Some characteristic “trails” obtained by this method of test are shown in Fig. 78. The first trail is that given by a camera rigidly fastened to the fuselage. The second and third show hand camera trails, made by an inexperienced and by an experienced observer, respectively. They show by comparison with the other figures that the human body is an excellent block to vibration, but in unskilled hands a poor check to rapid erratic (probably rotational) motions of the camera. The fourth is the trail given by a camera supported by gimbals bedded in sponge rubber accurately in the plane of the camera's center of gravity. Other trails are shown in the next chapter in connection with the description of practical camera mountings. Clearly the best suspension is that giving the smallest amplitude of displacement during the interval of time covered by an average exposure. It is, in fact, possible to determine from these trails the permissible exposure for any assumed permissible blurring of the image. The rigid mounting trail indicates very bad conditions, calling for literally instantaneous exposures. The center of gravity trail, at the other extreme, shows practically no limitation of exposure in so far as vibration is concerned, thus bearing out the theoretical conditions above discussed. An interesting conclusion from these experiments is that a rapidly running motor gives less harmful vibration than a slow one, although in the war it was a common practice to throttle the motor before exposing. As might be expected, the greater the number of cylinders, the shorter the period and the smaller the amplitude of the vibration. =Pendular Camera Supports.=—The design of the camera support may be approached from a different standpoint, namely, with the aim of carrying the camera so that it will tend to hang always vertical. In mapping this is of fundamental importance. It is, indeed, a question whether aerial mapping will ever be worthy of ranking as a precision method unless the camera can be mounted so that its pictures are taken in the horizontal, undistorted position. The simplest way to hold the camera vertical is to mount it on gimbals, with its center of gravity below the point of support. When so mounted the camera swings as a pendulum. Delicacy of response to variation of level is obtained by leaving a considerable distance between the center of gravity and the center of support. Oscillation about the vertical position is to be prevented by some system of dash pots or other damping. A suspension of this kind is furnished with the Brock film camera (Fig. 60).