(208) 
deformation of the support, photometry of filters, efficiency curve of shutters, 
centering of cameras, etc., — it is necessary to dispense with this analytical 
approach to the matter in establishing the proposed specification and to con 
centrate on the delineation of the properties of the apparatus as a whole. 
From this statement, conclusions such as the following may be drawn: 
A film camera should be tested with film: it is of little importance that its 
lens has outstanding properties when it is studied on a surface which cannot be 
distorted, if the conditions under which it is to be used do not reveal them. 
The proposed test, for example, would show no difference between the SOM- 
Poivilliers, Aquilor plate cone, F = 125 mm, 19X19 cm, and the film cone 
having the same geometric attributes; one could not expect the same precision 
of the two pieces of equipment. The test of resolving power should be made on 
targets which offer the contrasts and brilliances encountered in practice, using 
the apertures and exposure times which are used in practice. A shutter of low 
efficiency will make it necessary to use a large aperture with the consequent 
inconvenience of a large vignetting or an extended time of exposure with the 
risk of registering the image motions which we will discuss in another com 
munication. Why not even use moving targets of the type used in the investi 
gation I just mentioned, or of any other type? That would be much better 
than eliminating, by an arbitrary choice of emulsion (speed and granularity) 
the perhaps disastrous effect of the shutter. 
The tests should be made under conditions of temperature and hygro- 
metry similar to those of normal use; of what importance is it that the distor 
tion measured in laboratory tests was in a well-defined plane, if in practice, one 
cannot endow the sensitive surface with this well-defined form? 
Let us cite a second point of view which may explain some of the diffi 
culties encountered. It would appear that the authors of the last proposal had 
in mind the definition of the typical camera of a production series: they wor 
ried hardly at all about lack of symmetry in the distortion, they measured only 
along one or two diagonals, they are opposed to the particular control of the 
prototype or rather of the type of apparatus, demanding particular care in the 
routine control of apparatuses of the same type. Now the photogrammetrist 
does not use a standard camera with a standard lens, a standard shutter and an 
ideal sensitive surface. He buys one or two apparatuses, or more, and is quite 
justifiably concerned with their particular properties. It is much more interest 
ing to know that such and such a make of camera is famous because its lenses 
have the same distortion within very small tolerances, its shutters are very 
efficient and perfectly standardized, its centerings are very precise, etc., that to 
know the distortion of the theoretical lens, the efficiency curve of the theore 
tical shutter, etc. The user therefore should demand that the quanlities are as 
described by the manufacturer within tolerances which may be depended on in 
practice. This implies complete control of all pieces of equipment. 
There is also a third explanation for the deviations which are apt to cause 
disagreement. Whether we want it or not, the proposed standardization is to 
be the basis for comparing different equipment. Of what use are such com 
parisons? The problems which face photogrammetrists are manifold and 
require different solutions. There is no one method for the control of equip 
ment for universal plotting which can be applied to all apparatuses with equal 
success, just as there is probably no one ideal method for aerial traversing with