164 PHYSICAL GEOGRAPHY. 



is this source of error, that in every case the bearing of a line should be 

 taken at two places at least, and these should be as far apart as possible. 

 If the two bearings agree, it may be safely concluded that they are cor- 

 rect, and not affected by local attraction. Yet it must not be forgotten 

 that it is possible, though highly improbable, that two bearings thus taken 

 should be equally affected. If the line is very short, or if no two points 

 on it can be found at which the bearings agree, a point may be taken out 

 of the line in any direction, and at a suitable distance ; and, if the direct 

 and reverse bearings from it to any point of the line be found to agree, 

 those points may be considered free from local attraction. In some 

 places, as in the vicinity of iron mines, it will be found impracticable to 

 use the needle at all for the determination of bearings ; but even in this 

 case, the figure and area of a field may be found by so placing the com- 

 pass at each angle as to take the bearings of the two adjacent sides from 

 the same point This will give the angle between these sides without 

 error from local attraction. 



The most difficult problem ever presented to the surveyor is that which 

 asks him to retrace a lost line, with but one point known, and the bearing 

 from some old deed. To add to his perplexity, the parties in interest are 

 usually too much excited by the apprehension of being robbed of a square 

 rod of rocky pasture, or of swamp rich in mud and brakes, to be able to 

 give correctly such facts as might be serviceable in the solution of the 

 problem. In such case, if the parties cannot be induced to agree upon a 

 second bound and thus determine the line, there is no way but to "run 

 by tJie needle" after making due allowance for change in declination since 

 the previous survey. Running in this way may lead to the discovery of 

 some old landmark, nearly obliterated, and thus settle the dispute ; but if 

 not, though the error in the bearing is likely to be 15' to 30', it is better 

 than a lawsuit ; and if, in such case, the parties in their ignorance believe 

 that to be "true as the needle to the pole" is to be true enough it is 

 certainly an occasion where "'tis folly to be wise." 



Determination of a true meridian. That the surveyor may be able to 

 test his compass by some well established meridian, it would be an 

 economical measure if the state should locate and permanently mark a 

 true meridian in one or more of the principal towns of each county, 

 and then require by law all surveyors to record the declination with the 



