pancies between the adjacent strips along a profile, this profile is so adjusted that 
the sum of the corrections for each of X, Y and Z along the profile is equal to zero. 
Points lying between these profiles are adjusted by linear interpolation. A separate 
profile-wise adjustment of the block is undertaken for X, Y and Z. 
4 — ACCURACY OF THE WORK 
4.1. Relative Stability of the Block. — Fig. 2 shows the discrepancies in X, Y 
and Z for tie points, of adjacent strips, along the 18 transversal profiles mentioned 
in paragraph 3.3. These differences are the discrepancies noted after the strip adjust 
ment but before the block adjustment. 
Fig. 2 - Remarks: Discrepancies listed above are: a) Always given as A1 , 
A Y (in meters) 
A Z ' 
b) Lower strip value minus upper strip value 
4.2. Absolute Accuracy of the Block. — Excluding all the points whose identifi 
cation was in doubt (I.Q.: 53, 73, 84, 90, 119, 107 and 108), and the following 
points with abnormally big discrepancies due to obvious, but yet undiscovered blund 
ers, (50, 102, 106, 109, 126, 129, 130, 141, 143 and 232), we get the following mean 
square values: . . _ _ 
m x = i 10-3 m 
m 7 — zb 10.8 m 
m p = 14.9 m representing 1.8 % 0 Z (209 points) 
m z = di 9.9 m representing 1.2 % 0 Z (200 points) 
It may be 
3 and 4: 
of interest to see the abnormally large mean square errors in strips 
m Y = zt 16.2 m 
m y = zb 31.2 m 
m p = 35.1 m (17 points) 
m z = zb 15.0 m (17 points) 
Such high discrepancies reflect the circumstances detailed in paragraph 2.2.