C lr 4 1 



But it is evident neither of thefe motions can be ao 

 curately performed. In the third method, the ob- 

 ferver is only required to move both himfelf and the 

 quadrant, fo as to keep the Sun always in or near 

 the axis of virion* which may be performed very 

 well, becaufe the axis of virion is a vifible and cer- 

 tain direction for it. One exception, however, mould 

 be made to this general rule, namely, in taking the 

 Sun's altitude when very low, by the back obfer- 

 vation $ in which cafe it will be befi to ufe the fecond 

 method, or elfe to hold the quadrant perpendicular 

 by judgment j which will be much facilitated by 

 ufing a telefcope containing wires in its focus parallel 

 to the plane of the quadrant, as defcribed in p. 106 : 

 for, in this cafe, the perpendicular pofition of the 

 quadrant cannot be attained fo near by the method of 

 turning the quadrant on a line going to the Sun as 

 an axis, as it can by the other method. 



It remains to treat of the errors which may arife 

 from a defect of parallelism in the two .furfaces of 

 the index-glafs, and to point out the means of ob- 

 viating them in the celefliai obfervations. It is well 

 known, that if a pencil of parallel rays falls upon a 

 glafs whofe two furfaces are inclined to one another, 

 and fome of the rays are reflected at the fore-furface, 

 and others palling into the glafs and fufrering a re- 

 flection at the back-furface and two refractions at 

 the fore-furface emerge again from the glafs, thefe 

 latter rays will not be parallel to thofe reflected at the 

 fore-furface, as they would have been if the furfaces 

 of the glafs had been parallel, but will be inclined to 

 the lame. I find that the angle of their mutual in- 

 clination, which may be called the deviation of the 



rays 



