Control of eye movement reflexes

We investigated the effect of strategic suppression of reflexive eye movements on external control over fixation using a fixation offset paradigm. A visual signal at fixation facilitates the fixation reflex and inhibits eye movements. Certain preparatory states render the fixation reflex less reactive to visual stimulation at fixation, as evidenced by a reduction in the fixation offset effect (FOE). For example, past studies have suggested that the reduced FOE during anti-saccade tasks results from the requirement to inhibit reflexive eye movements. We tested whether suppressing reflexive saccades reduces external control over ocular fixation using a go-nogo saccade paradigm. During each trial, one of two targets appeared in the periphery. Participants were instructed to saccade to one target (go), but when the other target appeared they either had to maintain fixation (nogo) or move their eyes in the direction opposite the target (anti). When nogo trials were admixed with target-directed saccades a large FOE was observed compared to when target-directed saccades occurred alone (experiment 1); however, when anti-saccades were mixed with target-directed saccades, a small FOE was observed for both types of eye movements (experiment 2). We conclude that suppressing reflexive eye movements does not reduce external control over fixation. Further research is necessary to elucidate which other component of preparing to make an anti-saccade diminishes the FOE.     

Latency of saccades during smooth-pursuit eye movement in man Directional asymmetries

To examine the effects of smooth-pursuit eye movements on the initiation of saccades, their latency was measured when subjects initially fixated or pursued a target. In half of the block of trials, the fixation or pursuit target was extinguished 200 ms before the saccade target was illuminated (gap trials). Reduction of the mean saccade latency in the gap trials (the “gap effect”) was evident even when the subjects were pursuing a moving target, consistent with previous observations. The effect of pursuit direction on saccade latency was also examined. Saccades in the same direction as the preceding pursuit (forward saccades) had shorter latencies than those in the opposite direction (backward saccades). This asymmetry was observed in both the gap and nongap trials. Although the forward-backward asymmetry was much smaller than the “gap effect”, it was statistically significant in six of eight cases. These results suggest that the preparation of saccades is affected by smooth-pursuit eye movements. Received: 2 June 1997 / Accepted: 6 November 1997     

Control of fixation and saccades during an anti-saccade task: an investigation in humans with chronic lesions of oculomotor cortex

Fourteen patients with a chronic, unilateral lesion restricted to the frontal lobe (twelve involving the frontal eye field (FEF)), nine patients with a chronic, unilateral lesion restricted to posterior association cortex (eight involving the intraparietal sulcus (IPS)), and twelve neurologically normal control subjects were studied in an anti-saccade task. A combination of manipulating cuing and fixation offset enabled us to examine the effects of chronic oculomotor lesions on both saccade preparation and voluntary control over ocular fixation. Patients with lesions of the FEF made more errors (reflexive glances) toward contralesional targets, whereas patients with IPS lesions made fewer errors toward contralesional targets. Patients with IPS lesions had increased latencies to initiate saccades away from contralesional targets. For FEF patients, the presence of a fixation point inhibited the initiation of contralesionally directed saccades less than those directed ipsilesionally. Saccade preparation in response to a cue did not reduce the inhibitory effect of a fixation point on initiating anti-saccades directed either ipsilesionally or contralesionally for either patient group. We conclude that chronic IPS lesions result in a reduced contralesional visual grasp reflex (VGR) and delayed utilization of visual signals in the contralesional field for planning voluntary eye movements. In contrast, patients with chronic FEF lesions are impaired in inhibiting the VGR toward contralesional signals, and manifest an asymmetry in the balance between fixation and saccade activity. Moreover, voluntary control of fixation is compromised after chronic damage to either frontal or parietal oculomotor cortex.     

Effects of stimulus conditions on the performance of antisaccades in man

 We investigated the effect of different spatial and temporal parameters on the saccadic reaction times (SRTs) of the antisaccades and on the frequency and the SRTs of erratic prosaccades in five adult human subjects. The subjects were instructed to aim their saccades to the side opposite to where a visual go-stimulus occurred. Parameters under consideration were: the gap duration (between 0 and 600 ms, and an overlap paradigm); the stimulus size (sizes of 0.1°, 0.2°, and 0.4°, using the gap 200-ms paradigm); and the stimulus eccentricity (1°, 2°,4°, 8°, and 12°, with the gap 200-ms paradigm). A decrease in the anti SRTs and an increase in the error rate were observed with medium gap durations (200 ms, 250 ms), while the anti-SRTs were longer and the error rates lower with the shorter values (0 ms, 100 ms, and with the overlap paradigm) and with the long values (600 ms). A slight decrease in the anti-SRTs and an increase in the error frequency occurred with increasing eccentricity; the SRT distributions of the errors resembled closely those of prosaccades in corresponding prosaccade tasks with the same eccentricities. The stimulus size had only modest or no effects at all. Analysis of the distributions of the correction times of the erratic prosaccades showed that the intersaccadic intervals could be very short: in the range of express saccades, with a peak at 100 ms; or in some subjects even shorter, with a peak at 40–50 ms. It is concluded that the performance of antisaccades is influenced by parameters that interact with the fixation and/or attention system of oculomotor control. Parameters supporting a disengagement of fixation at the time of stimulus onset provoke a reduction of the saccadic reaction times not only of prosaccades but also of antisaccades. Moreover, a certain state of disengagement seems to facilitate the occurrence of reflex-like errors. Received: 3 July 1996 / Accepted: 29 March 1997     

The gap effect and express saccades in the auditory modality

Latencies of eye movements to peripheral targets are reduced when there is a short delay (typically 200 ms) between the offset of a central visual fixation point and the target onset. This has been termed the gap effect. In addition, some subjects, usually with practice, exhibit a separate population of very short latency saccades, called express saccades. Both these phenomena have been attributed to disengagement of visual attention when the fixation point is extinguished. A competing theory of the gap effect attributes it to disengagement of oculomotor fixation during the temporal gap. It is known that auditory targets are effective in eliciting saccadic eye movements, and also that covert attention operates in the auditory modality. If the gap effect and express saccades are due to disengagement of spatial attention, both should persist in the auditory modality. However, fixation of gaze is largely under visual control. If the gap effect results from disengagement of fixation, then at least a reduced effect should be seen in the auditory modality. Human subjects performed the gap task and a control task in the dark, using auditory fixation points and saccadic targets, on five successive days. Despite this practice, express saccades were not observed. There was a reliable gap effect, but the reduction in saccadic latency was only 17 ms, compared with 32 ms for the same subjects in the visual modality. This suggests that about half the gap effect is due to disengagement of visual fixation. The remainder was not due to non-specific warning effects and could be attributed to offset of the auditory fixation stimulus. Received: 1 March 1996 / Accepted: 11 July 1997     

Event-related potentials and saccadic reaction times: effects of fixation point offset or change

Previous studies have shown that saccadic reaction times (SRTs) are reduced if the initial fixation point (FP) disappears 200 ms (gap period) before a peripheral target is presented. This gap saccade task is associated with a negative cortical potential at the end of the gap period. To determine whether the neural processes underlying this potential account for the reduction of SRTs during gap saccade tasks, we recorded event-related potentials (ERPs) in 19 subjects performing a gap saccade task (gap duration 200 ms), a warning saccade task (the color of the FP changed 200 ms prior to target appearance) and an overlap task (the FP remained visible during the trial). SRTs were shortest during the gap task, longest during the overlap task and intermediate during the warning task. The gap and warning tasks were accompanied by the same widespread negative cortical potential with a maximum at the time of stimulus presentation. These findings indicate that the warning effect mediated by the disappearance of the FP during gap saccade tasks is responsible for the gap negativity which was observed by several authors. Our findings of shorter SRTs during the gap task than the warning task, however, suggest that the gap has an additional effect that probably depends on subcortical mechanisms. Received: 01 June 1998 / Accepted: 12 March 1999     

Effects of warning signals and fixation point offsets on the latencies of pro- versus antisaccades: implications for an interpretation of the gap effect

The present study was designed to evaluate whether fixation point offsets have the same effects on the average latencies of prosaccades (responses towards target) and antisaccades (responses away from target). Gap and overlap conditions were run with and without an acoustic warning signal. The gap effect was taken to be the difference in mean reaction time between gap and overlap trials. This effect was dramatically reduced by the presentation of the warning signal. Without this signal, fixation offsets can serve as warning signals themselves, which artifactually inflates the magnitude of the gap effect. The warning effect of fixation offsets was equivalent for pro and antisaccades. A significant gap effect is still evident with the acoustic warning signal; however, in this case it is associated primarily with prosaccades. These results replicate and extend our previous work demonstrating that, if their warning effects are controlled, the facilitatory effects of fixation point offsets are response dependent, and suggesting the existence of a component process (fixation release) which is closely linked with the processing architecture underlying target-directed saccades.Portions of this work were presented at the annual meeting of the Association for Research on Vision and Ophthalmology, Sarasoto, 1993     

Cortical control of saccades

Saccadic eye movements are controlled by a cortical network composed of several oculomotor areas that are now accurately localized. Clinical and experimental studies have enabled us to understand their specific roles better. These areas are: (1) the parietal eye field (PEF) located in the intraparietal sulcus involved in visuospatial integration and in reflexive saccade triggering; (2) the frontal eye field (FEF), located in the precentral gyrus, involved in the preparation and the triggering of purposive saccades; and (3) the supplementary eye field (SEF) on the medial wall of the frontal lobe, probably involved in the temporal control of sequences of visually guided saccades and in eye-hand coordination. A putative cingulate eye field (CEF), located in the anterior cingulate cortex, would be involved in motivational modulation of voluntary saccades. Besides these motor areas, the dorsolateral prefrontal cortex (dlPFC) in the midfrontal gyrus is involved in reflexive saccade inhibition and visual shortterm memory.     

On the gap effect for saccades evoked by electrical microstimulation of frontal eye fields in monkeys

To investigate the mechanisms of fixation disengagement and saccade initiation, we electrically stimulated the macaque frontal eye fields (FEF) while monkeys performed a visual fixation task. We tested the effect of introducing a temporal gap between fixation target offset and the onset of the electrical stimulus. We found that the duration of the gap had a pronounced effect on the probability of producing electrically evoked saccades at a given current level. The highest probability was found for gaps of 200 ms duration. There were also effects of gap duration on saccade latency and amplitude for most of the stimulation sites. The increase in saccade probability may be associated with lower current thresholds for evoking saccades.     

Representation of averaging saccades in the superior colliculus of the monkey

We tested the hypothesis that averaging saccades occur when two different saccades are prepared and executed simultaneously. The activity of saccade-related burst neurons (SRBNs) in the primate superior colliculus was recorded while monkeys made both non-averaging saccades to single targets and averaging saccades which directed the gaze between two simultaneously presented visual targets. For movements of comparable direction and amplitude, the activity measured during averaging and non-averaging saccades was statistically indistinguishable. These results are not consistent with the hypothesis that averaging saccades result from the simultaneous execution of two different saccades at the level of the collicular SRBNs. Instead, these findings indicate that averaging saccades are represented as single intermediate movements within the topographically organized map of these collicular cells.     

Intrasaccadic target steps during the deceleration of primary saccades affect the latency of corrective saccades

This study investigates how visually guided saccades and subsequent corrective saccades are affected by a secondary target step occurring at different times during the primary saccade. Eye movements of human subjects were measured by means of a differential infrared light reflection technique while the subjects performed visually guided saccades to a laser spot in darkness. The target was stepped backward or onward during the targeting saccade. While the intrasaccadic target step did not influence gain, peak velocity or skewness of the primary saccade, it had a significant effect on the subsequent corrective saccade when the secondary target step occurred during the deceleration phase of the primary saccade: the latency of the corrective saccade was significantly increased compared with the one performed under the single-step control condition. This increase also occurred when single target steps were presented randomly intermixed with backward and onward double target steps and even between selected sub-samples of saccades with identical postsaccadic visual error. If the target step occurred early during the primary saccade, the latency of the corrective saccade was not changed. This indicates that visual information sampled during the deceleration phase of a saccade can lead to a cancellation of the normal trigger mode of corrective saccades. Received: 9 April 1999 / Accepted: 23 June 1999     

Amplitude of human head movements associated with horizontal saccades

 Human saccades may or may not be associated with head movements. To date, little attention has been devoted to the mechanisms determining head movement recruitment and scaling. Normal human subjects made horizontal, centrifugal saccades along an encircling array of light-emitting diodes. Measurements of gaze, head, and eye-in-head angle were made at the conclusion of the head movement (or at the end of the eye movement in eye-only saccades). We found that head movement amplitude (ΔH) related in a simple fashion to the eye eccentricity that would have resulted if the gaze shift had been performed without a head movement. Plots of ΔH vs this predicted eye eccentricity (E _PRED) had a central flat region in which gaze shifts were unaccompanied by head movements (the eye-only range) and two flanking lobes in which ΔH was a linear function of E _PRED (the eye-head ranges). ΔH correlated with E _PRED better than with gaze shift amplitude, as would be expected if head movements were controlled so as to keep eye eccentricity within a particular range. Head movement tendencies were quantified by the width of the eye-only range, the slope of the eye-head range, and the width of the region within which the eye was likely to be found at the conclusion of the completed gaze-shifting behavior (the customary ocular motor range). The measures ranged widely in these normal subjects: 35.8±31.9° for the eye-only range (mean±SD), 0.77±0.16 for the slope of the eye-head range, and 44.0±23.8° for the customary ocular motor range. Yet for a given subject, the measurements were reproducible across experimental sessions, with the customary ocular motor range being the most consistent measure of the three. The form of the ΔH vs E _PRED plots suggests that the neural circuitry underlying eye-head coordination carries out two distinct functions – gating the head movement and scaling the head movement. The reason for the large intersubject variability of head movement tendencies is unknown. It does not parallel intersubject differences in full-scale eye (in orbit) range or full-scale neck range. Received: 25 June 1998 / Accepted: 23 November 1998     

Visual responses with and without fixation: neurons in premotor cortex encode spatial locations independently of eye position

 The ventral premotor cortex (PMv) of the macaque monkey contains neurons that respond both to visual and to tactile stimuli. For almost all of these “bimodal” cells, the visual receptive field is anchored to the tactile receptive field on the head or the arms, and remains stationary when the eyes fixate different locations. This study compared the responses of bimodal PMv neurons to a visual stimulus when the monkey was required to fixate a spot of light and when no fixation was required. Even when the monkey was not fixating and the eyes were moving, the visual receptive fields remained in the same location, near the associated tactile receptive field. For many of the neurons, the response to the visual stimulus was significantly larger when the monkey was not performing the fixation task. In control tests, the presence or absence of the fixation spot itself had little or no effect on the response to the visual stimulus. These results show that even when the monkey’s eye position is continuously changing, the neurons in PMv have visual receptive fields that are stable and fixed to the relevant body part. The reduction in response during fixation may reflect a shift of attention from the visual stimulus to the demands of the fixation task. Received: 8 April 1997 / Accepted: 16 July 1997     

Independent contributions of the orienting of attention,fixation offset and bilateral stimulation on human saccadic latencies

In a series of experiments we examined the effects of the endogenous orienting of visual attention on human saccade latency. Three separate manipulations were performed: the orienting of visual attention, the prior offset of fixation (gap paradigm) and the bilateral presentation of saccade targets. Each of these manipulations was shown to make an independent contribution to saccade latency. In experiments 1 and 2 subjects were instructed to orient their attention covertly to a location by a verbal pre-cue; targets could appear in the attended hemifield (valid) or in the non-attended hemifield (invalid) together with a no-instruction (neutral) condition. Saccades were made under fixation gap and overlap conditions, to either single targets or two bilaterally presented targets which appeared at equal and opposite eccentricities in both hemifields. The results showed a large increase (cost) of saccade latency to invalid targets and a small non-significant decrease (benefit) of saccade latency to valid targets. The cost associated with invalid targets replicates the meridian crossing effect shown in manual reaction time experiments and is consistent with the hemifield inhibition and premotor models of attentional orienting. The use of a gap procedure produced a generalised facilitation of saccade latency, which was not modified by the prior orienting of visual attention. The magnitude of the gap effect was similar for saccades made to attended and non-attended stimulis. This suggests that the gap effect may be due to ocular motor disengagement, or a warning signal effect, rather than to the prior disengagement of visual attention. When two targets were presented simultaneously, one in each hemifield, saccade latency was slowed compared with the single target condition. The magnitude of this slowing was unaffected by the prior orienting of visual attention or by the fixation condition. The slowing was examined in more detail in experiment 3, by presenting targets with brief offset delays. The latency increase was maximal if the two targets were presented simultaneously and decreased if the distractor appeared at short intervals (20–80 ms) before or after the saccade target onset. If the non-attended stimulus was presented at greater intervals (160, 240 ms) before the saccade target, then a facilitation effect was observed. This demonstrates that the onset of a distractor in the non-attended hemifield can have both an inhibitory and a facilitatory effect on a saccade production.     

Accuracies of saccades to moving targets during pursuit initiation and maintenance

The overall goals of the studies presented here were to compare (1) the accuracies of saccades to moving targets with either a novel or a known target motion, and (2) the relationships between the measures of target motion and saccadic amplitude during pursuit initiation and maintenance. Since resampling of position error just prior to saccade initiation can confound the interpretation of results, the target ramp was masked during the planning and execution of the saccade. The results suggest that saccades to moving targets were significantly more accurate if the target motion was known from the early part of the trial (e.g., during pursuit maintenance) than in the case of novel target motion (e.g., during pursuit initiation); both these types of saccades were more accuate than those when target motion information was not available. Using target velocity in space as a rough estimate of the magnitude of the extra-retinal signal during pursuit maintenance, the saccadic amplitude was significantly associated with the extra-retinal target motion information after accounting for the position error. In most subjects, this association was stronger than the one between retinal slip velocity and saccadic amplitude during pursuit initiation. The results were similar even when the smooth eye motion prior to the saccade was controlled. These results suggest that different sources of target motion information (retinal image velocity vs internal representation of previous target motion in space) are used in planning saccades during different stages of pursuit. The association between retinal slip velocity and saccadic amplitude is weak during initiation, thus explaining poor saccadic accuracy during this stage of pursuit.     

Attentional facilitation of response is impaired for antisaccades but not for saccades in patients with schizophrenia: implications for cortical dysfunction

The facilitation of response known as the gap effect (a decrease of response latency), observed for saccades and antisaccades when attention is modulated prior to such eye movements, was studied in patients with schizophrenia and in controls. The hypothesis tested was whether patients would show a deficient attentional facilitation in response latency. Fifteen patients with schizophrenia and 17 healthy controls performed blocks of saccades and antisaccades in a gap condition and an overlap condition. Saccade and antisaccade response latencies as well as the error rate for antisaccades were measured for each subject. A similar gap effect (decrease in latency for the gap compared to the overlap condition) was present in the saccade task for patients and controls. In contrast the gap effect in the antisaccade task was absent in 50% of patients compared to none of the controls. This finding was interpreted as indicative of deficient preprocessing in antisaccade-specific cortical areas in schizophrenia (such as the prefrontal cortex), while visually guided saccade processing remained intact. Our results, in addition to many other recent findings, could lead to specific hypotheses on cortical dysfunction in schizophrenia.     

The effects of bottom-up target luminance and top-down spatial target predictability on saccadic reaction times

Saccadic reaction times (SRT) are composed of the sum of multiple processes, including bottom-up sensory processing, top-down goal oriented processing, and afferent and efferent conduction delays. In order to determine the timing dependencies and potential interactions between bottom-up and top-down processes on SRTs, we trained monkeys to perform several variants of visually guided saccade tasks. Bottom-up components of SRT were manipulated by varying target luminance from near detection threshold to supra-threshold ranges (i.e., 0.001–42.5 cd/m²). There was a significant reduction of mean SRT with increases in target luminance up to 3.5–17.5 cd/m². Luminance increases above these ranges produced significant increases in SRT when the target was within 6° from the fovea. Top-down components were assessed by manipulating spatial target predictability across blocks of trials using either 1, 2, 4 or 8 possible target locations. Decreasing spatial target predictability increased SRT across target luminances from 1 to 4 targets in the gap task, but then paradoxically decreased SRT again when there were 8 possible targets in both the gap and step tasks. Finally, a gap task (200 ms gap) was used to determine the dependence of target luminance on the magnitude of the gap effect. Decreasing target luminance significantly reduced the magnitude of the gap effect indicating that the gap effect is strongly influenced by bottom-up factors.     

Effect of eye movements on the magnitude of functional magnetic resonance imaging responses in extrastriate cortex during visual motion perception

We have studied the effects of pursuit eye movements on the functional magnetic resonance imaging (fMRI) responses in extrastriate visual areas during visual motion perception. Echoplanar imaging of 10–12 image planes through visual cortex was acquired in nine subjects while they viewed sequences of random-dot motion. Images obtained during stimulation periods were compared with baseline images, where subjects viewed a blank field. In a subsidiary experiment, responses to moving dots, viewed under conditions of fixation or pursuit, were compared with those evoked by static dots. Eye movements were recorded with MR-compatible electro-oculographic (EOG) electrodes. Our findings show an enhanced level of activation (as indexed by blood-oxygen level-dependent contrast) during pursuit compared with fixation in two extrastriate areas. The results support earlier findings on a motion-specific area in lateral occipitotemporal cortex (human V5). They also point to a further site of activation in a region approximately 12 mm dorsal of V5. The fMRI response in V5 during pursuit is significantly enhanced. This increased response may represent additional processing demands required for the control of eye movements. Received: 16 July 1997 / Accepted: 14 October 1997     

Peak velocities of visually and nonvisually guided saccades in smooth-pursuit and saccadic tasks

 Smooth pursuit typically includes corrective catch-up saccades, but may also include such intrusive saccades away from the target as anticipatory or large overshooting saccades. We sought to differentiate catch-up from anticipatory and overshooting saccades by their peak velocities, to see whether the higher velocities of visually rather than nonvisually guided saccades in saccadic tasks may be found also in saccades in pursuit. In experiment 1, 12 subjects showed catch-up, anticipatory, and overshooting saccades to comprise 70.4% of all saccades in pursuit of periodic, 30°/s constant-velocity targets. Catch-up saccades were faster than the others. Saccadic tasks were run as well, on 19 subjects, including the 12 whose pursuit data were analyzed, with target-onset, target-remaining (saccade to the remaining target when the other three extinguish), and antisaccade tasks. For 17 of the 19 subjects, antisaccade velocities were lower than for either target-onset or target-remaining tasks. Velocities for the target-remaining task were near those for target onset, indicating that target presence, not its onset, defines visually guided saccades. Error and reaction-time data suggest greater cognitive difficulty for target remaining than for target onset, so that the cognitive difficulty of typical nonvisually guided saccade tasks is not sufficient to produce their lowered velocity. To produce reliably, in each subject, catch-up and anticipatory saccades with comparable amplitude distributions, nine new subjects were asked in experiment 2 to make intentional catch-up and anticipatory saccades in pursuit, and were presented with embedded target jumps to elicit catch-up saccades, all with periodic target trajectories of 15°/s and 30°/s. Velocities of intentional anticipatory saccades were lower than velocities of intentional catch-up saccades, while velocities of intentional and embedded catch-up saccades were similar. Target-onset and remembered-target saccadic tasks were run, showing the expected higher velocity for the target-onset task in each subject. Both experiments demonstrate higher peak velocities for catch-up saccades than for anticipatory saccades, suggesting that cortical structures preferentially involved in nonvisually guided saccades may initiate the anticipatory and overshooting saccades in pursuit. Received: 1 December 1995 / Accepted: 25 February 1997     

Visual perceptions of vertical and intrinsic longitudinal axes

 The purpose of these experiments was to investigate whether visual perceptions of the earth-fixed vertical axis are more accurate than those of intrinsic body-fixed axes. In one experiment, nine neurologically normal young adult subjects’ abilities to position a luminescent rod vertically or parallel to the longitudinal axis of the head or trunk were studied in four conditions: (1) earth-fixed – subjects stood erect with the head aligned to the trunk and visually aligned a hand-held rod to vertical; (2) earth – subjects aligned the rod to vertical as in 1, but the orientations of the head and trunk were varied in the sagittal and frontal planes on each trial; (3) head – frontal and/or sagittal plane orientation of the subject’s head was varied on each trial and the rod was aligned parallel to the longitudinal axis of the head; (4) trunk – frontal and/or sagittal plane orientation of the subject’s trunk was varied on each trial and the rod was aligned parallel to the longitudinal axis of the trunk. Note that in conditions 2, 3, and 4 the head and trunk were never aligned with each other. Also, each condition was carried out in normal light and in complete darkness. Perceptual errors were measured in both the frontal and the sagittal planes. The results showed that the variable errors were significantly lower when subjects aligned the rod to vertical rather than to the longitudinal axis of the head or trunk. Also, errors were similar in size in the two planes and were unaffected by vision of the surrounding environment. In a second experiment, subjects were seated and controlled the position of a luminescent rod held by a robot. They aligned the rod either to the longitudinal axis of their head or to the vertical in complete darkness, under three conditions similar to those described above: (1) earth-fixed, (2) earth, and (3) head. There was no possibility of use of kinesthetic information for controlling rod position in this experiment as in the first experiment. The results were similar to those of the first experiment, as subjects aligned the rod more accurately to vertical than to the longitudinal axis of the head. These results show convincingly that visual perceptions of earth-fixed vertical are more accurate than perceptions of intrinsic axes fixed to the head or trunk. Received: 17 February 1997 / Accepted: 28 April 1997