(185) 
Sustained rate of climb or descent will, of course, result in an unaccelerated flight 
on an inclined path. It can be proven that, in such a case, the combination of 
two rates of climb indicators, connected respectively to impact and static pres 
sure, will produce true air speed differentials independent of path inclination. 
The configuration shows that, in order to satisfy a three-minute accuracy for 
longitudinal deflections of the pendulum, it is necessary to measure the rate of 
climb with an accuracy of about 10 feet/minute in a 600-knot aircraft. 
We have seen that the first-order horizontal components of accelerations 
which disturb our true vertical direction in flight can be measured by rate of 
turn in the transverse direction and by rate of climb in the longitudinal direction 
of the aircraft. The accuracies of measurement required are high, but they are 
not prohibitive at the present state of the art. 
To be complete, I should mention that there are also second-order disturb 
ances involved. There are Coriolis and wind accelerations. These secondary 
components, however, are small and easily predictable and do not influence the 
picture very much unless the required accuracy of verticality is increased to 
considerably better than 6 minutes of arc. 
The vertical corrector thus operates in a manner that reduces the indica 
tions of a pendulum in flight by independently measured horizontal components 
of longitudinal and transverse accelerations. The gyro, instead of being moni 
tored directly by the pendulum, is erected to a reference line, which is corrected 
by the measurements of horizontal accelerations and therefore is continuously 
much closer to the true vertical than the direction of the pendulum. The gyro, 
by virtue of its angular momentum, also will fulfill the task of a low-pass filter 
eliminating instantaneous errors of measurement Thus a fairly high noise level 
and a slight time delay of the output of the measuring instruments can be tole 
rated, which reduces conditions imposed upon the output of the measuring in 
struments. 
This system of achieving considerably improved true vertical indication 
from an ordinary horizon gyro, which has been very briefly described is being 
realized and an experimental model will undergo flight tests soon. It is hoped 
that a fairly simple and reliable piece of equipment will be able to furnish ver 
ticality within one- tenth of a degree under photogrammetric flight conditions. 
How soon this will be achieved and how good an accuracy of verticality will 
be finally accomplished cannot be predicted before the equipment has gone 
through extensive flight tests. 
I would not want to close without giving credit and appreciation to 
Mssrs. B. K. Wernicke and H. R. Mestwerdt of the Photo Reconnaissance Lab 
oratory for their decisive part in developing the theory of the described system, 
and to Aeroflex Laboratories, Inc., Long Island City, New York, for adven 
turing into the problem of transferring theory into hardware Anyone familiar 
with the subject knows that there are plenty of problems involved, but we are 
confident that finally they can be solved. 
It has not been my intention to present to you a fully developed piece of 
equipment but rather to arouse your interst in, and your original thinking 
about, a problem which is urgent. It may be of more importance in the mil 
itary field where the factors of simplicity and speed determine the value of 
procedures. I feel, however, that the mentioned development has more chances 
of achieving commercial importance than any other solution to the true vertical