

246 Mr J. Sang on a Systematic Method of 



sition of the points in the two surveys differed from one another, 

 but the difference on being analyzed, was found to be nearly in 

 a constant ratio to the distance of the points from the first sta- 

 tion in the systematic survey. In fact, had the base line been 

 conceived to be a little longer than it actually was, the points in 

 both surveys would have coincided so nearly, that for practical 

 purposes it would have been unnecessary to alter them. 



A convenient arrangement for avoiding parallax in the in- 

 clination is to have all the signal staves of the height of 7 J links, 

 and divided from the top downwards into alternate white and 

 black spaces of one link each, leaving the remaining third for 

 being pushed into the ground. When observing the inclina- 

 tion, the height of the instrument is measured, and the telescope 

 is directed to a point at the same height on the signal stave ; 

 there is consequently no other correction required, except in long 

 lines, for the curvature of the earth. By this simple arrange- 

 ment some tedious work is saved ; and the depression at one 

 end of the line can be compared, on the ground, with the ele- 

 vation at the other. If landed proprietors were aware how 

 small an additional expense is incurred in measuring, along 

 with the ordinary subjects of a survey, the height of every sta- 

 tion above the sea or other conspicuous level, they would seldom 

 omit having it done. The closes in altitude seldom require 

 correction. This evidently proceeds from the inclinations being 

 always small, and from their being referred to the same zero at 

 each station. 



The method of co-ordinates allows the whole attention to be 

 devoted to having the surveying-lines convenient for offsets, 

 and over even ground, except that they must also be as long 

 as possible. They should be over even ground to avoid errors 

 in length, and the too frequent use of the instrument for hori- 

 zontal reduction. If possible, they should be so arranged, that 

 the inclination can be measured from one termination, or, at 

 most, from both terminations at the same time with the bear- 

 ing. The calculations already described render certain the de- 

 tection of any blunder in the measurement ; and the very 

 trouble of dividing the small errors, and the satisfaction de- 

 rived from perfect closes, induce great care in the manipulation 

 of the chain. The surveyor is also forced to verify 'his chain 



