plate and adjustment computations according to the principles which have been 
treated above, the standard error of unit weight of the measurements was deter- 
mined. For the big surface plate which is visible in Fig. 4 the standard error was 
found to be of the magnitude 4 microns. 
  
Ws 
FiG. 4. The Leitz Incident Light Illuminator Ultropak. 
This accuracy ought to be sufficient for the determination of the flatness of 
many types of surfaces in photogrammetry. 
A great number of photogrammetrié diapositives have been measured with 
this method. In the Tables A and B two examples are demonstrated. 
Table A 
Deflections of the emulsion side of a diapositive. 
Microscope method 
Diapositive, emulsion down, supported by a surface plate of high precision. Deflections 
\ PI a 0) I gh [ 
are demonstrated in microns and positive upwards. Confidence limit about 15 microns. 
Confidence level 5 per cent. Absolute orientation with the aid of the points which are 
p I 
denoted 0 
o 0 o+15 o+6 04-9 Qcl-3 
040-9 o+24 o-+23 o+16 04-19 
o+1 04-21 o+18 04-25 o 0 
o —10 04-8 o+19 o+5 o+12 
o 0 o+13 o+ 10 o+ 14 o+12 
Table B 
Deflections of the emulsion side of a diapositive 
Microscope method 
Diapositive, emulsion down, supported by a surface plate of high precision. Deflections 
are demonstrated in microns and positive upwards. Confidence limit about 15 microns. 
Confidence level 5 per cent. Absolute orientation with the aid of the points which are 
denoted 0. 
o0 o--27 04-17 o —6 Q,— 35 
o+12 o+33 o0 4- 46 o 4- 22 o — 14 
04-27 04-40 04-45 o--39 o 0 
o+7 o+32 o+51 04-3] 04-1 
o 0 043-6 0 13 Oo 11 0 39