1921-22.] Gyroscope and “ Vertical ” Problem on Aircraft. 307 
the rapidly revolving ball system, each ball, when in the outer compart- 
ment of its slot, possesses very little freedom. In the case of the erector 
proper, which, it is to be remembered, revolves slowl}^ in the direction of 
spin of the gyroscope, the balls are allowed sufficient clearance, when in 
the outside compartments of their slots, to allow of the revolving couple 
h coming into existence 
when required. 
There is little to 
choose between the two 
main forms of erector 
which have been de- 
scribed, the double-slot 
erector and the track 
erector. When the ap- 
paratus is properly set 
Fig. 29. 
1 
of a 
up and balanced, each is capable of supplying an accuracy of 
degree under the conditions which prevail in aircraft. 
In ligs. 29 and 30 is shown, in elevation and plan respectively, one 
method of mounting a camera c or sight as part of a Gray stabiliser. The 
frame / is rectangular, and is pivoted at p^2>2 uprights u, u. Pivoted 
at pjj ^ , within the frame /, is the gyroscope g and erector e. The camera 
c or sight S is pivoted to the frame / at p.^p .^ , and is attached to the gyro- 
scopic system by a pivoted connecting rod c' as shown. This construction 
was convenient in cases where the observer’s cockpit was long and narrow. 
In tig. 31 is shown the arrangement which was adopted for use on 
aeroplanes in the cockpits of which the space in the fore-and-aft direction 
was limited. The gyroscopic system composed of gyroscope g and erector 
e was attached rigidly to the sight S, and the entire system pivoted at 
PiPi to the frame / in such manner that the system was in neutral equi- 
librium with respect to the axis p^Pi . The frame was so pivoted to the 
