The proper way to eliminate vibrational effects is to devise a mounting that will transmit only the translational shocks or that will transform the rotational ones into translations. Platforms pivoted and cross-linked so as to be free to move only parallel to themselves (described in the next chapter) represent one attempt to reach this result. Quite the simplest and most scientific form of mounting to achieve this end is to support the camera solely _in the plane of the center of gravity_. The principle here involved is easily grasped if we note that when we jar a camera supported above or below its center of gravity, the effect is to start the camera vibrating with the center of gravity oscillating pendulum-like about the point of support. The closer the center of gravity to the center of support, the smaller the moment of the body about the latter point. =Experimental Study of Methods of Camera Support.=—Conclusive evidence as to the merits of any system of camera mounting can be obtained only under conditions that eliminate the effect of other variables which may be equally efficacious in diminishing the effects of vibration, but which have only limited application. Very brief exposures—1/500 second and less—will, for instance, result in good pictures with almost any condition of vibration. Hence a sharp picture offers no proof of the merits of a camera mounting unless it is known that the exposure was no shorter than the limit set by the ground speed of the plane. In fact it may be said that the chief object of studying methods of camera suspension is to increase the allowable exposure to a maximum, thus lengthening the working hours and multiplying the useful working days for aerial photography. The most satisfactory method of test yet developed is to fly over a light or a group of lights on the ground with the camera shutter open. In the first use of this method, which originated in the English Service, such flights were made at night, but later it was found that good results could be got by placing the lights in a forest and making the tests when the sun was fairly low. One of the group of lights must be periodically interrupted, at a known rate, to furnish the time intervals. [Illustration: FIG. 78.—Tests of camera mounting, made by flying over a bright light against a dark background. (_a_) Rigid fastening on side of plane; (_b_) held in the hand, inexperienced observer; (_c_) held in the hand, experienced observer; (_d_) camera mounted at center of gravity on gimbals bedded in sponge rubber.]