Having collimated the horizontal wire by the experimental test 

 described under V, (bl), run the instruments over the pegs and 

 in the final step move the bubble- -not the diaphragm. The 

 peculiar purpose of a set of pegs, as used for this method, is to 

 establish an artificial horizon. If the collimated sight line is made 

 to occupy a position in this horizon and the bubble checked up 

 to agree with tl.ese conditions, the sight line will describe a level 

 plane when the bubble occupies the center of its scale. 



(b) If one has a reliable level at hand, a satisfactory adjust- 

 ment of the telescope bubble may be accomplished by setting up 

 both instruments so that the eyepieces are at the same height. 

 Sight some distant point in the hori/on through the level, bring 

 the telescope of the transit to bear upon the same point, and adjust 

 the bubble to this condition. 



(c) Prof. S. D. Sarason of Syracuse University, in order to 

 obviate the peg method, advocates the use of a reversion level to 

 secure the adjustment for parallelism. The hori/ontal cross wire 

 being collimated as directed under V, (b-1), proceed as follows: 



Level the telescope with the reversion bubble in a normal 

 position by use of the tangent screw. Take a careful reading on a 

 rod, say 300 ft. distant. Reverse the instrument in altitude and 

 azimuth so that the telescope will be re-directed to the rod with the 

 reversion bubble above. Center the bubble again and take a 

 second reading. If the scales of the level vial register accurately, 

 a mean position between the two observations on the rod will be a 

 true horizon for that particular set-up. Bring the horizontal wire 

 to read on this mean position by means of the vertical tangent 

 screw, then, center the bubble with its own adjusting screws. 

 Now reverse the instrument, and having set the telescope on 

 the mean position a second time note the position of the 

 bubble on its scale. If it is not perfectly centered, either the 

 cdllimation adjustment for the horizontal wire is in error, or the 

 tangential axes of the bubble scales are not parallel. 



The Reversion Bubble (see p. 93) is a spirit level ground 

 barrel shape on the inside surface so that the tangent axes of the 

 opposite scales will be. parallel. All manufacturers will acknowl- 

 edge that this is a difficult task, and to get both bubble races ground 

 to exactly the same curvature is even more difficult. We will 

 not, however, send out a reversion bubble unless the scales register. 

 A slight variation in the sensitiveness of the scales is of no 

 consequence but the bubble must occupy the center, when the 

 telescope is in the horizon, either erect or inverted. The bubble 

 inu.t, therefore, be adjusted to a pioperly collimated cross wire. 



70 



