May, 1960 
ANALYSIS OF CONTOUR ERRORS 
TABLE | — UNINTERPRETED GROSS SUM DISCREPANCIES 
163 
  
  
  
  
  
  
  
  
Instrument Specimen Flat area, Hilly area, Profile, 
millimeters millimeters meters 
A-8 3* 155.5 14.2 38.5 
20 13.8 0.7 24 6 
25 15.5 2.4 14.9 
27 31.5 3:3 15 7 
28 26.9 2.7 39.0 
29 8.2 4.4 12 7 
30 48.1 0.8 15.0 
mean + rms 24+15 2.3+1.2 20+10 
A-7 2 22.2 271 24.9 
6 c 0.7 16.4 
9 79.5 7.3 37:5 
11 34.6 8.1 20:5 
21 57.1 4.2 — 
26 113.5 3.4 24.8 
35 4.9 2-0 23.2 
522-52 4.0+2.8 25+7 
C-8 22 6.9 1.4 23.7 
23 lh EE —— 15.9 
Kelsh 7 9.9 6.6 32.9 
10 41.9 2.0 17.6 
12 34.7 3.8 27.3 
14 105.5 16.3 32.6 
31 22.6 8.2 22 9 
33 19.9 0.9 29.9 
34 56.6 1.8 36.6 
38 120.3 45.8 17.2 
41 c 8.8 29.9 
51 +39 10.4+14 29 4-5 
Balplex 5 250.5 55.1 41.0 
8 159.3 16.5 18.9 
40 58.2 26.0 46.3 
156 +94 32+21 35+10 
*?Not used in computing mean. 
  
  
and 296-303; and also as explained by Finsterwalder in the same journal in 1957, page 
390. Some minor lack of understanding exists with regard to the absolute value and 
meanings of the error constants, particularly the curvature error, but their relative 
values are nevertheless valid bases for comparison. The definitions of the 4 . . . B 
terms given in the accompanying table are as follows: 
Vertical error: 
Horizontal error: 
Direction error: 
Curvature error: 
nn — * (B + A tan a) meters 
m, = * (A + B cot a) meters 
m, -— * (À 4 B cota) grads 
m, — * (A + B cot a) units of curvature 
The method of computation of the 4... B values is not indicated here; it is explained 
in the reference. Considerable simplification of computation was afforded by the fact 
that all specimens were tested in the same places. The plus-minus variance of the 
individual 4 ... B terms was not evaluated because (1) the sample was so small the 
value would have little meaning and (2) the time for making the additional evaluation