[ ,01 ] 



allel to the back-horizon-glafs ; when this ad- 

 justment may be performed in the fame manner as 

 the correfponding adjuflment of the fore-horizon- 

 glafs. But the bending of the index, that follows 

 the fetting it off the arch, Is a very difagreeable cir- 

 eumitance, having a tendency, efpecially on board of 

 ihip, to expofe both the index and centre work to.' 

 damage j and may even, without extraordinary pre- 

 cautions taken by the inftrument maker in placing 

 the plane of the index-glafs exactly according to the 

 length of the index, difturb its perpendicularity to- 

 the plane of the quadrant : on thefe accounts it would 

 be much better if thisadjuftmentof the back-horizon- 

 glafs could be performed, like thofe of the fore- 

 horizon-glafs, with the index remaining upon the 

 arch of the quadrant. Fortunately, this defideratum- 

 has been lately efTecled by an ingenious contrivance 

 invented by Mr. Dollond, which he has given an.: 

 account of in a letter addreiTed to me*, which I have 

 prefented to this Society, by means of an additional 

 index applied- to the back-horizon •' glials ;■ whereby 

 both the adjuft ments may be made by the fame oh— 

 fervations and with nearly the fame exa6inefs. as thofe 

 of the fore- horizon -glafs :. for a farther knowledge, o£- 

 which, fee the account itfelf. 



Befides the difficulty of adjufting the back-hori- 

 zon-glafs, the want of a method of directing the, 

 line, of fight parallel to the plane of the quadrant 

 lias proved al'fd a confiderable obftacle. to ihe tile 

 of the back-obfervation :. thk will ealily appear- 

 from the following propofition, that the error of 

 the angle meafured arifing "from any fm all de~. 



f- See the XIV th paper, which-, immediately precedes this., 



- viatica: 



