473
Replicated measurements gave a setting precision of 6.46 ¡xm, which
does not differ significantly from sx,y-
Then, a small wire of 1.3 mm 0 was introduced between the frame
and the glass plate. Four plates were exposed with the wire in various
positions. The results of the computations are shown in table 1. A
comparison between the standard error of unit weight s 0 and the
precision sx,y can be performed with an F-test. If the ratio s 0 2 /s 2 x, y
exceeds F0.95 (2,40) = 3.23, it is necessary to investigate whether the
plate has been bent or lacks of flatness. However, it would be best to
discard that plate.
Table 1.
Plate
dc lx m
dX H / W m
dY H'
SO n m
F
Note
1
+ 710
— 1538
+ 1300
103
149.5
Cannot be used.
2
+ 679
— 1355
— 323
29
11.8
Can be used after correc-
3
+ 594
— 543
+ 920
11
1.70
tion.
Can be used after correc-
4
+ 1158
+ 213
+ 3130
7
0.69
tion.
References:
1. Finsterwalder, S.: Über die Konstruktion von Hôhenkarten aus Ballonauf-
nahmen. K. Bayer. Akad. d. Wiss .1900, Bd XXX Heft II, S. 149—164.
2. Hallert-Ottoson: General differential formulae of the complete projective
relations between planes. Transactions of the Royal Institute of Tech
nology Stockholm, Sweden. No 140, 1959. Udc 526.918.7.