274 



ciently accurate to form the basis of a permanent survey ; 

 that numerous errors had been caused by depending on the 

 compass ; and that the true meridian ought to be the datum 

 line, from which all the bearings of a survey should be 

 reckoned. 



These vieAvs have been generally accepted as correct, and it 

 becomes a matter of interest to know, what means a surveyor 

 has at his disposal for ascertaining the true bearing of his 

 lines. 



No doubt the best and easiest way is to connect with some 

 line, the bearing of which has been ascertained with accuracy. 

 But such a line is not always accessible, and a surveyor 

 should have it in his power to establish his bearings by inde- 

 pendent observation. This can be done, within a very small 

 limit of error, in fact as near as the theodolite can be read, 

 which is sufficiently accurate. 



Observations to ascertain the variation of the compass are 

 being constantly taken at sea, but the same degree of accuracy 

 is not required as in a survey on land. The altitude of the 

 sun or a star is used in most of these methods. In using alti- 

 tudes we are liable to error from refraction ; moreover the 

 theodolite, the land surveyor's instrument, is better suited for 

 measuring horizontal angles or azimuths than altitudes. For 

 these. reasons it is desirable to use a method, equally good, in 

 which altitudes are not required. Again, it will be well if 

 we can dispense with a knowledge of the latitude. 



It is not a very difficult problem to find the latitude ; still 

 there is advantage in having one problem to solve instead of 

 two, and if the surveyor can find the meridian without first 

 finding the latitude, so much the better. 



There is another point to be considered in estimating the 

 values of different methods ; that is the amount of time con- 

 sumed in observing. At an observatory this is a secondary 

 consideration. The main point is to obtain perfect accuracy, 

 and for this purpose transits are observed, day after day, and 

 corrections made by the help of the clock. But a surveyor 

 requires something more expeditious. 



There are three well-known methods available for him, and 

 I propose to consider them, with reference to the tests 

 already mentioned ; and then to apply the same tests to a 

 fourth method, which is the subject of my paper. 



The first method is by a single observation of the sun or a 

 star. For this an altitude is required, also a knowledge of 

 the latitude. It is, therefore, open to objection on both 

 accounts. 



The second method is by equal altitudes. This requires 

 two observations at an interval of several hours, and is there- 

 fore inconvenient in point of time. It often happens, more- 



