11. Future Experiments. 
  
We will first discuss the value of International Controlled Photogrammetric Experiments, 
Photogrammetry is an empirical science and so it could only by studied with observation and repeated tests of hypo- 
thesis. The mathematical model of photogrammetry and its reliability in practice must be the principal problem. 
  
The mathematical model used in photogrammetric theory is ordinarily assumed to have normal error distribution 
  
with homogeneous variance, no gross errors and mostly also no systematic errors. It is treated according to the method 
ofleast squares. The interest is usually concentrated upon some single partial error or upon small groups of few par- 
tial errors. Photogrammetric practice from the consumers point of view must deal with the total error, which includes 
  
all systematic and partial stochastic errors. The International Controlled Photogrammetric Experiments made it possi- 
  
ble to assemble so many empirical data, that photogrammetric errors could be studied sucessfully with the aid of 
modern statistics. They were alternatives to the usual theoretical way of studying photogrammetric errors. They 
opened new ways of tackling problems and gave some interesting new results. Among them we mention the Cumula - 
tive pz distributions and the Empirical Photogrammetrie x -distribution. The great number of participants in the 
Experiments must be an indicator of an international interest in the idea. We believe in a future for such Experi - 
ments and in the value of them. 
The following four problems in photogrammetry are recommended for'study with the aid of a great number of 
assembled empirical data. 
11.1. The influence upon accuracy of refraction, light conditions, aerial cameras, the speed and vibration of 
aeroplanes, exposure conditions, the geometrical dimensions of a negative at the time for exposure,during the 
photographic procedure,up to the time for the copying of diapositive, 
11.1 The influence upon accuracy of the number of pass-points. 
€ 2 11.3 The error of distance for lengths shorter than one base length. 
1 11.4 The error transition from model 1 to model 2, using the projection centre when orienting model two to mo- 
del one. 
A fifth problem is 
11.5 A system for the study of the economy of international photogrammetry. This system must be thoroughly in- 
vestigated and discussed at the XII International Congress of Photogrammetry before any economic study will be 
started.