The effects of repetition on the post-rotational turning sensation and nystagmus obtained after deceleration from constant angular velocities are described. In a first series of tests decelerative stimuli from constant angular velocities of 60, 30, 15 and 7-50/sec. were given to two groups of subjects but in a different order, and the mean responses obtained from both groups showed that habituation might occur in the course of a single test.In a second series with a different group of subjects, the test was repeated on each of nine consecutive days. The effects of habituation were again seen and there was a progressive decline in the responses which were obtained. The administration of l-hyoscine (0 7 mg. base) 1 hr. before the test on the sixth day produced little additional effect in habituated subjects.POST-ROTATIONAL turning sensation and nystagmus may be readily elicited in the normal individual by sudden deceleration from a constant angular velocity using a turning chair or similar device. This test of labyrinthine function was first described by van Egmond, Groen and Jongkees [1948] and called by them 'Cupulometry' because the measurable responses produced by such impulsive decelerations are caused by the deflection of the cupulae of the horizontal semi-circular canals by the momentum of the endolymph. It was intended for clinical use as a substitute for the Barany test because it allowed the stimulus response relationship of the end organ to be determined with 'physiological' stimuli, which were said to lie between 3 and 600/sec.The duration of the post-rotational turning sensation elicited by the chosen decelerative stimuli was reported by the subject, and the duration of the nystagmic response was estimated directly by an observer in the turning room. According to Hulk and Jongkees [1948] plots of the duration of turning and of the duration of nystagmus against logarithm of turntable speed were linear for impulses of between 3 and 60°/sec.However, in constructing plots of the mean sensation and nystagmus 'cupulogram', Hulk and Jongkees did not give the individual points from which the graphs were derived, nor did they give estimates of scatter. They expressed their results as 'slope values' (1I/l) using a formula developed by Groen and quoted a mean of 7 sec. (range 5-16) for turning sensation and of 11 (range 6-24) for nystagmus. Further, they took the intercept on the abscissa as indicating the threshold of the individual both for subjective turning sensation and for nystagmus, giving a mean of 2-50/sec. for turning sensation (range 0-75-5), and of 8 for nystagmus (range 5-30). The values * This work was carried out with the aid of a grant from the Clinical Studies