[ I03 J 



alio were it properly conftruded, it would require to be ad- 

 jufted every time it is ufed or ftirred from its place, unlefs it were 

 fet upon another level plane, and to make this adjuftment would 

 require the light of the day or of a candle ; fo that it would not 

 anfwer to make by night and in ^ny place obfervations of the 

 flars, which are often the only ones a traveller has an opportunity 

 of making. I apprehend it would require at every obfervacion a 

 confiderable part of the trouble of adjuftment, which in the 

 level here defcribed ferves once for all ; for after being adjufted, 

 it requires no farther care than to keep the needle duly 

 pointed. 



As an inftrument which does not, like this one, deted its 

 own errors, may gain credit for an accuracy which it has not, it 

 will be unfair to judge of the prefent level only from a com- 

 parifon of it with fuch a one : the faireft trial of its adjuftment, 

 &c. will be made by obfervations taken in calm weather from 

 the furface of a fluid of fufficient extent, and not covered by 

 glafs planes. If the weight were to be fattened to the ftem 

 without the intervention of a cord or hook, the tremors of the 

 mirror would be fuppreifed, and the wind would communicate no 



motion 



radii will be as i to 4 : if the fines were as the radii, or as i to 4, the angles 

 or arches J D, AE would be equal, but (ince .SZ) is by fuppofition only half BE, 

 and the angles are fmall, the arch of aberration AE will be little more than 

 half the arch A D : and hence it appears that the error of the level will' be q. p. in 

 nearly the inverfe fubduplicate ratio of the radius of its curvature ; fo that when 

 the radius is as I, 4, 16, &c. the aberration will be as 4, 2, i, &c. or as i, i, 

 I, &c. ■ 



