258 HYDROGRAPHICAL SURVEYING [chap. xi. 



Y's, when the mean of these two observations for each object 

 will be the correct amount of elevation or depression. This 

 is the best way, and eliminates all error. It may, however, be 

 sometimes convenient to proceed as follows : 



Ascertain the collimation error by directing the telescope 

 on to an object in elevation, reading the vernier, then turning 

 the telescope round until the level is uppermost, and again 

 adjusting for the object and reading the vernier again. Half 

 the difference between the readings is the collimation error, 

 which, when the reading taken with the level uppermost is 

 greatest, will be added to observations of elevations made with 

 the telescope in its normal position, and subtracted from 

 depressions. This collimation error is permanent for all 

 positions of the horizontal arc. 



For level error, at each observation of each separate object, 

 the telescope must be brought horizontal by the level attached 

 to it, and the vernier of the vertical arc read. Whatever it 

 reads will be the level error. 



The sign of the correction to be applied for this error is, 

 for elevation, +, when the 0° of the arc is above the zero of 

 the vernier when the tube is level, and — when below. For 

 depressions the signs will be reversed. Care must be taken 

 that no mistakes are made as to these signs. For a tyro it is 

 slightly confusing. 



Both level and collimation error must be applied to each 

 observation. 

 Sextant When the ship can be well fixed, sextant angles of elevation 

 from her with a sea horizon will be very good, as good, in fact, 

 as elevations with a small theodohte, as they are free from all 

 possible errors of levelling, etc., and a sextant measures angles 

 to ten seconds, w^hereas a small theodolite is only cut to minutes. 

 Even when the ship is within the limits of the sea horizon, the 

 results A^ ill be good, providing the distance of the shore line is 

 well kno^n, and is not under half a mile. By observing from 

 the lowest step of the accommodation ladder, we can use a 

 shore horizon at even less distances. 



Sextant elevations, then, are very useful, but we do not 

 generally get so many opportunities of obtaining series of 

 heights by it ; and when at any distance from the land, only 

 the skyline of hills will be clearly seen, so that it is principally 



Eleva- 

 tions 



