BY W. A. TULLY, ESQ., B.A. F.R.G.8., ETC. 137 
necessarily greater weight than if the base were calculated in 
the other way. 
[At this stage the author exhibited, for the information of 
members, a book in which the results of the computations had 
been entered, and also some other papers relating to the 
subject. | 
The base line being measured on a plain with an average 
height of 1298 feet above the sea level had to be reduced to 
that level. This reduction was effected by the formula:— 
“Multiply together the base and height, and divide the product 
by the radius of the earth, and the quotient will be the correc- 
tion to be deducted from the base.” This correction amounted 
to 2°2520 feet which had to be subtracted from the measured 
length. 
The probable sources of error were also investigated. None 
of them were likely to affect the accuracy of the work, and as 
they would not act in the same direction, the result would be 
according to the law of compensation that one would tend to 
counteract the others. 
The whole work depends on the correctness of value given to 
the standard bar used on the survey. It will be understood 
that a very small error in the length of the bar will more than 
overbalance all other errors that can arise. 
The several sources of error are as follow :— 
1. Error in determining correct temperature, 
2. The co-efficient used for expansion not being applicable to 
the steel of which the tapes were composed. 
3. The liability of the tapes being out of line. It may be 
mentioned that a declination of one inch out of the line in three 
places throughout the whole base will give a possible error of 
‘64 inches. A declination of one inch from the straight line 
in the middle of each tape would produce an error of 0°05 inches 
in the whole base. 
4, Frictional error by affecting the tension. 
