the center of gravitational attraction not to coincide with the center of mass or 
the geometric center. The peak deviation of the vertical resulting from this 
cause is of the order of 6 minutes of arc, but the way in which this enters into 
the system causes the system to deviate from the plumb bob vertical by about 
2/3 minute of arc with the worst course and speed. 
Complete analysis of the errors of the PVRU show that the nature of the 
error to be expected in flight is an 84 minute period deviation from true verti- 
eal with an amplitude less than 3 minutes of arc. Two such systems have been 
constructed and have been tested in the laboratory for environmental and func- 
tional integrity. Under simulated flight conditions the errors have been less 
than 3 minutes of arc. Flight testing will occur in the near future. 
The Aeroflex Photogrammetric Vertical Reference System is a much 
simpler system than the Navigation Systems. This simplicity arises, not from 
some accidental features of the proposed system, but from fundamental differ- 
ences between the requirements of Navigation and the requirements of photo- 
grammetry. If we refer back to the system Block Diagram, we see that no- 
where is there an indication of either geographic position or of integrated dis- 
tance traversed. The latter is implicit in the system, but cannot be made avail - 
able without complete and fundamental changes. Because the equipment neces- 
sary to obtain these numbers need not be present, the system can be much 
simpler. 
The azimuth axis in the vertical system uses an essentially free gyro and 
one of relatively non-stringent requirements. This is due to the fact that we do 
not care in which directions the accelerometers measure, so long as the rate 
> 16 —