This article reviews the author's investigations on the perception and control of spatial relations if the carriers of the relevant sense organs are mobile and controlled independently of each other. In the dragonfly, head rotation is controlled by the head's inertia, as well as by cervicocollic, optokinetic, and dorsal light reflexes and, in turn, controls trunk rotation by means of neck reflexes on the wings. In humans, invariance of head-referenced visual direction under eye-to-head rotation is attained by feedforward of an efference copy. In the pigeon, invariance of responses to trunk tilt under head-to-trunk rotation is, in flight, achieved by feedforward of head-to-trunk information yielded by neck receptors. But in standing or walking, this is accomplished by means of gravity sense organs in the trunk. Such organs are also shown to exist in the human trunk by means of experiments on a sled centrifuge. From tests with paraplegic and neuromectomized subjects, it is concluded that truncal graviception 1) is not influenced by mechanoreceptors in the legs, the skin, and between the vertebrae, but 2) is affected by at least two afferent inputs, one originating in the kidneys, another in the tissues that support the large blood vessels against the gravitational load. These conclusions are corroborated by experiments with bilaterally nephrectomized subjects and by means of positive air pressure to the legs, respectively. Recent results under application of positive and negative air pressure to the entire lower body indicate that yet another source of somatic graviception may exist, for example, one that exploits the hydrostatics of blood pressure or the inertia of the mass of the abdominal viscera.