If there is a difference, the True Horizon for this H, I. at this 



station will be determined by taking the mean of both 



observations. 

 In this case, bring the telescope to read on the True Horizon, thus 



determined, when the line of sight must be truly horizontal. 



Bring the bubble to the center of its scale by its own adjusting 



nuts and the test is complete. 

 As a final test, reverse the instrument once more and, centering the 



instrument on the horizon line a second time, see that the bubble 



is perfectly centered. 



The adjustment of the bubble may be tested and rectified in 

 this way in three to five minutes any time during the day, on any 

 sight, but these directions do not assume any such necessity. 



Being satisfied with the collimationand bubble adjustment, the 

 engineer will leave the eyepiece at one end and the sun shade at the 

 other for continuous work as in ordinary practice. 



The instrument illustrated in Fig. 15 has a telescope 25 cm in 

 length and an aperture of 30 mm. It has been designed to combine 

 unusual compactness with utmost accuracy, and the construction is 

 such that the telecope can be instantly transformed from the erecting 

 to the inverting type by a simple exchange of eyepieces. We are 

 prepared to supply two orthoscopic (inverting) oculars producing 

 18 and 26 magnifications, also an erecting eyepiece (not shown in the 

 illustration) with a special sun shade to balance, giving 20 magnifi- 

 cations. 



For convenience in manipulation, the normal position for the 

 Vertical Control should be determined, as directed on pp. 15 or 23. 

 This will be most conveniently accomplished by swinging the instru- 

 ment exactly 180 by use of the release key in the clamp-and-tan- 

 gent movement. In the lugs of this mechanism are small set screws 

 by which 180 revolutions can be very exactly accomplished. This 

 arrangement is designed to bring the instrument directly back on 

 the rod when reversed for double observations. 



The normal adjustment of the Vertical Control is one of 

 convenience and not of importance. The collimation and bubble 

 axes being parallel, they are brought, by this means, into a plane 

 that is horizontal in all directions. The w T ork is further facilitated 

 by restoring the bubble to a normal position with the vertical axis 

 before setting up at any station. Taking this precaution, it should 

 not be necessary to turn the fine motion screw over half a turn, in 

 either direction, to center the bubble accurately at each observation 

 and to overcome, in this way, slight errors in the adjustment of the 

 vertical axis. 



If the focusing device is unimpaired and the bubble in working 

 order, perfect results can be obtained with this instrument in remote 

 frontier districts or elsewhere, on the assumption that the instrument 



32 



