(366) 
to adopt a method of testing with grids by means of monocular observations 
which would permit appreciable lowering of the personal equation, and grids 
with an interval of 5 mm. 
In reference to the measurements, the Italian report thinks it preferable to 
compare the mean value of six observations made on each point with the cal- 
culated theoretical values. The general characteristics of the stereoscopic image 
would be better shown by the average of all the absolute values of the discrep- 
ancies. 
In Paragraph 5, determination of the mean square errors, one foresees that 
for the apparatus using the Porro-Koppe principle allowance would be made 
for the distortion of the lenses used in plotting and others, could be made 
without correction for distortion. The trials would have to be made with 
allowance for the distortions of each of the lenses of the air cameras (multiple 
cameras). Moreover, the last paragraph is not satisfactory because it does not 
consider the scale of the original photographs. The report furnishes different 
calculation of the theoretical precision of different plotting machines; the theo- 
f 
retical error of the instruments to the scale of the photography is e — s X — 
zZ 
s being read at the scale of the plotting, f the focal distance, z the height of 
flight to the scale of the photographs. The report next describes the method of 
testing used at the IGM and the Officine Galileo. It does not differ essentially 
from the proposed method but we will not discuss the details. 
The IGN of France believes that this testing proposal applies only to a 
very small part of the numerous operations which compose an aerial survey. 
Such a survey includes all of a series of operations which should be made by 
coherent means with corresponding precision for each step (photography, con- 
trol, plotting) considering the specified objectives: a map at a given scale satis- 
fying high standards of precision. 
The standard test of the proposed procedure is only a test of the mechan- 
ical and optical assembly and adjustment of plotting machines which might 
be obtained in other ways; for example in the Poivilliers B by reading of 
divided circles; for machines not having divided circles, it is possible to use 
adjustment methods which refer to the fixed points of a simple grid, without 
reference to the dimensions of a sub-divided one. 
Finally, if the method of inspection proposed is practically applicable to 
machines using a mechanical correction (Stereotopographe Poivilliers D or 
Stereocartographes Santoni) or optical correction of the distortion (Wild Auto- 
graphes A5 and A6) when these suffice to cancel the diapositive errors; it is 
not so practical for machines using the Porro-Koppe principle. 
The phrase of the last paragraph which says *the theoretical determination 
(calculated values?) of the coordinates takes account of the distortion of the 
lens it is proposed to compensate for," appears quite soothing. However, it will 
be long and difficult to transcribe to linear corrections in the plane of the grid, 
the corrections to be made to the calculated coordinates to eliminate the influ- 
ences of the distortion and this with a precision such that the errors thus intro- 
duced will remain neglible in relation to the errors to be measured. 
A discrepancy in the distortion of 0.5 Centigrade (16 seconds) in a point 
of the photograph superimposed on a cross of the grid produced in reference 
to the theoretical position will be calculated as an error of the order of 
nir 
ref 
gre