AN INVESTIGATION INTO THE EFFECTS OF ERRORS IN SURVEYING. 877" 



fractional error in distance-transmission of the scheme as a whole. Such an assumption 

 will only be on the safe side when the number of triangles is less than five, and will 

 become rapidly more incorrect the more that number exceeds five [(40) et seq.~\. A table 

 is given (p. 873) which allows of the difference between the accuracy of distance-trans- 

 mission and that of the check-base being studied for different numbers of approximately 

 equilateral triangles, when arranged in the form of a straight chain. 



No matter what may be the number of triangles, it is possible, after the fractional 

 difference between the measured and calculated length of the check-base has been 

 computed, to assess roughly a value for the error in distance-transmission for any scheme. 

 This may be done by the aid of (29) and (40). 



TRANS. ROY. SOC. EDIN., VOL. XLVII. PART IV. (NO. 28). 129- 



