117 



Next we assume that the graduation of the compass is a good one and that the 

 pivot on which the needle rests is in the center of the graduation, and that both ends of 

 the needle read precisely 180 apart. 



Next, that the needle will be so sensitive that when deflected from its pointing by 

 the outside attraction of a piece of iron held about a foot or so away from it, it will settle 

 to its original position several times in succession. This sensitiveness depends on the 

 form and sharpness of the pivot, strength of its magnetism and its bearing upon the 

 finely polished jewel or steel cap. (ft is generally owing to the dulling of the point and 

 the scratching of the cap that a needle becomes sluggish and refuses to return to the sam* 

 point.) 



Then the extreme ends of the needle should come close to the graduation and, 

 together with the point of suspension, lie in the same plane with it, so as avoid parallax 

 in reading. It is also important that the center of gravity of the needle be as far below 

 this plane as possible in order that the quivering of the needle, so necessary to insure 

 the proper settling of the needle on the pivot, shall not be annoying. If the extreme 

 ends of the needle and the point of suspension are in the plane of the graduation, the 

 quivering motion will not be annoying, since the extreme ends lie in the axis of quiver- 

 ing and consequently are stationary, as shown in Figs. 3, 4, 7 and 9. In Figs. 1 and 2 

 the plane in which the needle and its ends are contained is much below the axis of 

 quivering which lies in the point of suspension and in consequence such needles can 

 only be read when they have ceased quivering. 



With a compass constructed as above we see no reason why its needle, or any 

 number of needles of the same shape as shown in Fig. 7, should not give the same 

 reading as long as the compass and its immediate surroundings remain undisturbed. 



Our task of showing that in a theoretically perfect compass the needle will always 

 assume magnetic North and South, debarring erratic oscillations due to earth cur- 

 rents would, therefore seem to be ended were it not for the fact that in a surveying 

 instrument a sighting arrangement, telescopic or otherwise, must be provided, by 

 means of which natural objects can be viewed and their relative positions in azimuth 

 determined with regard to the stationary pointing of the needle in the magnetic 

 meridian, affording a means for measuring angles and tracing lines. Seen from this 

 standpoint, it is therefore all-important that the plane passing through the slits of a 

 surveyor's compass, in which the line of sight is contained, be truly in line with the 

 zero points of the graduated ring and at right angles to the plane of the same. With 

 the transit, the line of sight, as denned by the optical axis of the object-glass and the 

 cross wires of the telescope, must not only revolve in a vertical plane in which the line 

 of collimation is contained, but this vertical plane must also either cut the zeros of the 

 graduations or at least be parallel to the plane passing through them. This condition 

 is, however, so difficult of attainment in instruments fitted with a telescope, and in 

 cases where it may exist it is so very apt to become deranged, that the writer is sure 

 that right here the principal cause can be found for most of the differences observed in 

 the reading of the needles of different instruments, though observed at the same time 

 and place. To illustrate this: suppose we bisect a distant natural object with the 

 cross wires of different instruments whose lines of collimation are in perfect adjust- 

 ment, then the readings of the needles assumed to have both of their principal axes 

 in coincidence would be the same if the zero points of the compasses are contained 

 in the vertical plane of the line of sight. But, if on the other hand the initial points of 

 the compass are not contained in the vertical plane of the line of sight the reading must 

 differ by an amount equal to the angle of this divergence in any one of the different 

 instruments. This index error, while often of no account in ordinary instruments, will 

 exist nevertheless, and must always be looked for even in the best class of instruments, 

 since in the field use of an instrument and by rough handling it is very liable to 

 change. 



This is particularly the case where a telescope is involved, since in its construction 

 we have to deal with the optical axis, which, of itself, is of a very compiex nature and 

 subject to changes affecting its line of collimation in relation to the compass, and 

 because of the changes occurring by rough handling in the position of the standards 

 carrying the wye bearings of the telescope's axis of revolution in the customary instru- 

 ments. Therefore, in order to eliminate all the errors due to the change of the line of 

 sight with regard to the zero points of the compass, the instrument should be tested 

 before any important work is undertaken with the needle. 



Before proceeding to determine the index error or constant of the compass of a 

 transit, it will b'e well to verify the adjustment of the line of collimation as well as its 

 motion in the vertical plane. Then, having previously established a meridian line by 



