ERRATUM 



Page 83, Hue 19 to line 23. — Erase, and substitute the 

 following : 



In measuring an angle between an elevated object and one 

 on the horizon, a direct measurement may introduce more 

 error than would result from reflecting to each object in 

 succession a third object also on the horizon, making an angle 

 as nearly as possible 90° with the elevated object, and taking 

 the difference between the two observed angles. It should be 

 noted that whatever may be the elevation of the object, if 

 another object be on the horizon and distant exactly 90°, the 

 true horizontal angular distance is also 90°, and therefore equal 

 to the observed distance. 



This follows from the formula given on p. 357. 



Cos horizontal angle =cos angular distance X sec. alt. 

 If angular distance =90°, 



Then cos horizontal angle =cos 90°xsec. alt.^^0. 

 Therefore horizontal angle = 90°= obsd. angular distance. 



It also follows that if the observed angular distance is greater 

 than 90°, the true horizontal angle is greater than the observed 

 angle, and vice versa. 



In the event of both objects being elevated, a natural mark 

 on the horizon should be selected as nearly as possible; 90° 

 from one of the objects; and a second mark on the horizon, 

 making an angle with the other object, also as nearly as possible 

 90°. From measurement of the angles between the horizon 

 marks and both elevated objects, tlie required horizontal angle 

 between the latter may be deduced with approximate accuracy 

 dependent on the degree in which the necessary conditions 

 governing the selection of the horizon marks have been 

 fulfilled. 



In the above figure, let A and B be two elevated objects; 

 required A'B' the horizontal angular distance between them. 



