165 
The computations of M and A for the 15 above mentioned tests were carried 
out by hand-computer engaging two calculators for about 50 working days. The 
mean values have been written on a planimetrie plot of the control points of the polygon 
at the scale 1:200.000. The distribution of the values M appeared in any case slow 
and regular with wide zones of maximum and minimum variously placed. On the 
other hand, the distribution of the differences A appeared completely accidental. 
It was practically impossible to find any systematic behaviour. These remarks, 
660 670 680 680 700 710 720 730 740 750 
390 
760 
390 
380 
370 
360- 
8.7 *14.0 .17.3 
f.i: 
+ ♦ 09 
♦16 .36 
j: 
B.4 *1V 
2.6 «6.2 *8.9 
• 1*0 
♦ lfl.7 *% 7 
02.2 *4.8 
-1' *0*7 
-4.4 -4.1 
• • 
-as 
♦14 *0.6 »1.2 
-0.2 
e 
♦2,5 *74 -0.2 
-4.2 -13 
-8-5 -6.6 
* V 
-4.7 -1.6 
-380 
-370 
360 
350- 
♦ 6.2+7 7 -*8.7 +10,7 
>• t o a. 
-7.1 
-13.7 
-6.3 
-7.3 
- 73 -7.3 
• -&3* 
i 2 -u., 
-350 
340- 
330- 
320- 
* 4 l 2 .8.7 *8.8 
j • • • 
♦ 2,3 
►5.6 +0.3 
‘ • +73 +63 
♦79 I* * 
• *7J <5 9 
I +6.2 
^ 1 
* 2 i *3# 
>8.5 .86 
>3 j .4*8* 4 i 
♦2fi ‘'i 
*27 00 
.5,4 
« 
5.7 4 4.6 
0.1 -0.6 -2, 
• • • 
-1.1 41.3 
♦27 
► 4.9 
♦4 B +5.7 
• • 
♦71 *70 
• 3 ♦ 4,7 *7,4 I 
-, f • • 
*1.2 .0.3 
• • 
tìù— 
*1,4 ‘2.4 -2.. 
.1 * * 
340 
♦3.7 ♦ 98 
*8.2 +7.6 
• « 
*4.4 
♦ 2.9 
♦2.3 • 
• 
-1.5 
o * up 
* .01 
• * 
-1.2 
-2.0 
OP 
A 7 -I.® 
330 
320 
310- 
300- 
.3.5 
**? »*2.4 
-0.2 
-1.1 
.4.7 -4-9 
1.1 -0.2 -1.1 
310 
-300 
660 670 680 690 700 710 720 
Fig. 31 
730 740 750 760 
even if superficial, are sufficient to establish that the dimension of the circle has been 
chosen in a suitable way since a clear distinction resulted between the behaviour of 
the general character of the errors and the behaviour of errors of local and accidental 
character. 
We did not think it suitable to reproduce all these maps (in total 78: 39 for M 
and 39 for A) in this report; they are available for further eventual studies at the Office 
Secretary of the Group. To give an example, in figs. 30 and 31 we represent the plani 
metrie distribution of the means M and of the differences A concerning the coordinate 
Z of test 17.