97 



dicular to the plane of the transverse axis by the process described on page 48. 

 The striding-level is then placed upon the collars and is secured to the telescope by 

 means of its fastenings. The instrument being leveled up, the telescope is clamped 

 and the bubble of the strid ing-level is brought to the center of its graduation. 

 Now turn the instrument on its vertical axis and find, through the telescope, a suit- 

 able and well defined object, distant about 400 feet, which the horizontal wire bisects 

 when the bubble of the striding-level is in the center. Then remove the striding, 

 level, and, if the telescope reverses through the standards, as is usual in these 

 transits, unclamp the telescope and plates, reverse the telescope, turn the instru- 

 ment on its vertical center and direct the telescope again towards the object previ- 

 ously bisected. Clamp the telescope, replace the striding-level on the collars and 

 bring its bubble to the center. If the horizontal wire again bisects the same object, 

 its adjustment is correct, but if not, move the reticule bearing the wires over half 

 the error by means of the capstan- headed screws of the telescope, as explained in 

 the footnote on page 49, and repeat the operation until the adjustment is correct. 

 Jf the telescope is not reversible through the standards, as in Theodolite No. 14, 

 (or as in most Plane-tables), the horizontal axis of revolution should bo reversed 

 end for end in its bearings, and the plates should not be undamped. In other re- 

 spects the process is the same as above described. The instrument is now in ad- 

 justment for leveling with the striding-level. To insure good leveling with the 

 striding-level, it is necessary that its points of contact on the collars be free from 

 dust, grit, and oxide. 



The adjustment of the vertical wire should now be made as explained on page 

 49, by reversing through the standards, or if the instrument has low standards, over 

 the bearings of the horizontal axis of revolution, as explained on p. 100, after which 

 the adjustment of the horizontal wire should be examined again, and if found to be 

 disturbed, the two wires should be adjusted in turn, successively, until the adjust- 

 ment of each is found to be correct. See "Some Remarks Concerning Instrument 

 Adjustments," Part I. 



Tests of the assumed conditions. 



The Collimation. It has been assumed in the above methods of adjustment that 

 the line of collimation passes through the geometrical transverse axis of revolution 

 of the telescope ; otherwise, a change in its height would occur when the telescope 

 is reversed, giving two parallel lines of sight, of which the distance apart is double 

 the distance between the line of collimation and the geometrical transverse axis. 



An error from this source in a well designed instrument of modern construction 

 will be very small, and will be manifest only with short sights and in most cases 

 will be within the limits of error of observation for long sights, since the space cov- 

 ered by the cross-wire is proportional to the distance to the object, and will soon 

 amount to considerably more than this error. It would not, therefore, affect the 

 adjustment of the instrument for long sights. 



It is also assumed that the principal optical axis coincides with the geometrical 

 axis of the telescope for all focussing positions of the object-slide ; that is, whether 

 the telescope be focussed on a distant or on a near object ; otherwise the telescope 

 would give incorrect readings for objects at different distances ; more fully explained 

 in Part I. , Some Facts Worth Knowing/' 



The existence of these two sources of error may be detected by repeating the 

 test of the adjustment of the horizontal wire with the striding-level, using then a 

 very short sight. If the horizontal wire does not bisect the near object when the 

 telescope is both in a direct and in a reversed position, there is an error which 

 affects readings on near points, caused by one or both of the above mentioned 

 sources of error. This error may be eliminated with the striding-level, by the 

 method used in precise geodetic leveling to remove errors of collimation, which is 

 to take two readings of the rod at each point, one with the telescope direct, and the 

 other with it reversed ; the mean of the two being the true reading. This process 

 may be used for distant points if it is suspected that the adjustment of the horizon- 

 tal wire has been disturbed. This error cannot be eliminated with a fixed level ; nor, 

 as it might be supposed, with a reversion level, that is, a fixed level accurately 

 ground and graduated upon the top and bottom (for an illustration see elsewhere) 

 since both horizontal axes of such a level are adjusted to the line of sight for long dis 

 ances, and because the level cannot be reversed on the telescope tube. The effect of an 

 error of this kind in short distances can, however, be easily eliminated when 

 backsights and foresights are equal in length. 



