[ iO, ] 



underftand it) a plano-concave lens, whofe plane fide is uppermoft 

 and a refledling level ; covering a box filled with fpirit of wine, 

 except a bubble, which refts under the higheft part of the concave 

 furface ; being a fort of fpirit-level ; but to procure it the ac- 

 curacy of fuch, the radius of its concavity fhould be as great as 

 that of the fpirit-level ; and I fuppofe it mufl be difficult to 

 grind it truly to fo great a radius * ; and to find that point in the 



concave 



* If any kind of fpirit-level be made ufe of to fix a reflefting plane horizontal, 

 and its bubble were, from the attraftion of cohefion between the glafs and the 

 included liquor, to reft below the fummit of the concave arch of the tube, Sec. the 

 error this will produce will be of an angle, double that meafured by the arch of the 

 tube between the point of reft of the bubble and the fumrait of the tube, (as can be 

 coUedled from what has and will be mentioned) ; now I have intimated above, that 

 not only the fenGbility but the accuracy of the fpirit-level, or its exemption from 

 errors arifing from this caufe, will be improved by enlarging the radius of curvature 

 of its tube : and becaufe it has been aflerted, that '< it could not be depended upon 

 " to greater accuracy than one minute ; for that when the radius of curvature of 

 " the level is very great, the force with which the bubble rifes to the higheft point 

 " is fo very fmall, that the leaft adhefion of the liquor to the glafs will ftop it, and 

 " its place of reft will be uncertain,"* which feenis to intimate, that enlarging the 

 radius of curvature would be of little advantage: I hope I ifliall be pardoned for 

 inferting the following ftatement, which inclines me to think otherwife ; though 

 an acquiefcence in this as a fa£t would be a recommendation of the level here de- 

 fcribed, which is not liable to fuch objedion, and though it is only with diffidence 

 and reluftance I cjn dilTent from the very accurate and ingenious perfon who 

 aflerted this, and whofe pofition I would not venture to controvert, had he ftated 

 his having afcertained the faft by experiment. 



Let there be two fuch levels, whofe curvature is reprefented by the arches JF, 

 AG, fig. 1, AC being the radius of the former, and the point ^ the fummit of 



both ; 

 • Vide Ludlams Remarks on his Artronomical Obfervations. 



