136 



Reversion Level. 



Applicable to any of our Engineers' Wye Levels. 



The spirit level used in this feature differs from the ordinary one in that it is 

 ground to the true shape of a barrel so that the tangents to the level bubblo curves, at the 

 zero points of the scales, are parallel and diametrically opposite. By the use ot this rever- 

 sion level, attachable to any of our Engineers' Wye Levels, in place of the single read- 

 .ing level, it is possible to do good leveling, though the adjustments of the spirit level 

 and cross wires are entirely deranged and the collars worn (see remark below), by first 

 making the level bubble central and taking a reading, then by revolving the telescope 

 180 in its wyes, which point is indicated by an adjustable stop,* making the bubblo 

 again central and taking another reading. The arithmetical mean is the correct result. 



This device will, in an emergency, be appreciated when it is known that by the use 

 of the method above the work will average as good as that done with an ordinary good 

 wye level, in adjustment. The adjustment of an instrument provided with a reversion 

 level is made in precisely the same manner as if the spirit level was of the single read- 

 ing kind, since the adjustment of the level when it is reversed will take care of itself. 



The reversion level is guarded by a revolvable outer tube (Patented, not shown in 

 cut) leaving a space of air, as a non-conductor of heat between it and the ordinary level 

 mounting tube. This exterior tube serves both as a protection against breakage and 

 sudden changes of temperature, and, as its inner surface is painted white, it also acts 

 as a reflector which facilitates the reading of the bubble. 



Remark : The inequality of worn collars cannot be eliminated in a strict sense 

 by using the reversion level, yet for ordinary good work it may be said to be. Nor can 

 the test for the equality of the collars be directly tested in this way but should be done 

 as in the case of the ordinary wye level ; viz , by the two-peg method described under 

 the adjustment of the Dumpy Level (pages 54 and 55 of our handbook). The follow- 

 ing modification is to be noted: After the line of collimation has been adjusted for dis- 

 tant objects by rotating the telescope in its wyes and the spirit level has been adjusted 

 by reversing end for end and adjusted laterally^ (the telescope having the sunshade at- 

 tached, as it serves to balance the telescope when the object slide is drawn in), the in- 

 strument is set up close to the near target, and a reading is taken with the level tube 

 in the direct position. In order to eliminate the error of collimation for nearer objects, 

 should any exist, another reading is taken with the telescope rotated 180 in the wyes, 

 and the mean taken as the true reading. If, now, the horizontal wire also bisects the 

 distant target and the bubble remains central in each position of the telescope, the 

 cjllar.s are of equal diameter. Should the latter not be the case, the error may be cor- 

 rected thus : Bisect the distant target with the telescope in its direct position, and adjust 

 the level till the bubble is central. Rotate the telescope 180 in its wyes, indicated by 

 the siop, and note the number of divisions through which the bubble moves in ordei 

 that the distant target remains bisected, so that a correction can be made when most 

 precise work is required. 



It is assumed that in making this test the temperature of the two collars has been 

 alike run! that the telescope has been in proper balance by being focussed for a distance 

 of about 800 feet with sunshade attached. A scratch on the telescope or object slide 

 indicates the focus which the maker used in equalizing the collars. An apparent crn.i 

 found as above may be due to a change in the shape of the level tube which may occur 

 in time (for which the maker, of course, cannot be held responsible), as well as to a 

 worn condition of collars, or these causes combined. 



In order to trace the error to its source the only sure lest is made with a striding 

 level _'in r'a Precise Level.) 



l*ri'<', ;i< above* if ordered with our Wye Level in place of the single reading 

 kind #L>0.00. 



* It is an extremely difficult matter to grind a level of this kind so that the bubble will remain central at 

 all positions during this rotation through 180. The stop just mentioned is so adjusted, however, by the maker, 

 that when the level has been turned exactly 180 it gives a correct reading. 



