303 SURVEY OF THE COAST 



as before the true angle between the verticals of the observed 

 objects. 



I his operation will besides bring; the horizontal angle ex- 

 actly to the points of the circle diametiically opposite to the 

 former, and will again act as a correction in the same sense 

 as the two mentioned before. 



An error in the line of collimation will of course combine 

 with that of the verticality of the circle descrii)ed by the teles- 

 cope, and be corrected partially by the double operation ; but 

 this adjustment is easily verified in this instiument. 



An iriegularity in the axes of motion of the instrument 

 would of course have an influence on the observations ; t)ut 

 in the horizontal angles, it is evident that this influence 

 would be destroyed by the exactly inverted jjositio'is of the 

 instrument in which the observations are made. 'I his error 

 therefore falls in with all the otiiers, wliich appear of course 

 always combined, and are finally compensated by a certain 

 determined number of observations systetnatically arranged. 



The only cause of error still remaining is the accidental 

 error of any particular division, which might have been used 

 in the series of angles. This chance can occur only by 

 some particular accident in the work, as the dividing engine 

 of Mr. Troughton is so regular, and his attention and care 

 so great, that the error of the division may be supposed a 

 minimum. 



The displacement of my microscopes from their exact 

 distance of 130° was too small to produce any influence on 

 the accuracy of the compensations. 



Upon the |)rinciples here demonstrated I grounded the 

 following method of observing all horizontal angles with the 

 two-feet theodolite, 



1. Having carefully adjusted the instrument in all respects, 

 the telescope is placed in such a position as to brino; its eve 

 eiid perpendiculatly over the microscope A. In this posi- 

 tion, which I call direct, 1 observe all the objects between 

 which I intend to determine the angles at the time. 



2. Then I turn the telescope round vertically, so as to 



