Optokinetic and pursuit system: a case report

Behavioural Brain Research, 57 (1993) 21-29 © 1993 Elsevier Science Publishers B.V. All rights reserved. 0166-4328/93/$06.00

21

BBR 01487

Optokinetic and pursuit system: a case report Uwe J. Ilg*, Frank Bremmer, Klaus-Peter Hoffmann Department of Zoology and Neurobiology, University of Bochum, Boehum, Germany (Received 25 January 1992) (Revised version received 6 October 1992) (Accepted 25 May 1993)

Key words: Optokinetic system; Pursuit system; Eye movement; Pretectum; Lesion; Rhesus monkey

Several studies have demonstrated the importance of the pretectal Nucleus of the Optic Tract (NOT) and the Dorsal Terminal Nucleus of the accessory optic system (DTN) for the generation of horizontal optokinetic nystagmus (OKN). Although single unit data from trained rhesus monkey N O T / D T N cells are available it is still unclear if there is a link between the pursuit and the optokinetie system at this level of motion analysis. In order to address the question whether the N O T / D T N is important for the optokinetic as well as the pursuit system an electrolytic lesion was placed where N O T / D T N activity was recorded previously. The monkey was tested on optokinetic and pursuit paradigms. Immediately following the lesion the monkey performed a spontaneous nystagmus with slow phases directed away from the lesioned side. This spontaneous nystagmus persisted even during optokinetic stimulation in the opposite direction. During the first week postlesion the spontaneous nystagmus disappeared and the monkey regained the ability to perform optokinetic nystagmus toward the lesioned side. The gain of the mean slow phase eye velocity was, however, largely reduced for this stimulus direction. The onset of OKN following the onset of optokinetic stimulation was not affected by the lesion. During smooth pursuit the mean eye velocity was more reduced for pursuit towards the lesioned side. The resulting position error was compensated by an increase in the number of catch-up saccades. In addition to the confirmation of the wellknown directional deficits of the optokinetic system caused by a lesion of the pretectum, a directional deficit in the pursuit system was demonstrated.

INTRODUCTION

Several studies in various animals provide evidence for the participation of the pretectal N O T / D T N complex in the gaze stabilization reflex (for review see33). In primates the N O T / D T N complex has been examined by microelectrode recordings in paralyzed and anaesthetized monkeys 9, by electrical stimulation experiments 3~, by lesion studies 14'15'32, and by single unit recordings in awake and trained rhesus monkeys 1o,~,28. In accordance to previous work there is no possibility to distinguish between N O T and D T N neurons 9"32. N O T / D T N neurons could be shown to respond directionally selective with ipsiversive preferred direction. The velocity tuning of N O T / D T N neurons for retinal image motion displays a maximum at 26°/s 1° which is close to a proposed maximum in sensitivity for models of the optokinetic system 25. It therefore seems reasonable to suggest that N O T / D T N neurons provide retinal slip information to the optokinetic system. By ex*Corresponding author. Abt. Neurologie, Neurologische Universitatsklinik, Hoppe-Seyler-Str. 3, D-72076 Tabingen, Germany.

amining the activities of N O T / D T N neurons during optokinetic afternystagmus (OKAN) it has been excluded that the N O T / D T N neurons already carry a non-visual eye velocity signal 1°. Although the functional role of N O T / D T N neurons is clear with respect to the optokinetic system, the importance of these neurons for the smooth pursuit system is quite unclear on the basis of single unit data. In rhesus monkeys trained for smooth pursuit eye movements N O T / D T N neurons respond to retinal slip of the pursuit target 1°'11'28. Again, neurons do not carry an eye movement signal, as proved by the blink paradigm 28. If the monkey performs pursuit across a structured visual background, some neurons respond to the target induced slip (type I neurons) whereas others respond to the passive movement of the background induced by the eye movement (type II neurons) 11. The latter are excited if the monkey performs contraversive pursuit eye movements, an apparent inversion of the preferred direction. This apparent inversion is due to the direction of passive background movement opposite to the direction of eye movement. In addition to this dichotomy of neuronal responses the parameters of eye

22 movements are affected by the structure of the background: smooth eye velocity is reduced and the number of saccades is increased when the monkey has to track a target moving across a structured background~,, ~1,~ As a consequence of the existence of two types of N O T / D T N responses during smooth pursuit eye movements the question arises whether these slipencoding N O T / D T N neurons provide an input signal to the pursuit system as well. The lesion studies already published 14'~5'32 were done with monkeys that had not been trained on a pursuit paradigm and therefore did not address the question whether the N O T / D T N complex is involved in the processing of retinal slip information for the initiation or maintenance of smooth pursuit eye movements. In one study a raisin was used as pursuit target ~4. The tracking of this raisin was not affected by the pretectal lesion. To address the relation between the N O T / D T N and pursuit eye movements in a monkey trained on a pursuit paradigm, a pretectal lesion which caused clear deficits in the optokinetic system was produced. In addition to the already described optokinetic deficits 14,15,32 a directional deficit in the pursuit system was demonstrated.

MATERIALS A N D M E T H O D S

A male rhesus monkey (M. mulatta, 6.5 kg) was trained on a fixation paradigm 26"39 in a training set-up using a C-64 home computer. As soon as the monkey learned to release the touch-bar within 500 ms after the fixation target has been dimmed the monkey underwent surgery. Two eye coils, headholder, and recording cylinders were implanted aseptically under pentobarbital anaesthesia 7 (Nembutal 60 mg/ml, Sanofi). A recording cylinder was placed stereotactically at anterior 4 mm on the midline. During the experiments the monkey was seated in a primate chair with the head fixed. Single unit responses were recorded using tungsten-in-glass microelectrodes within a transdural guiding-tube. The reconstruction of an electrode track is shown in Fig. 1 based on a stereotactic a t l a s 36 and single unit response properties. Following the localization of the N O T / D T N based on ipsiversive directional responses an electrolytic lesion was produced in the right pretectum at the recording site. Anodal current pulses (5 times 1-s pulses at 1 mA) were applied. Eye position was determined using the magnetic search coil technique (Eye Position Meter, S k a l a r ) 13"3°. Spatial resolution was 2 min of arc, temporal cut-off was 250 Hz, and sampling rate was 500 Hz. The ex-

~/~@|.~"~/'electr°de

tlac~

/ " \ Inferior

Pulvlnar

-

k

f~-~

direction selective

responle

°'n='2:m'2°

Fig. 1. Reconstruction of an electrode track. In a frontal section from a stereotactic atlas 36 at anterior 2 mm the electrode track is drawn at the specified stereotactic location. Single unit response types are indicated as they were obtained during microelectrode penetration.

O0

Fig. 2. The size of the pretectal lesion. Three Nissl stained frontal sections of the midbrain are shown. The distance between the anterior section and the middle section is 1,200 t~m, the distance between the middle and the posterior section is 800/~m. Note that the lesion extends into the putvinar, but does not affect the fibers to the colliculus superior running in the brachium.

23 step ramp paradigm. Following one second of fixation the central target was turned off. Simultaneously the pursuit target was turned on at 10 ° either left, right, up, or down and moved for two s at 15°/s in centripetal direction. The target moved in front of homogeneous or structured background (see above). The sequence of trials was random. The data files contained at least 10 trials with identical target movement. During the trial the eye position was tracked by the computer and as soon as the gaze deviated from the target, the trial was aborted. Only data of successful trials were stored. The processing of pursuit data included saccade elimination and mean smooth eye velocity calculation during three different phases: first during the intial fixation period, second for the first 200 ms of pursuit, and third for the remaining 1.8 s of steady-state pursuit. Only data sets with mean eye velocity larger than 10 °/s in the steady-state pursuit in at least one direction were considered in this study. The animal was sacrificed at day 76 postlesion by pentobarbital overdose and perfused transcardiacally. The brain was removed from the skull, frozen at -70 ° Celcius in isopentane, cut at 50-/~m sections, and Nissl stained (for detailed description see ref. 9). Wherever statistical procedures were necessary we used the distribution free Wilcoxon Rank Test a.

periments were controlled by a laboratory computer CED1401 connected to a personal computer COMPAQ 386/20 equipped with data acquisition and data analysis (DADA) software. Visual stimuli were back-projected onto a tangent screen extending 90 ° × 90 °. The viewing distance was 35 cm. The background was generated by a random dot pattern with a dot size of 1 ° and a luminance of the bright dots of 20 cd/m 2. The position of the background was controlled by a 2D galvanometer system (General Scanning). To elicit horizontal O K N the background was moved at constant velocity in a given direction. The duration of data collection was 20 s, either 20 s of O K N or 10 s O K N and successive 10 s OKAN. The fixation and pursuit targets were also backprojected onto the screen. The location of the fixation target, straight ahead, was fixed to the center of the screen. The position of the pursuit target was controlled by a second 2D galvanometer system. All galvanometers were driven by digital-to-analog converters of the CED1401 controlled by the D A D A software. Eye velocity was obtained using digital filter technique. Saccades either as quick-phases during O K N or during pursuit were detected and removed automatically by an eye acceleration threshold algorithm. In the eye velocity channel the detected saccades were replaced by the mean eye velocity 50 ms prior to the saccade. The frequency of quick-phases was used to estimate the frequency of either spontaneous or visually elicited nystagmus. Mean slow phase eye velocity of O K N was obtained by calculating the mean of slow phase eye velocity over 20 s. Smooth pursuit eye movements were elicited using a

RESULTS

Fig. 2 shows the extent of the pretectal lesion in three Nissl stained sections. The lesion affected the entire pretectum and parts of the pulvinar dorsal to the preI00.0

right

(~

~m 80.0 60.0 40.0

20.0 right

melt up

'~0 )

'-20.0 J'=="

40

do

s01

~00

aa~uw ~

~ ' -40.0

'-60.0 '

-80.0 -I00.0

2000

4000

6000

8000

~.0000 12000 J4000 18000 ).8000 I

Fig. 3. Spontaneous nystagmus. A shows an example of 20 s spontaneous nystagrnus. Horizontal and vertical eye position are displayed. In B the mean eye velocity and standard deviation for different stimulus velocities is plotted. Note that the lesion is located in the right hemisphere producing a spontaneous nystagrnus with slow phases towards left. This spontaneous nystagmus persisted even when optokinetic stimulation was applied in opposite direction (rightward).

24 i.O

tectum. The brachium of the colliculus superior was not affected.

Gain

0.B

Optokinetic nystagmus

0.

0.

~-'~

~

Immediately after the lesion a spontaneous nystagmus with slow phases directed towards the nonlesioned side was observed. The mean eye velocity of this spontaneous nystagmus was 9.7°/s in darkness. When a structured background was presented, the mean eye velocity of spontaneous nystagmus was reduced to 7.4 °/s. These data represent mean values for three data collections of 20 s each at the lesion day. The spontaneous nystagmus persisted even when an optokinetic stimulation was applied in the opposite direction.

pre leelon

0.2

~

poet lesion

dght

len i

-t00

J

~

l

l

-B0

-60

-40

-20

i

0

i

20

i

J

i

40 60 80[*/~ec| lalmuluevek~clty

Fig. 4. Pre- and postlesion gain of the optokinetic system. Mean gain (eye velocity divided by stimulus velocity) and standard deviation for 5 experimental sessions (20 s each) are shown for different stimulus velocities. The slow phases towards the lesion (right) were decreased following the lesion.

®~

OKAN

OKN

Day 0

OKN

'I~

OKAN

], I. I' ,,,,g',,~o.[

10

$

el

~position

80

BO

'In

40

0

40

0

['/~l 40

40

F/eec] 60 2000

4000

6000

OKN

BOOO

10000 ,~ooo 14000 ~ooo 1~ooo ~:0000

ipslverslve

80

2000 4000

ms

OKAN

6000

,

10000

contmverslve

OKN

Day 49

B000

i

,

I2000

14000

16000

18000

20000

ms

OKAN

hI

f

10

,

•

~i~

80

40

E

'°

i

: ;

i

0

['/Ncl 4O 80

2000 no00 6000

,,!

!

8000 10000 I2000 14000 16000 18000 20000 ms

~o

2000 4000 6000 8 0 0 0

0000 12000 14000 16000 18000 20000 ms

Fig. 5. Recovery of OKAN. Horizontal eye position and velocity are shown for 10 s optokinetic stimulation (following 1 min stimulation) and 10 s of O K A N in darkness. In A and B the experiment was done immediately following the lesion (day 0), in C and D the experiment was done at day 49 postlesion. In A and C the optokinetic stimulation was towards the lesioned side (fight), in B and D the stimulation was away from the lesioned side (left). At day 49 rightward and leftward OKAN could be observed.

25

®

®

pre lesion

eye velocity iO0

right

r'lwl

right

I

'80

I "

60

-

' 4n I

t

len

I

.. I pre leslon

. ~ ' - ' ~ . . . . . . . ~ ' - " ]

I

right

left

-ioo

I I

I

right

I

post lesion

ao ' -20

40

60

B0r/secl~00 stimulus velocity

I • -40

lOOms

I - ~

• -60

I lOOms

• -80

left -t00

post lesion

Fig. 6. Initial rise of eye velocity. In A five examples of eye velocity profiles are shown for the onset of optokinetic stimulation at 60 °/s towards right and left pre- and postlesion, respectively. In B the mean eye velocity and standard deviation during the first 200 ms of the optokinetic stimulation is displayed for different stimulus velocities pre- and postlesion. The lesion did not affect the early eye movements.

In Fig. 3 an example of this spontaneous nystagmus as well as the mean slow phase eye velocity for different optokinetic stimulation velocities are displayed. These data were obtained immediately following the lesion.

m

A

~ ~,...........~ ._

-d

B

-

IA Ici

.

.

.

.

.

.

.

.

D II II uII Jy-j ........... -f

0

•

.

.

~..~m

.--~ -

'

\

1000

,lOd

nil

2000

3000

4000

5000m

F ,'g

I ^

Fig. 7. Pre-, acute, and postlesion smooth pursuit. Horizontal eye position and velocity are shown in register to target position and velocity for individual trials of pursuit experiments. A and B display prelesion data. C and D display pursuit immediately following the lesion• E and F show the postlesion pursuit from day 49. Note that pursuit towards the lesion (right) shown in E displays a reduced tracking velocity and an increase in saccades. For legend and scale see F.

The frequency of the spontaneous nystagmus was 24 beats per 20 s. During optokinetic stimulation opposite to the spontaneous nystagmus, the frequency was reduced to 14 beats per 20 s. Optokinetic stimulation parallel to the spontaneous nystagmus yielded 39 beats per 20 s. Mean slow phase velocity of the spontaneous nystagmus decreased following the lesion. At day 6 postlesion the mean velocity was smaller than l°/s. The spontaneous nystagmus disappeared when the monkey was put in darkness and when a structured background was displayed. The spontaneous nystagmus never returned. At the time of recovery of gaze stability also a recovery of O K N towards the lesioned side (ipsiversive) was observed. The mean slow phase velocity monotonically increased and reached a stable but reduced level at day 38 postlesion. Fig. 4 displays this sustained reduction in eye velocity for different stimulus velocities. Although the reduction is more pronounced for ipsiversive (rightward) stimulation, the eye speed also was reduced for contraversive (leftward) stimulation. Immediately following the lesion and during the presence of spontaneous nystagmus no O K A N could be elicited. For rightward stimulation the slow phases reversed after the cessation of stimulation indicating the presence of the spontaneous nystagmus (see Fig. 5). For leftward stimulation it was not possible to distinguish between after-nystagmus and spontaneous nystagmus. But, since there was no decay in eye velocity over time, this eye movement was more likely to be spontaneous in nature than after-nystagmus.

26 Parallel to the recovery of ipsiversive nystagnms the velocity storage mechanism recovered. Following day 38 postlesion O K A N could be elicited again. Fig. 5 shows OKN and O K A N following prolonged optokinetic stimulation (about 1 min) immediately after the lesion and at day 49 postlesion. The mean eye velocity for 200 ms following the onset of optokinetic stimulation was calculated in order to address the initial build-up of slow phase eye velocity. There was no lesion-induced reduction in mean eye velocity nor an increase in latency of eye movement onset (see Fig. 6). The latency of eye movement onset was determined to be 64 ms (S.D. = 11 ms, n = 132) prior to the lesion, following the lesion the latency was 55 ms (S.D. = 7 ms, n--304) for stimulus velocity of 60°/s in horizontal directions. In the course of other studies four rhesus monkeys had been examined in our laboratory so far, all showing equal latencies to the onset of background movement.

Smooth pursuit At the lesion day the spontaneous nystagmus persisted also during the presentation of a pursuit target. The monkey performed saccades to keep the fovea directed towards the target and to compensate for the drift of the eyes. Fig. 7 shows representative examples for horizontal pursuit eye movements towards right and left before the lesion, at the day of the lesion (with the spontaneous nystagmus), and day 49 postlesion after the recovery of gaze stabilization. Quantitative data of pursuit parameters are shown in Fig. 8. The mean eye velocity and the frequency of saccades are displayed for pursuit towards the left (contraversive) and the right (ipsiversive). Pursuit was performed across either homogeneous or structured backgrounds. Data of 34 days before and 11 days after the lesion were summarized. The postlesion data were scattered between day 38 and 76 postlesion to ensure that the dynamics of recovery of gaze stabilization did not affect the results. Only data with mean eye velocity larger than 10°/s in at least one condition were considered. Unfortunately the monkey displayed a slight asymmetry in mean eye velocity before the lesion; 11.3 °/s for leftward pursuit and 10.0 °/s rightward pursuit. But following the lesion the difference in mean eye velocity for pursuit towards left and right across the homogeneous background was more pronounced; 11.3 °/s reduced to 10.5°/s for leftward pursuit and 10.0°/s reduced to 8.3 °/s for rightward pursuit. Parallel to this reduction in eye velocity, the number of saccades was increased for ipsiversive; for rightward pursuit from 1.8 saccades per trial to 3.0.

contraversive

(left)

ipsiversive (right)

mean eye velocity 15"/s

lO*/s 5*Is

saccades per trial

• ,

p ~0~35

[] preleeion

• ..

p~O.m

I

~

Fig. 8. Mean pursuit data. Mean data for smooth eye velocity and number of saccades per trial (1.8 s) are displayed for 34 days prelesion and 11 days postlesion after recovery of gaze stability (between day 38 and 76 postlesion). Pursuit was performed across a homogeneous or structured background as indicated by the icons. Asterics indicate significant differences between preo and postlesion data (Wilcoxon Rank Test). Note the lesion induced reduction in mean eye velocity for rightward, ipsiversive pursuit across homogeneous background and the increase in velocity for leftward pursuit across structered background.

As also shown by others 6"~s the presence of a structured background reduced the eye velocity and increased the number of saccades in all conditions. Interestingly, the background-induced reduction in mean eye velocity became smaller following the lesion for contraversive pursuit; prelesionary 11.3 °/s reduced to 5.9°/s and postlesionary 10.5°/s reduced to 8.9°/s. The step-ramp paradigm offered the possibilty to examine the monkey for retinotopic deficits. While the ramp direction was kept constant, the step direction and amplitude was changed. No retinitopic deficits were observed following the pretectal lesion.

DISCUSSION

As has been described earlier, optokinetic deficits were shown after lesioning the pretectum. Additionally, a deficit of the pursuit system caused by this lesion was shown.

27 The observed spontaneous nystagmus is also reported in other studies 14,15,32with lower eye speeds and exclusively in darkness. The time course of recovery from this spontaneous eye drift is quite similar to the recovery of gaze stabilization following an hemilabyrinthectomy 34 although the direction of slow phases is opposite following hemilabyrinthectomy. After hemilabyrinthectomy the slow phases are directed towards the lesion side 34, after a pretectal lesion the slow phases are directed towards the non-lesioned side. The efferent projection of N O T / D T N neurons may explain both the similarity in recovery of gaze stability and the different directions of the spontaneous nystagmus. The major efference of N O T / D T N neurons consists in the projection to the ipsilateral inferior olive providing the climbing fiber input to the contralateral cerebeUum2'12'21'24'29. The cerebellar purkinje cells project to the contralateral nucleus vestibularis ~9 (in cat 1, see also Fig. 9.27 in ref. 5). In addition, other N O T / D T N neurons project to the ipsilateral nucleus prepositus hypoglossus 24. The efference of this nucleus crosses the midline to the contralateral nucleus vestibularis 3. Similar to the hemilabyrinthectomy the cause for the spontaneous nystagmus following the pretectal lesion may be related to an imbalance of neuronal activity between the vestibular nuclei of the two hemispheres 34. In case of an hemilabyrinthectomy the nucleus vestibularis ipsilateral to the lesion loses its input from the VIIlth cranial nerve and thereby its spontaneous activity. In case of a pretectal lesion the nucleus vestibularis contralateral to the lesion loses its input from the NOT/ DTN complex. The recovery of gaze stabilization as well as the recovery of O K N towards the lesion may be related to a readjustment of the balance of neuronal activity between the vestibular nuclei. The unchanged parameters of the slow phase eye velocity build-up may contradict one report in the literature 32. However, Schiff and colleagues determined the mean eye velocity 1-2 s following the onset of stimulation, whereas we calculated the mean eye velocity for 200 ms. Secondly, not all animals in the study by Schiff et al. showed the reduction of the initial rise. A different report states no reduction in initial rise in two of four animals 15. An important comment at this point concerns the substrate of the initial build-up of eye velocity. Although a pursuit deficit has been shown by our results, no reduction of initial rise could be observed. The presented pre- and postlesion latencies and the time course of the initial rise are similar to the properties described by Miles and colleagues for the ocular following response (OFR) 27. Thus the initial rise is pos-

sibly identical to the OFR. From our results we can exclude the N O T / D T N as the neuronal substrate of OFR. Early data has been published concerning the neuronal substrate of the OFR. Neurons within the medial superior temporal area (MST) and the dorsolateral pontine nucleus (DLPN) respond prior to the onset of the OFR. Furthermore, some MST neurons could be antidromically identified by electrical stimulation of the D L P N 16. Recently it has been proposed that in primates NOT/ D T N neurons may provide a meaningful signal to the pursuit system 11. The anatomical basis for this proposal is a multiple efference of N O T / D T N neurons: first to structures in close relationship to the gaze stabilization system, i.e. nucleus prepositus hypoglossus, inferior olive (see above) and the nucleus reticularis tegmenti ponti 22'23. Second there are pretectal efferents to the dorsal lateral pontine nucleus 38 which transmit retinal slip information from the posterior parietal cortex to the cerebellum 8'37. An indirect hint to a connection between the N O T / D T N and the pontine nuclei is the similar response properties of N O T / D T N neurons and neurons in the medial lateral pontine nucleus 17. The deficits in the optokinetic as well as the pursuit system support the idea that the N O T / D T N efference is also used to transmit velocity error information to the pursuit system. A previous study addressing the influence of a pretectal lesion on smooth pursuit eye movements reported unchanged pursuit following damage of the pretectum 15. There might be two reasons for the contradiction to our results: first, Kato and colleagues did not analyse the pursuit parameters as mean eye velocity and number of catch-up saccades. Since the effects reported here are so small, they might have been overlooked by Kato and colleagues' work. Second, the monkey reported here was intensively trained on the pursuit paradigm while Kato's was not. The background-induced reduction in mean pursuit velocity is thought to be due to an inhibitory influence of the contralateral optokinetic system driven by the passive movement of the background. This idea is supported by our data showing a lesion-induced increase in eye velocity during pursuit across structured background away from the lesion. For this direction of pursuit the lesioned, non-functioning N O T / D T N would be responsible, at least partially, for the inhibitory influence of the optokinetic system. So it is obvious that the inhibitory influence was weakened in this direction. These data together with other data presented here suggest that the N O T / D T N provides input for the optokinetic as well as the pursuit system. So the question whether the N O T / D T N is part of the pursuit or opto-

28

kinetic system becomes void, moreover as it was suggested by others 35, our data support the hypothesis of only one common slow system. A final comment is related to the co-existence of the slow system (i.e. pursuit system) and the fast system (i.e. saccade system) in the nomenclature of Steinman et al. 35. As we have shown the pretectal lesion initially introduced a spontaneous nystagmus and affected the optokinetic and pursuit system permanently. The retinal errors introduced by these deficits were compensated by saccades. We showed also that the presence of a structured background reduced the mean smooth eye velocity during pursuit and increased the occurence of saccades (see also 6'11A8). The counter-balancing behaviour of the saccadic system has also been shown by lesions of the frontal eye field 2°. In cases affecting the smooth pursuit system, either by changing the visual parameters or by lesions of brain tissue, the saccadic system replaces the weakened pursuit system. Since the experimental procedure includes only the control of eye position and not eye velocity, the oculomotor system is always able to satisfy the paradigm task.

ACKNOWLEDGEMENTS

We would like to thank M. Helming and H. Korbmacher for their excellent and skilful assistance in many stages of this study. In addition we would like to thank one of the anonymous reviewers who encouraged us to analyse our data related to pursuit across structured background. This work was supported by the Esprit Basic Research Action of the European Community and the Deutsche Forschungsgemeinschaft.

REFERENCES 1 Angaut, P. and Brodal, A., The projection of the "vestibulocerebellum" onto the vestibular nuclei in the cat, Arch. Ital. Biol., 105 (1967) 441-479. 2 Ballas, I. and Hoffmann, K.-P., A correlation between receptive field properties and morphological structures in the pretectum of the cat, J. Comp. Neurol., 238 (1985) 417-428. 3 Belknap, D.B. and McCrea, R.A., Anatomical connections of the prepositus and abducens nuclei in the squirrel monkey, J. Comp. Neurol., 268 (1988) 13-28. 4 Bronstein, I.N. and Semendjajew, K.A., Taschenbuch der Mathematik, Verlag Henry Deutsch, Thun und Frankfurt, 1981. 5 Carpenter, R.H.S., Movement of the Eyes, 2nd edn., Pion Limited, London, 1988. 6 Collewijn, H. and Tamminga, E.P., Human smooth pursuit and saccadic eye movements during voluntary pursuit of different target motions on different backgrounds, J. Physiol., 351 (1984) 217250.

7 Evarts, E.V., Methods ['or recording activity of individual neuron s in moving animals, Methods Med. Reds.. 11 (196(0 241 2 5 0 . 8 Glickstein, M.. Cohen, J.L., Dixon, B., Gibson, \ . , tlollins, Mr LaBossiere, E. and Robsinson, F., Corticopontmc visual projections in macaque monkeys, J. Comp. NeuroL, 190 (1980) 2(/9--229. 9 Hoffmann, K-P., Distler, C., Erickson, R.G. and Mader, W., Physiological and anatomical identification of the nucleus of the optic tract and dorsal terminal nucleus of the accessory optic tract in monkeys, E.~7). Brah7 Res., 69 (1988) 635-644. l0 llg, U.J. and Hofl)nann, K.-P.. Response properties of monkey NOT neurons, Soc. Neurosci. Ahstr., 16 (1990) 904. 11 llg, U.J. and Hoffmann, K.-P., Responses of monkey nucleus of the optic tract neurons during pursuit and fixation, Neurosci. Res. 12, (1991) 101-110. 12 Ito, M., Cerebellar contro[ofthevestibular-ocular refiex--around the flocculus hypothesis, Annu. Rev. Neurosci., 4 (1982) 275-296. 13 Judge, S.R., Richmond, B.J. and Chu, F.C., hnplantation of magnetic scleral search coils for measurement of eye position: an improved method. Vision Res., 20 (1980) 535-548. 14 Kato, I., Harada, K., Hasegawa, T., Igarashi. T., Koike, Y. and Kawasaki, T., Role of the nucleus of the optic tract in monkeys in relation to optokinetic nystagmus. BraO~ Res., 364 (1986) 1222. 15 Kato, I., Harada, K., Hasegawa, T. and lkarashi, T., Role of the nucleus of the optic tract of monkeys in optokinetic nystagmus and optokinetic after-nystagmus, Brain Res., 474 (1988) 1626. 16 Kawano, K., Watanabe, Y., Kaji, S. and Yamane, S., Neuronal activity in the posterior parietal cortex and pontine nucleus of alert monkey during ocular following response. In Vision, Memoo,', and the Temporal Lobe, Elsevier, New York, 1990, pp~ 311315. 17 Keller, E.L. and Crandall, W., Neuronal responses to optokinetic stimuli in pontine nuclei of behaving monkey, J. Neurophysiol., 49 (1983) 169-187. 18 Keller, E.L. and Khan, N.S., Smooth-pursuit initiation in the presence of a textured background in monkey, Vision Res., 26 (1986) 943-955. 19 Langer, T., Fuchs, A.F., Chubb, M.C., Scudder, C.A. and Lisberger, S.G., Floccular efferents in the rhesus macaque as revealed by autoradiography and horseradish peroxidase, J. Comp. NeuroL, 235 (1985) 26-37, 20 MacAvoy, M.G., Gottlieb, J.P. and Bruce, C.J., Smooth pursuit eye movement representation in the primate frontal eye field, Cereb. Cortex, 1 (1991) 95-102. 21 Maekawa, K. and Simpson, J.l., Climbing fiber responses evoked in vestibulocerebellum of rabbit from visual system, J. Neurophysiol., 36 (1973) 649-666. 22 Maekawa, K. and Kimura, M., Electrophysiological study of the nucleus of the optic tract that transfers optic signals to the nucleus reticularis tegmenti ponti--the visual mossy fiber pathway to the cerebellar flocculus, Brain Res., 211 (1981) 456-462. 23 Maekawa, K., Takeda, T. and Kimura, M., Responses of the nucleus of the optic tract neurons projecting to the nucleus reticularis tegmenti pontis upon optokinetic stimulation in the rabbit, Neurosci. Res., 2 (1984) 1-25. 24 Magnin, M., Kennedy, H. and Hoffmann, K.-P., A doublelabeling investigation of the pretectal visuo-vestibular pathways, Vis. Neurosci., 3 (1989) 53-58. 25 Maioli, C., Optokinetic nystagmus: modeling the velocity storage mechanism, J. Neurosci., 8 (1988) 821-832: 26 Mikami, A., Newsome, W.T. and Wurtz, R.H., Motion selectivity in macaque visual cortex. 1. Mechanisms of direction and speed

29

27

28

29

30

31

32

selectivity in extrastriate area MT, J. Neurophysiol., 55 (1986) 1308-1327. Miles, F.A., Kawano, K. and Optican, L.M., Short-latency ocular following responses of monkey. I. Dependence on temporospatial properties of visual input, J. Neurophysiol., 65 (1986) 13211354. Mustari, M.J. and Fuchs, A.F., Discharge patterns of neurons in the pretectal nucleus of the optic tract (NOT) in the behaving primate, J. Neurophysiol., 64 (1990) 77-90. Mustari, M.J., Fuchs, A.F. and Kaneko, C.R.S., Descending connections of the nucleus of the optic tract, Soc. Neurosci. Abstr., 16 (1990) 904. Robinson, D.A. A method of measuring eye movement using a scleral search coil in a magnetic field, IEEE Trans. Biomed. Eng., 10 (1963) 137-145. Schiff, D., Cohen, B. and Raphan, T., Nystagmus induced by stimulation of the nucleus of the optic tract in the monkey, Exp. Brain Res., 70 (1988) 1-14. Schiff, D., Buettner-Ennever, J. and Matsuo, V., Effects of lesions of the nucleus of the optic tract on optokinetic nystagmus and after-nystagmus in the monkey, Exp. Brain Res., 79 (1990) 225239.

33 Simpson, J.l., Giolli, R.A. and Blanks, R.H.I., The pretectal nuclear complex and the accessory system. In J.A. BuettnerEnnever (Ed.), Reviews of Oculomotor Research, Vol. 2, Elsevier, Amsterdam, 1988, pp. 335-364. 34 Smith, P.F. and Curtboys, I.S., Mechanisms of recovery following unilateral labyrinthectomy: a review, Brain Res. Rev., 14 ( 1989) 155-180. 35 Steinman, R.M., Kowler, E. and Collewijn, H., New directions for oculomotor research, Vision Res., 30 (1990) 1845-1864. 36 Szabo, J. and Cowan, W.M., A stereotaxic atlas of the brain' of the cynomolgus monkey (Macaca fascicularis), J. Comp. NeuroL, 222 (1984) 265-300. 37 Thier, P., Koehler, W. and Buettner, O.W., Neuronal activity in the dorsolateral pontine nucleus of the alert monkey modulated by visual stimuli and eye movements, Exp. Brain Res., 70 (1988) 496-512. 38 Weber, J.T. and Harting, J.K. The efferent projections of the pretectal complex: an autoradiographic and horseradish peroxidase analysis, Brain Res., 194 (1980) 1-28. 39 Wurtz, R.H., Visual receptive fields of striate cortex neurons in awake monkeys, J. Neurophysiol., 32 (1969) 727-742.