Visual compensation for torsional head movements (about a naso-occipital axis) is relatively complex and less well understood than for lateral and vertical head movements (see e.g. Gresty et al., 1992). Torsional eye movements are harder to measure, not amenable to voluntary control, and subject to otolith as well as semicircular canal, proprioceptive and visual influences. As with other forms of movement, much of the compensation that affects subjective percepts is central. We have set up a number of psychophysical tests that do not require direct measurement of torsional eye movements, in a single apparatus.  These combine conventional and novel procedures, and can be used to explore normal and abnormal function with either voluntary or driven head movements.

      The orientation of the subject's head is measured with a potentiometer mounted on a helmet, with its spindle kept at a nearly constant orientation (ñca. 1o) by a light rigid arm and elastic thread attached to the ceiling. Measurements are made every 1ms by computer, using a data acquisition board (Microstar DAP2400-5).  Using half-silvered mirrors, the subject simultaneously views 3 displays in the same optical plane at about 30cm distance. One display consists of two photographic flash tubes, operated under computer control and viewed through slits (ca. 250æm x 12mm) to generate vertical afterimages 5mm above and below a fixation point. A yellow filter is used to absorb blue and ultraviolet components of the flash emission. The flashes can be programmed to occur independently or together, either at set times or set head orientations. The second display is a uniform field that can be illuminated with stroboscope flashes at ca. 4Hz to refresh the afterimages, which are then usually clearly visible (first bright, then dark) for ca. 20s. The third display is an oscilloscope screen with rows of dots forming one or more lines whose orientation is under computer control, with a small (0.6mm) central fixation cross incorporating a pattern of movement that can only be seen when viewed in sharp focus with the fovea.

      The 3 displays can be used in various programmed combinations and sequences.  The subject controls a knob that adjusts either the orientation of the oscilloscope display or the gain relating orientation changes on the oscilloscope to the measurements of head orientation. Using the afterimages in combination with the computer- controlled oscilloscope display it is possible to measure both static and dynamic gains of ocular counter-rolling eye movements. Without using afterimages it is possible to measure the static displacement of the subjective visual vertical (normally in the opposite direction to static head tilt: the M?ller effect) and a dynamic counterpart that can be in the opposite direction: the gain of the orientation changes needed to stabilise the percept of a line during oscillatory head roll (at ca. 1Hz).

Acknowledgement:
The flash circuitry was designed by W. Potter. Supported by the MRC. This study has ethical approval.

Gresty M.A., Bronstein A.M., Brandt T., & Dieterich M. (1992) Brain 115: 647-673