construction of a spirit-level the upper interior sur- 

 face of a straight (hollow) cylinder of glass is ground in the 

 direction of its axis to a perfectly circular arch. Either end 

 being permanently closed, the cylinder is nearly filled with 

 spirits of wine or ether, and the other end hermetically sealed. 

 Hence it is evident that any section of our circular vessel 

 perpendicular to its sides, when nearly filled with the pro- 

 per liquid, and securely closed up, would be equivalent to 

 a similar spirit-level. In this instrument the atmospheric 

 air incumbent on the ether, &c. or rather their surface of 

 contact, is termed the air-biibble, or simply the bubble; and 

 is represented in our circular vessel by the horizontal surface 

 of the liquid on which is drawn the straight line LZ. 



Having learned from our experiments with the circular 

 vessel, that the length (or figure) of the bubble LZ would not 

 alter in a constant temperature, we may restrict ourselves, in 

 lieu of finding the zenith-points v and t/, to the marking of 

 either end of the bubble, as L or /, before^ and the same end 



subsequent 



