Activity of cells in the lateral vestibular nucleus as a function of head position.

1. The spike activity of cells in the lateral vestibular nucleus was recorded in cats anaesthetized with pentobarbital sodium. Natural labyrinthine stimulation was applied by fixing the animal at different positions reached through roations about a longitudinal or transverse axis.2. The majority of cells responded to rotations only about the longitudinal axis. Two types of response were found. The first was characterized by a transient change in activity which occurred only during the movement. The second type had an initial transient component and a subsequent steady component that persisted as long as the head remained fixed.3. The interspike interval means, standard deviations, histograms and autocorrelograms of the steady response components of cells sensitive to lateral tilt were calculated. In every cell the relation between the head position with respect to gravity and the mean interspike interval of the steady discharge showed two main features. (a) ;Directional sensitivity': the mean interval increased following rotation in one sense, and decreased following rotation in the other. In twenty-two out of thirty-three cells, the mean increased when the recording side was raised. The remaining cells showed the opposite relation. (b) ;Multivaluedness': each particular position is associated with several different values of mean interval and these values had a relatively wide scatter. The curve that resulted from joining points in the order in which they occurred during the experiment was either closed, open, or combined closed and open portions.4. The standard deviations, histograms and autocorrelograms also showed directional sensitivity and multivaluedness with respect to position. Several types of interspike interval histograms and autocorrelograms characterized lateral vestibular activity. The forms of the histogram and the autocorrelogram of the discharge from each cell usually remained unchanged during stimulation.5. The extensive spread of the multivaluedness implies that observation of the mean interspike interval of a specified cell can lead only to a probabilistic statement about head position. Joint observations of mean, standard deviation and histogram, or joint observation of several cells or knowledge of the previous history of the movement may lead to a more accurate determination of position.6. Experiments in animals with only one labyrinth showed that each labyrinth can both accelerate and decelerate activity in both vestibular nuclei.