The effectiveness of the statolith organs in human spatial orientation

THE EFFECTIVENESS OF THE ST.iTOLITH IN HUMAN 3PATIAL ORIUFTATIOH

Summary.

In a dark environment the direction of the visual apparent vertical largely depends on the stimulation of the statolith (otolith) organs.

Apart from the direction of the apparent vertical its steadiness in

time also seems to depend on the influence of these sense organs. The decreasing steadiness, occurring With increasing degree of lateral tilt, points to a decline of their influence. This is substantiated by findings on the effect of non—statolithic stimulation: visual, somatic and semicircular canal stimulation excert a stronger influence on the apparent vertical with the head inverted than with the head upright. The statolith organs therefore seem to become less effective

in indicating the vertical with increasing degree of tilt. The question is, whether or not this is due to malfunctioning of the statolith

organs themselves at positions with the head inverted. On account of electrophysiological findings and measurements on the countertorsion reflex of the eyes, it is concludsd that these sense organs are

functioning equally well with head upright and head inverted. It is

implied that the changes in effectiveness are due to processes on the

brain stem level.

Introduction.

With lateral tilt of the head the visual world does not change its apparent orientation, despite the rotation of the visual image on the retina. This phenomenon does not only depend on the fact that the

visual field consiste of objects, of which the orientation to gravity is known. Also in a dark environment the apparent orientation of a single luminous line changes relatively little with tilt of the head: if a luminous line is adjusted to apparent verticality at various positions of the head, the deviations from gravity will be relatively small. This phenomenon is referred to as vertical constancy.

For a causal investigation of the vertical constancy phenomenon it^ is important to measure the apparent vertical (AV) relative to the subject and not relative to gravity. The adjustment of a lin-- to apparent verticality should be regarded as a response of the subject; it is a response to stimulation of different sensory organs, which deliver information about our position in space. If therefore with increasing degree of tilt the line is adjusted approximately correct, this involves increasing counterrotatiori of the line by the subject and consequently an increasing response (see fig. 1a; op. Feilchenfeld,

1903; Mesker, 1953s Schöne, 1962, 1964; Colenbrander, 1964; Correia et al., 1965). Measured in this way, the ,xV response is plotted in fig. 1b as a function of the lateral tilt position of the body for a number of subjects. For an easier understanding of the figure a dotted

Pig. 1a. Body position as stimulus for the gravireceptors and. apparent vortical (AV) as response to this stimulation; the position of the body is measured as the

angle between the gravitational vortical (GV) and tho median plane of the

body; the apparent vertical is measured as the angle between the median plane

of the body rand tho adjusted lino.

b. AV response of 13 subjects as a function of tho position of tho body. Tho dotted line represents the direction of gravity. (Prom Udo de Haes, in press a).

Fig. 2a. Intra-test variance of AV of 13 subjects as a function of body position

(from Udo de Haes, in press a).

b. Moan deviation of tho AV of 3 subjects as a function of body position, with

the subjects submerged under water (from Schöne, 1964).

10 60 9O 120 BODYPOSITION (degr...)

150 180

2a

2b

30 60 90 120 150

BODY POSITION (degrees)

180

60 90 120 150 BODYPOSITION (degrees)

— 2 —

line has been drawn, indicating gravity; points on this line represent correct adjustment, points over it adjustments beyond the true vertical (E—phenomenon after Müller, 1916), and points under it adjustments on the same side of the true vertical to which the body is tilted (A—phenomenon

after Aubert, 1861).

Différent sensory systems contribute to the perception of the vertical.

Of these the visual system and the gravireceptors are of primary importance. The latter can be divided into two groups: statolith (otolith) organs

in the inner ear and somatic receptors in various parts of the body. The role of the statolith organs has been investigated mainly by the employment of two methods: (a) use of a human centrifuge, by which the magnitude of

the gravitational force can be enlarged as an independent variable at

different positions of the body, i.e. at different directions of the force

(Schöne, 1962, 1964; Colenbrander, 1964; Correia et al., 1965; Miller & Graybiel, 1966b; Schone et al., 1967; cp. Wade & "3ay, 1967); and (b) compa-rison of responses of normal and labyrinthine—defective subjects (Kreidl, 1893; Fischer, 193Ob; Graybiel, 1956; Graybiel & Clark, 1965; Clark &

Graybiel, 1966ab; Graybiel et al., 1968). Prom these investigations it

has been concluded that the statolith organs play an important role in the

perception of the vertical in a dark environment. This also applies to positions with the head inverted. The statolith organs contribute to the perception of being in the upside-down position, as Graybiel & Kellogg

(1966) showed by investigating the "inversion illusion" during parabolic

flight manoevr3s with normal and labyrinthine-defective subjects. The sense of being in the upside-down position during the short states of

weightless-ness was perceived only by normal subjects.

Apart from the direction, also the steadiness in time is a characteristic feature of the -iV. It can be measured as the variance of the adjustments

of the line during a certain period of time. In many investigations it

has been shown that the AV is more steady with the subject upright than when he is tilted (Nagel, 1898; Fischer, 1930a; Witkin & Asch, 1948a; Miller & Graybiel, 1963a; 1966'a; Schone, 1964; Colenbrander, 1964;

Miller et al., 1966; Schone & Udo de Haes, 1968; Udo de Haes, in press a). In fig. 2a AV variance values are plotted as a function of the tilt

position of the body for a number of subjects. The fluctuations of the AV gradually increased for all subjects from 0°to 90 of tilt. Prom thereon the subjects behaved rather differently. Maximal variance values

were mostly obtained at 150 of tilt. At 180 the line was again

adjusted consistently more steady by a number of subjects. In this position the JAV coincides more or less with the median plane of the

body, which may have resulted in the more steady adjustments. As can be

seen in the previous figure (fig. 1b) the variability between the

subjects also reaches a maximum at 150 .

Prom a subjective point of view the uV is accordingly less "certain"

or "convincing" when the subject is in a position with the head

3

-indicate great disorientation under this condition (Quix & Eijsvogel, 1929).

These phenomena can be compared with results of experiments concerning the so called "oculogravic illusion" (Graybiel, 1952), which is an other way of measuring the AV response. In this set up the direction of the force relative to the subject is varied by centrifugation as in a merry-go-round, instead of by tilting the subject. The threshold spoed of rotation, giving rise to a perceivable change of force direction, turned out to be much greater with the head of the subject inverted than with the head upright (Graybiel & Patterson, 1955j Graybiel & Clark,

1962).

It has also been inferred that these changes in steadiness of the AV are related to changes in stimulation of the statolith organs. The increased fluctuations of the AV might indicate a declining effectiveness of these sense organs in indicating the vertical (Quix & Eijsvogel, 1929ï Graybiel & Patterson, 1955; Brown, 1961; 3chone, 1962, 1964; Schone & Udo de Haes,

1968; Udo de Haes, in press a). In accordance with this hypothesis it has been shown that the steadiness of the _iV is mainly dependent on the

position of the head and only to a limited extent on the position of the trunk (Schöne & Udo de Haes, 1968); and further that the same increa.se

in variance occurs with the subject submerged under water (Brown, 1961; Schöne, 1964; see fig. 2b). However, these findings do not disprove the possible involvement of other less specific factors. For instance, an increase in blood pressure in the head might also play a role. Unfor-tunately experiments with labyrinthine-defective subjects are sparse in this context; the data of Miller et al. (1966) perhaps suggest a

constant variance level of the ^V from 10 to 90 of tilt for the labyrinthine-defective subjects.

The hypothesis of a declining effectiveness of the statolith organs has been further investigated by three lines of research: (a) measurement of the effect on the AV of different stimuli, which "interfere" with the statolithic influence; (b) measurement of the influence of the statolith organs on the generation of different eye reflexes; and (c) measurement of the steadiness of the firing rate of single afférents from the statolith organs. In each case the effects were studied as a function of the degree of lateral tilt of the subject.

a. Effect of different typos of non-statolithic stimulation on the apparent vertical.

Many factors, internal as well as external in origin, affect the AV. Here the effect of different stimuli will be discussed in as much as this

provides information about the relative influence of the statolith organs. To this end a constant non-statolithic stimulus has to be applied at

4

-the statolith organs. The difficulty -therefore is to achieve such constant interfering stimulation at all positions of the body for each of the

modalities of the interfering stimula sunder study.

Somesthetic stimulation.

The stimulation of skin- and joint-receptors does affect the jiV. For

labyrinthine defective subjects these receptors even provide the only

information (in the absence of visual cues) about the direction of gravity. Correspondingly the orientation of these subjects is rather impaired under water, where differential stimulation of the skin receptors is largely reduced (James, 1887; Padden, 1959; Graybiel et al., 1968). But also the AV of normal subjects may be slightly affected by the subaquatic condition

(Graybiel et al., 1968). Furthermore the position of the trunk relative to the head influences the AV of normal subjects (Fischer, 1927; Clark & Graybiel, 19660; Wade, 19685 Schöne & Udo de Haes, 1968).

A more precise analysis of the interaction of somatic receptors and

,-Wade and Day,

statolith organs has been performed by Wade and Day (:1968 ; Wade, 1968, in

press; Day ft Wade, 1968, 1969).They come to an attractive simple "additive

hypothesis", implying a pure addition of statolithic, trxuak and neck

effects on the adjustment of a line to apparent vorticality.

This seems to be at variance, however, with the proposed variable effective-ness of the statolith organs. Taking this into account one would have rather predicted an increasing influence on the jiV of the somatic receptors,

with increasing degree of body tilt. Put in mathematical terms similar to those of Wade and .:)ay, this would imply the introduction of weighting

factors, instead of a pure addition of the effects (cp. von Hoist, 1950b| Bischof, 1966, in press). Now the experiments of Wade and Day have been performed over only a limited range of tilt (0 to 30 ); a wider range might indeed have shown that weighting factors are involved, which depend on the position of the body.

Over the whole range, from head-up to head-down, experiments have been performed by Schöne and Udo de Haes (1968). jtt different positions of

o c the head the trunk-to-head position was independently varied to -45» 0 and +45 » giving rise to a more or less constant interfering sonrct-sthetic stimulus situation at the different positions of the head (see fig. 3a). The difference in the A¥ between the two asymmetrical trunk-to-head

positions is plotted in fig. 3b as a function of the position of the head. The effect of a. change of trunk position was increased at positions with the head inverted, which indeed points to a decreased importance of the statolith organs relative to the influence of the somatic receptors. From 0 to 90 no such increase in the "trunk position effect" was ob-served, contrary to the increase of AV variance in this range of tilt. However, the experimental situation used was rather complex. Changing the trunk-to-head position involves a combined change of stimulation of both trunk- and neck-receptors. Furthermore a change of trunk position over a

Pig, 3a. Change of trunk-to-head position at a given position of the head.

b. Trunk position effect (difference in AV between trunk-to-head position +45°and —45 ) as a function of the position of the head. Prom 0 to 90 mean data of 7 subjects from 135 to 18O of 5 subjects. (Plotted after Schöne & Udo de Haos, 1968j only the first adjustment of every test is included here).

Pig. 4« Range of two adjustment types (a and b) at 120-150 of tilt, between which

many subjects have the choice. direction of gravity, .... median plane

of the body. (Schematized aftor the data of Udo de Haes and Schöne, cited in

this review). Note: the "median plane type" (b) becomes less likely in the

course of an 8-min test. This becomes apparent from tests in which jumps

between the two types occurred; of these 20 vs. 1 started with the median

plane type, whereas only 3 vs. 18 ended with it (

Y

2 = 24-5» P <^ .001).

A

This can be related to the proceeding adaptation of the somatic receptors, which in turn indicates their influence on the occurrence of the median plr:ne typo.

3b

3a

45 90 135

HEAD POSITION (degrees)

180 210 w

i

150 III 60-

30-oL-i^

/

/

/

/

30 60 90 120 150 180

BODY POSITION (degrees)

5

-r

relative to the head

certain anglefdoes not need to produce the same change in stimulation of the

somatic receptors at the different positions of the head. The "interfering stimulus" may therefore not have been constant at the different positions of the head.

Additional support for the notion that the relative influence of the somatic

receptors has increased in positions with the head inverted comes from the

o o

following peculiar phenomenon. Prom about 120 to 150 of tilt many subjects fool they have to choose between two distinctly different ways of adjustment (see fig. 4): one (a) involving positions of the target line in line with those at 60 to 90° of tilt; the other (b) involving adjustments close to the median plane of the body (op. Fischer, 1930a; Udo de Haes <*fc Schöne,

in press; Udo de Haes, in press "b). It seems to be the most plausible inference that type (a) principally depends on the stimulation of the statolith organs, because of the resemblance with adjustments at a smaller degree of tilt. It is proposed that the "median plane type" (b) mainly depends on the stimulation of somatic receptors (see legends fig. 4)« If this is so, the occurrence of this adjustment type at 120° and 150 again

would indicate that the statolith organs have lost importance at these

positions in favour of the somatic receptors.

Visual stimulation.

A number of investigations has dealt with the interaction between visual

and gravireceptor stimulation. The initial question concerned was which t»f

these two is decisive in space orientation. To this aim a luminous line,

or rod, was adjusted to apparent verticality in a stimulus situation in

which the direction of gravity ajid of the visual field did not coincide.

As visual field a whole room or some pattern of parallel contoiirs was

used. Erom adjustments made in such a situation tho importance of

gravi-receptor stimulation was inferred by Gibson & Mowrer (1938), Mann et al.

(1^.19), Moble (19^9), Passey (1950) and Kann (1952); the importance of

visual stimulation, on the other hand, by Asch and Nitkin (I948ab),

Wit ici n and Asch (I948ab), Witkin (1950) ?~

nc

* also by Bitterman and

Worchel (1953).

Gibson (1952) stressed the need to investigate more systematically the

interaction between these two types of stimulation. The question then is

on which factors their relative importance depends. Several investigations

deal with the influence of direction and structure of the visual field

(Kleint, 1936; Asch & Witkin, 1948a; Witkin & Asch, 1948b; Graybiel, 1952;

Mesker, 1953; Bischof & Scheerer, in press). Here we will concentrate on

the influence of changes in stimulation of the statolith organs, as

achieved by tilt of the subject.

The first experiment relevant in this context has been performed by

Witkin & Acch (I948a): the subject adjusted a rod, which was placed in

a 30 tilted room, while the subject himself was in the upright or 30

tilted position. In both these conditions the .»iV was "attracted" by the

6

-tilted room. However, with the subject -tilted himself the rod was adjusted more in line with the room than when the subject was upright. This points to a declined influence of the statolith organs at the tilted position of the body, in relation to visual stimulation.

For an investigation over the whole range of body tilt this method of

presenting a visual field can not bu used. It is not the angle between the visual field and gravity, which determines the relationship between the two cues. An over 30 tilted room may for a subject, tilted 120 himself, look perfectly upright vexactly/ because of the .a—phenomenon ; The

visual field should be set in a certain angle relative to the apparent vertical at the different positions of the body.

This has been accomplished in different ways. It would be the most

straight forward to set the visual field in a fixed angle to the

apparent vertical, as perceived in the absence of the visual field (iiV—dark; see fig. 5a); the deflection of the jiV, induced by the visual field,

could then be measured at the different positions of the body. However,

AV-dark is rather changeable, even in the course of a single test.

Mesker (1953) solved this difficulty be presenting the visual field

(a pattern of parallel stripes) in a variety of directions at all body

positions under study (from cPto 90 ). afterwards it could be calculated from the data which direction of the stripes had most nearly coincided with AV-dark, and consequently had given no deflection of the ^V. He

then measured the maximal AV deflections induced by the stripes to either side of AV-dark. The difference between those is plotted in fig. 5^ as a

function of the degree of body tilt. The AV-'appears to be increasingly affected by the visual field with increasing degree of tilt of the body from 0° to 90°.

Another technique of presenting a comparably interfering visual field at

the different positions of the body was used by Bischof & Scheerer (in prep.). They made use of a slowly rotating visual field of lines, in the presence of which a target line was continously adjusted to apparent verticality. Because of the rotation of the field the target line performed an latory movement, symmetrical about AV-dark. The amplitude* of this oscil-lation indicates, how strongly the target line "followed" the lines, and therefore inversely indicates the relative importance of the

stato-o o

lith organs. In all three subjects the amplitude increased from 0 to 60 of body tilt, thus pointing to a declining influence of the statolith organs. Surprisingly the oscillations became smaller again from 60 onwards, although at 180 they were still of a greater magnitude than at 0°.

Still another technique was used by ^do de Haes (in pross b). He sut the visual field, again consisting of parallel lines, in a fixad angla

relative to the apparent vertical as perceived in the presence of the visual field. This was achieved by fixing the visual field at either +15°

Pig. 5a« Apparent vertical (AV) as a compromise between direction of visual fiold and apparent direction of gravity, as percoivod in the dark (AV-dark, showing

A-phenomenon); g: gravity. (From Udo de Haos, in press b).

b. Visual fiold effect (difference between maximal AV deflections to both sides of iV-dark, induced by the visual field) as a function of body position; moan of responses of 3 subjects; (plotted after Mesker, 1953)«

Pig. 6a. Luminous lino with the fiold of lines fixed to it at -15 •

b. Comparison of tho AV with the field of lines fixed at rusp. +15 and -15° relative to the target line (AV+ and .aY-reap.), In thw fi,gure a sym-metric"! effect is assumed about j.V-dark.

c. Visual fiold effect (mean difference between A.V+ and ^V- of 10 subjects) as a function of body position; (from Udo de Haes, in press b).

5b visual field 30 Q LU

i

w

10

O 30 60 90

BODY POSITION (degrees)

6c

AV +

60 90 120 150

7

-until the target line appeared vertical (sac fig. 6a). Consequently the adjustments were always produced under the same apparent tilt of the visual field relative to the AV (+ or — 15 ) at all positions of the body. The size of the difference in adjustment between the two conditions of the

visual field (AV+ and AV- in fig. 6b) gives an indication of the influence

of the visual field on the AV; it shovied a gradual increase from 0 up to 90 of body tilt, comparable to the results of Mesker (see fig. 6c). The visual ifioid effect remained high from 9O onwards, indicating a

decreased relative importance of the statolith organs at body positions with the head inverted, as compared with the influence of visual stimulation. Thus both the variance of the AV (in a dark environment) and the effect on the AV of a visual field (which is superimposed on the variance) indicate a decreased influence of the statolith organs in the range of positions with the head inverted. Hoxvever, the increase in variance in this range was much more pronounced than the increase in the effect of the

visual field in both the experiment of Bischof and Schoerer and of

Udo de iïaes. It is the question whether this holds true for the effect of every visual field, or whether this depends on the fact that in both cases a field of parallel lines instead of a real structured picture has boen presented to the subject. The possibility exists that a field of parallel lines ceases to work as a cue for verticality in the absence of a reference which is offered by gravity stimulation: a pattern of parallel boards

on the ceiling, for instance, does not give a clear impression of ver-ticality. A field of parallel lines may therefore also cease to work as a cue for verticality in positions with the head inverted, in which the gravity-dependent reference frame is supposed to be rather "weak". A real structured picture, on the other hand, may continue working as a cue for vertic-,„lity if the gravity-dependent reference is absent or becomes "weak". It is to be investigated therefore, whether the effect on the AV of a reci] structured picture might show a greater difference between the ranges of nositions with the head upright and the head inverted.

Semicircular canal stimulation.

The interaction between statolith organs and semicirculo,r canals on the AV has only boen investigated to a limited extent. The canals are stimulated by angular accelerations, and may contribute to the perception of the vertical with rolling movements of the head; (cp. the early findings of Wagel, 1898). Here we will confine ourselves to the question as to how a particular angular stimulus to the canal system influences the AV as a function of th_ -^roe of lateral tilt of the body. To achieve this, rotation :hould be performed in a vertical plane about a sagittal axis

of the head; after one minute of rotation with constant speed (e.g. 60 /sec) the subject should be stopped abruptly fvt the required body position.

The canal system is then stimulated by the stop impulse ("postrotatory" stimulation) and signals "rotation", while the statolith organs give

8

-stopping. It may then be investigated as to how strongly the AV is affected by the canal stimulation at the different positions of tho body.

The first relevant experiments were performed by von Hoist and Grisebach

(1951)»using approximately the above described method. In a typical

example they show that the canal influence on the AV is stronger on

stopping at the 180 than at the 0 position (see fig. 7a). This experiment has been extended by Udo de Haes and Schöne (in press), who compared

the AV at stop after clockwise and after counterclockwise rotation for different stop—positions (see fig. ?b). The time course of the difference in the AV between the TWO rotation conditions was taken as a measure of the canal effect; it increased with increasing degree of tilt on stopping, reaching a maximum at 150 (see fig. 7c). This data, confirming the

earlier findings of von Hoist and Grisebach, indicate a declining

in-flxience of the statolith organs with increasing dogroe of tilt, in relation to tho influence from the semicircular canals.

b. Influence of the statolith organs on eye reflexes.

It seems an interesting question as to whether the decrease in effectiveness of the statolith organs with increasing tilt, as .described above, is limited to the perception of the vortical. In this respect investigations,

dealing with the influence of the statolith organs on the generation of certain eye reflexes, are relevant.

Suppression of postrotatory nystagmus.

The vestibular nystagmus -±o- primarily elicited by stimulation of the semicircular canals. However, when the usual horizontal orientation of the rotation plane ic changed to vertical the statolith organs also seem to influence this reflex. After an idential stop-impulse relative to the subject, the postrotatory nystagmus has a shorter duration if the rotation is performed in a vertical plane, than if it is performed in a horizontal plane (Correia & Guedry, 1964; Guedry, 1965; Benson <\ Bodin, 1966a).

Guedry (1965) explained this fact by assuming that a,fter stopping rotation in a vertical plane the "fixed position signal" from the statolith organs is in conflict with the "rotation signal" from the canals (see above);

the statolith organs will consequently suppress the nystagmus response. In a horizontal plane the statolith organs cannot differentiate between positions, and therefore no suppression of the nystagmus may take place. Although this hypothesis has not gone unchallenged (cp. Benson cc Bodin,

1966ab), it seems to be accepted now. However, critical surgical experi-ments, which may show an abolition of the suppression .-ifter selective extirpation of the statolith organs (cp. Janeke, 1968) have still to be performed.

The question is how postrotatory nystagmus, generated after vertical plane

rotation about the sagittal axis of the head (cp. previous chapter) depends

Fig. 72.« Typical example of AV after stopping at 0 (full line) and 180 ("broken line)

as a function of time; the AV is measured relative to gravity (from von Hoist

& Grisebach, 1951).

b. Apparent vertical, influenced by a stop-impulse after clockwise and after counterclockwise rotation in tho frontal plane.

c. Semicircular cajial effect on the AV (difference in timocourse of AV between the two rotation conditions) as a function of thu "body position on stopping. Mean of responses of 7 subjects (from Udo de Haes à Schöne, in press).

Fig. 8a. Example of two electro-nystagmograms of a rabbit, obtained after stopping of

rotation in the frontal plane in the 0 and in the 180 position. The

movements of the (laterally placed!) eyes are recorded with vertical leads;

speed of rotation: 288 /sec. (from Janeke, 1968).

b. Fumber of nystagmus beats (of upper eye) after 60 > sec stop-impulse, as a

7a

'«T 10 0) « 5 S» o •D •" D ^ -10 -15 / / P" ^ V N, ^"N,,, X '*^. X — -^ — — '• •• — ^ — • — * *—_'""r=nn»« i 30 60 90 120 150

TIME (sec)

7b

CCW —CW-STOP

50-7c

90 120 30 60

STOPPOSITION (OC)(degrees)

150 180

8a

| stopping moment

180e

J500//V — isec

8b

20

e-(O

D

!

U.

ia

m

30 60 90 120 150 180

STOPPOSITION (degrees)

_ 9 —

on the stop-position. This has been investigated with rabbits and with humans. In fig. 8a electro—nystagmograms of a rabbit are presented, which ara obtained after stopping of rotation at 0 and at 180 (from Janeke, 1968). In fig. 8b the number of nystagmus beats of human -subjects is plotted as

a function of stop-position (from Udo de Ilaos & Schöne, in press). In both

investigations an increase in the number of beats was observed with

in-creasing degree of tilt on stopping. This points to a dein-creasing suppression

of tho nystagmus by the statolith organs.

indicate a declining effectiveness Thus the data obtained on both AV and suppression of nystagmus^of the

statolith organs with increasing degree of lateral tilt, in relation to the influence of the semicircular canals.

/mother point is of interest. The change in AV and suppression of nystagmus, depending on lateral tilt,can be compared with a difference in related

responses between the prone and supine position of the subject. One should know that, with respect to the position of tho statolith organs towards gravity, the supine position involves a greater degree of tilt than the prone position (see fig. 9)« Now on the one hand space orientation as measured by verbal reports (Quioc & Eijsvogel, 1929) and by arm pointing (Brown, 1861), is more impaired in the supine than in the prone position. On thu other hand the suppression of postrotatory nystagmus (after vertical plane rotation in tho subjects yaw—plane) seems to be weaker in the supine than in tho prone position (Correia & Cue dry, 1964; Guedry, 1965; Benson <V. Bodin, 1966a). This supports the idea that the same processes underlie both the increasing disorientation and the decreasing suppression of nystagmus.

Ocular countertorsion.

The countertorsion (counterrolling) of the eyes, occurring with lateral tilt of the head, reduces the rotation of the visual image on the retina,

and it is to be considered therefore as a contribution to vertical con-stancy (cp. Colenbrander, 1964; Bischof, 1966; Udo de Haes, in press a). In contrast to many animal species in which this reflex is well devolopped, it is vestigial in man. Our present interest is accordingly not concerned with the quantitative contribution to the perception of the vertical, which may be neglected in first approximation. Investigations with labyrinthine defective subjects (Benjamins & Nienhuis, 1927; Fischer, 1930b; Miller & Graybiel, 1963b; Kellogg, 1965) and centrifuge experiments (Woellner à Graybiel, 1959? Schone, 1962; Colenbrander, 1964) have shown that the countertorsion reflex primarily depends on the stimulation of the stato-lith organs (see fig. 10a). This reflex can therefore be used as an in-dicator of the functioning of these sense organs. In this respect it seems worthwhile to compare the steadiness of countertorsion and AV as a function of the degree of lateral tilt.

The course of the reflex is plotted in fig. 10a; it reaches a maximum at about 60 to 75 of tilt. Now some authors have presented rather irregular curves in the range of positions with the body inverted, suggesting

Pig. 9- Orientation of the utricular sensory epithelia (maculae), seen in trans-section, in the upright, prone and supine position of the head; the supine position involves a greater degree of tilt of tho maculae than the prone position (based on the anatomical data of Corvera et al., 1958).

Pig. 10. Ocular countertorsion reflex as a function of body position, for two levels of magnitude of the gravitational (resp. gravito-inertial) force (ig and 2g).

The reflex is measured relative to the head (from Schöne, 1962).

Pig. 11. Intra-test variance of ocular countertorsion as a function of body position;

values of 2 subjects ('from Udo de Haes, in press a).

10

30 60 90 120 150

BODY POSITION (degrees)

180

11

Z_O0 . 8

55

K

e

_K LU | 0.6 U tt < J U O LU O 0.4 W LU a

l

9 ! O . O ' * l o o 0.0 30 60 90 120 BODYPOSITION (degrees) 150 180

10

-Kellogg, 1965). A mors systematic determination of the variance of the countertorsion reflex by Udo de Haes (in press a) did not confirm this. In his experimental set up, (with statical 8—min. tests in each position), no systematic change in the variance was observed from 30°to 180 of tilt

(see fig. 1l). This was in contrast to the sharp increase of the variance

of the simultaneously recorded AV.

c. Electrophysiological recordings.

The statolith organs consist of two membraneous sacs in the inner ear, the utricle and the saccule. The sensory epithelia (the maculae) of the utricles are located horizontally, when the head is pitched about 30

forward, the statoconial membranes lying on top of the maculae. The

maculae of the saccules are vertically oriented, approximately pa,rallel to the median plane of the head, with the statoconial membranes pointing outward; (Quix <% Tilerndley, 1924; de Burl et, 1930; Corvera et al., 1958).

It has boen shown that only displacements of the membranes in the plane

of the macula, brought about by the shear component of the acting force,

is effective in stimulating the receptors (von Hoist, 1950a; Trincker,

1959)- Neither traction nor pressure in a direction perpendicular to the macula has an effect (cp. Trincker, 1962).

Nith the shear component of the force zero, as is the case with the utricular maculae in the head-up position, most afferent fibres uho-,f a steady resting discharge, which changes with tilt of the head (Lowenstein & Roberts, 1950? ffiesen & Klinke, 1969)« However, for a single afferent fibre tilt in different directions does not produce the same change in the

of the utricular afférents,

resting discharge! This directional sensitivity could be related to the microstructure of the receptor cells (Flock, 1964; Lowenstein et al.,

19645 Gpvendlin, 1965)' livery receptor cell has a plane in which it is sensitive and a plane, perpendicular to the latter, in which it is not. In the sensitive plane the resting discharge increases with tilt to one side, reaching a maximum at about 90 and dropping back to the resting level at 180 5 with tilt in the opposite direction the activity is in-hibited. The fibres of the saccular maculae exhibit the resting discharge with the head tilted 90 to either side.

T.Te will now concentrate on the activity of units of the utricular maculae,

which arc known to be the most important part of the statolith system (Versteegh, 192?j Jongkees, 1950; von Holst, 1950a; Schoen, 1950; Janeke, 1968). One of the most surprising features of the afferent discharge activity is it:, a >.inesa in time. The res~bi-j.g discharge rate is highly constant, and there is hardly any adaptation of an increased firing rate during tilt (Lowenstein & Roberts, 1950). Of main interest here is the steadiness of the firing rate of utricular afférents, in dependence on the position of the preparation towards gravity. In this respect the recordings of Trincker (in prep.) of single fibre?, of the utricular nerve of the Cavia are the most detailed and encompass the greatest range of

-

11

-tilt positions, An estimation of the range of deviations from the mean inter—impulse time is gdvmi in Table 1. In the upright position the deviations amounted to 5/° only (cp. the data of Cîroen et sj.. , 1952). With an increase of the discharge frequency it beca.me slightly less

steady, with minimal steadiness in the region of 90 . Of special interest

here is that the firing rate became consistently more steady again with

the preparation inverted.

Discussion.

With regard to the biological significance of the processes of space orientation, their impairement in position: with the head inverted is not surprising. A proper functioning of the space orientation system in those positions, which are only taken up seldomly in daily life, clearly is not strongly selected for. Obviously this offers no answer to the question as to which physiological processos are involved.

The fact that the influence on the AV of the interfering stimuli in-creased with increasing degree of tilt points to a correspondingly de-creasing influence of the statolith organs. Thus the findings indicate that the statolith organs become less effective with respect to the AV response. In case other factors were involved, such as an increase in blood pressure in the head, a relative increase in the influence of the interfering stimuli vculd not ha,ve been likely to occur. Moreover,

Udo de Haes and. Schöne worked with an apparatus in which trunk and limbs were placed horizontally, in order to avoid discomfort for the subject

in positions with the head inverted (cp. Udo de Haes & Schöno, in press). Attention may be paid now to the processes possibly underlying the decline in effectiveness of the statolith organs. The original explanation of Qnix was that the statolith organs themselves were out of function in positions with the head inverted (the "blind spot" of the statolith organs; Quix •% ifcrndley, 1924; Quix & Eijsvo^el, 1929). Ho based this idea on the assumption that the pressure of the statoconial membranes would be the adequate stimulus for the macula.r receptors. Due to their orientation in the skull, the maculae of utricles and saccules would not be stimulated with the head inverted, as mentioned above this assumption about the adequate stimulus has been disproved by later investigations, which leads us to drop the "blind spot" idea.

In a modified form this hypothesis has again been put forward by Young (1967) and "teor (1969)« These authors assume that, although the shear force may be the adequate stimulus, a "pressure bias" is necessary for a proper functioning of the receptor cells. There are several findings which seem to be at variance with this idea. First Bos et al. (1963) found that the eye movements of a rabbit on a parallel swing were not smaller with the head inverted than with the head upright; in fact they were even larger, which they explained by assuming greater movements of the freely hanging utricular statoliths than when pressing on the maculae •

-

12

-Second the fluctuations in the ocular countertorsion reflex aeem to be of the same magnitude in body positions with the "pressure bias" present and with the "pressure bias" absent (fig. 1O). And third the steadiness of the firing rate of single utricular afférents seems hardly to be

influenced by the pressure component of the statoconial membrane on the

macula (Table 1). These findings lead ono to assume that the macular receptors aro functioning equally well with the head upright and the head inverted. For an explanation of the declining effectiveness attention should therefore be shifted to the central processing of the information from the statolith organs.

It seems plausible to formulate a hypothesis with respect to the output side: the greater the AV response (the angle between AV and head), the greater its variance and the greater its susceptability to interfering stimuli. Apart from the divergent small variance values at 180 this hypothesis does not seem to be easily reconciled with the following two points.

If the subjects of Udo de Haes and Schöne are ranked with respect to the

magnitude and to the variance of their AV responses, there appears no

inter-subject correlation between these two quantities (r (13) = 0.04 for 30-90° of tilt; r,(l3) = 0.08 for 120-150° of tilt; Spearman rank

cor-3

relation coefficient). Secondly the hypothesis mentioned suggests that

the changes in effectiveness are confined to the perception of the vertical.

The data on the diminishing suppression of p»strotatory nystagmus (figs. Sa

and b) suggests that this is not the case.

The data obtained on the suppression of nystagmus indicate that the

changes in effectiveness arc the outcome of processes on the brain stem

level, where tho signals from the different parts of the labyrinth are

integrated (cp. Duensing & Schaefer, 1959)« Several processes may be

involved. Prom 0 to 90 there is an even increase in AV variance for all

subjects, and a similar even increase in the offeet of visual and canal

stimulation. This may be related to the increasing deviations of utricular

activity from its resting level. Although the afferent activity may in

itself be well organized, the divergent deviations from the resting level to

either side may', apart from causing a greater response, render it also more

liable to interfering influences (cp. Bischof, in press). The latter are to

be sought internally with respect to the variance of the AV (cp. Udo de

lïaes, in press a).

'From 90 "to 180 , with the utricular activity going back to the resting

level again, this explanation falls short. The great differences between

AV variance of different subjects (fig. 2a) and between the outcome of

experiments with different interfering stimuli, suggest that more complex

processes are involved in this range of tilt.

A hypothesis can be formulated, based on observations of Szentàgothai (1952)

on the concsistency of the statoconial membranes. He found (in the dog)

-

13

-that the membranes are rather viscous and take different shapes with different positions of the head. This has been confirmed recently in the Cavia by Lindeman (1969)« Szentàgothai now proposed that these deformations

might influence the activity of the single receptor cells. A bubble shaped

utricular membrane at 180 might therefore, in comparison with the 0

position, produce an aberrant stimulation pattern of the afferent fibres, although the single receptor cells might function equally well in both positions. The aberrant pattern then could induce disorientation in positions with the head inverted.

However, it also seems possible that information from the saccules is involved. It has been shown by direct mechanical stimulation of the sac-cular maculae that they influence human space perception (Meyer zum

Gottesberge & Plaster, 1965); the patients reported a feeling of falling, or being in the inverted position. If we assume that the utricular acti-vity is about the same in the 0 and 180 position, a discrimination between up and down will not be possible without information from the

saccules. In fact, the microstructure of the saccular maculae, as revealed

by Spoendlin (1965)» seems to be especially adapted for a discrimination

between up and down. It is suggested that the interaction between

utricular and saccular activity, which thus can be supposed to produce an increase in the AV response at body positions c

that results in less steady orientation processes.

14

-Body position 0 - 3 0 3 0 - 6 0 60 - 90 9 0 - 1 8 0

(degrees)

Mean deviation 5 5 - 1 0 ca. 10 decreasing to (percent) about 5

Table 1. Mean deviation of inter-impulse time of single utricular

afférents at different body positions, in percentage of the

mean value at each position (after Trincker, in

prep.)-I n t e r a c t i o n b e t w e e n S t a t o l i t h O r g a n s

a n d S e m i c i r c u l a r c a n a l s o n A p p a r e n t

V e r t i c a l a n d N y s t a g m u s .

(Investigations on the Effectiveness of the Statolith Organs)

Helias A. Udo de Haes and Hermann Schöne

Max-Planck-Institut für Verhaltensphysiologie, Seewiesen, Germany

Supported by the Dutch Government; Wet Wetenschappelijk Onderwijs,

Art. 80 sub. 2.

2

-ABSTRACT

The effect of postrotatory vertical canal stimulation on nystagmus

and apparent vertical was investigated as a function of the

tilt-position at stop in a vertical roll—plane. Nystagmus duration and

number of beats as well as the effect on the apparent vertical

increased with change of stopposition from head—up to head—down«

The results are interpreted in terms of a hypothesis which proposes

a decline of effectiveness of the statolith organs with increasing

3

-INTRODUCTION

Responses euch as nystagmus, due to the stimulation of the

semi-circular canals, can be modified by a concomitant linear acceleration.

In man constant rotation in yaw about a horizontal axis (i.e. in a vertical rotation plane) caused a unidirectional fluctuating nystagmus which lasted as long as rotation continued« After stopping the nys-tagmus as well as the subjective phenomena were much shorter than those after yaw-rotation in the horizontal plane (Guedry, 1965»

Benson and Bodin, 1966a). Because these findings could not be explained by a direct effect of gravity on the cupula (Correia and Guedry, 19^4 5

Guedry, 1965» Benson and Bodin, 1966a) the following hypothesis was proposed by Guedry: during rotation in a vertical plane the responses to canal stimulation are enhanced by the statolith organs which are stimulated by the continuously re—orienting linear force. After

stopping, however, there is a contradiction between the information

from canal- and gravireceptors. The canal-receptors signal rotation,

that is change of position, whereas the statolith organs signal a

fixed position. This discrepancy causes the suppression of the

nys-tagmus. Benson and Bodin (1966 a) criticized this point of view.

They proposed, on the basis of different specific densities of

endolymph and perilymph a direct effect of gravity on the canals:

that component of gravity, which is co—planar to the stimulated

canals, would be effective in cupula restoration.

As far as the perrotatory responses are concerned Janeke (1968)

investigated partial labyrinth-ectomized rabbits; the canal system

was left intact, the saccular maculae and the utricular nerves were

bilaterally destroyed. During vertical plane rotation about a

sagittal (roll-) axis he observed only a transitory nystagmus, whic.

indicated the intactness of the canal system. The sustained

unidirect-ional nystagmus, however, had ceased completely. This indicates that

the continuous perrotatory nystagmus originates in the statolith organs

as was suggested by Guedry (1965).

Concerning the p o st ro t at o ry responses a critical behavioural

experiment with humans was performed by Benson (1966), the result

of which also favoured the hypothesis of Guedry.

This hypothesis implies that the degree of contradiction between th>~

signals from canal- and gravireceptors increases as the

(subject-fixed) plane of rotation is varied from horizontal to vertical.

No contradiction is caused after stop in a horizontal plane, because

the gravireceptors cannot differentiate between positions in this

plane. In a vertical rotation plane, however, gravireceptors

dif-

-4-ferentiate optimally between positions, therefore the degree of contradiction has a maximum value.

Another aspect of the interaction between the statolith organs and canal receptors concerns the effectiveness of the statolith organs. Whereas Guedry*s hypothesis postulates maximal influence of the statolith organs after stop in a vertical plane as compared to the horizontal plane, it bears no implication about the influence of the degree of tilt at stop in a vertical plane. In man this effect of stopposition on postrotatory nystagmus has been invest-igated only in the yaw-plane (Correia and Guedry, 19^4; Guedry, 1965; Benson and Bodin, 1966 a). The nystagmus seemed to be sixthly shorter in the prone (nose-down) as compared to the supine (nose-un) position, suggesting that the cjtatolith organs are more effective-in suppreoseffective-ing the nystagmus effective-in the prone than effective-in the supeffective-ine

position. To obtain further information about the effectiveness o:'

the statolith organs we studied postrotatory nystagmus in the

roll-plane, varying the stopposition over the whole range from

head—up t o head—down.

Thus far we have only discussed nystagmus, which is generally classified as a response to canal stimulation. Our investigations also concerned the apparent vertical (AV), for the perception of which the statolith organs play the principle rolej the relei'ant

stimulus parameters are direction and magnitude of the gravitational force (Schöne, 1962? Correia o.a., 1965; Killer and Graybiel, 1966b); Schöne o.a., 196?)»

As to the effectiveness of the statolith organs with respect to the AV Quix and Eijsvogel (1929) were the first to notice the high degree of uncertainty of subjective orientation at the inverted positions. This and similar phenomena have been confirmed by several authors, who measured the AV-variability in different ways (Fischer, 1930a; Brown, 1961; Schöne, 1964; Schöne and Udo de Haes, 1968; Udo de Haes, in press), including the threshold of the so called

"oculogravic illusion" (Graybiel and Patterson, 1955* Graybiel and Clark, 1962). In every instance the AV was much less stable in the inverted as compared to the normal positions.

The hypothesis proposing: a decline of effectiveness of tho statolith organs with increasing degree of tilt was further elabourâtcd by Schöne (1962). He related the increasing variability of the AV to the findings of von Hoist and Grisebach (1951)» In preliminary

experiments they found that the AV is affected by postrotatory stimulation of the vertical canals in a vertical roll-plane; the

5

-influence appeared to be much greater in the head—down than in tho head-up position. We repeated and extended these experiments.

Also in line with the idea of a changing effectiveness of the

statolith organs is the finding of Schöne and Udo de Haes (1968) that the somatoreceptors of the trunk affect the AV most strongly at the inverted positions of the head.

A theoretical analysis of this kind of interaction between sensor,

systems has recently been given by Bischof (1966).

In summary our investigations have three main aspects:

1) the interaction between the signals from canal and statolith

receptors on a response generally occurring with canal stimula- ion (nystagmus);

2) the same interaction on a response which is mainly dependent c:

the stimulation of the statolith organs (apparent vertical);

3) the decline of effectiveness of the statolith organs with

increa-sing degree of tilt with respect to the AV and to the sup-pression of nystagmus.

METHOD

General, The subject (S) was placed in a bed which enabled

rotation in the roll-plane of the head (fig« 1). After acceleration

of 1°/sec S was rotated with constant speed of 60°/sec for one

minute; his «yes wore closed. He was then stopped within 1 sec at

one bf the following positions: 0, (head-up), 30, 60, 90» 120,

150 and 180 (head-down). In the nystagmus series they were to the

left, with an error of maximal 6°; in the AV experiments they were

to the right, with an accepted error of maximal 3 • In both series

four independent tests, 2 clockwise (CW) and 2 counterclockwise

(CCW), were performed with every S.

Nystagmus. Four Ss were investigated. They wore Frenzel-spectaclos.

The fast rotatory nystagmus beats of the right eye were indicated by

the experimenter and recorded on a tape« Non-rotatory movements of

the eye were avoided by fixation of a tiny light o The testorder was

randomized, with two tests daily for each S. The tape records were

analysed for number of beats and nystagmus duration» Also the

time-pattern of the beats was investigated: in order to get an estimate

of both the timeconstant and the initial value of the slow phase

velocity exponential regression analyses were performed on the

reciprocal values of the intertime between successive beats. The

initial value is defined as the value at the time 2 sec after onset

of deceleration (t ).

Pig. 1. Apparatus and definition of angles,

Pig. 3- Apparent vertical (angle fi ; see figure 1 ) as a function of time, averaged for the results of all subjects.

a) f^ as a function of time, separately for the responses after CW- and after COW-rotation; (cp. inset figure).

b) The difference between corresponding tW- and COW-curves ( f i , . - , , ) as a function of time. Mean initial value of ft at ck = 0+30* differs

1 diff

significantly from the mean initial value of ft,., at 1A= 150+180 ; (p <^.05 for 5 of the 7 subjects; p /.001 for all subjects).

AVGV

3a 180

150

120 ut

90

60 30

—CW-STOP

2 4 6 8

TIME (minutes)

2 4 6

TIME (minutes)

6

-Apparent vertical« Seven Ss, including three from the previous series, were used. After stop they kept their eyes closed for 5 sec. Then a luminous line was adjusted to apparent verticality at 15 sec

inter-vals for 8 min, with the first adjustment 15 sec after stop (t ).

Between the adjustments the line was moved to and fro. The starting position of the line and the testorder were randomized, with two or three tests daily for each S. Prom the timecourse of the difference between the adjustments after CW and after CCW rotation (average values of all tests) the effect from the canals was estimated by means of exponential regression analyses.

RESULTS

Nystagmus. All Ss showed a consistent increase of the number of beats and of nystagmus duration with change of stopposition from O

to 180° (fig. 2a and "b).

The analysis of the inter—teat-times suggested that "both the

time-constant and the initial velocity of slow phase changes with stopposition: the greater the degree of tilt at stop, the greater the initial velocity and the slower its decline in time» The average

time constant of all experiments amounted to about 8 sec. The results of this time-analysis, however, should be considered with

some reserve, because of the limitations of the recording technique.

Apparent vertical« In fig« 3a the AV is plotted against time separately for the different stoppositions and for CW and CCW

rotation. As fig« 3b demonstrates, the difference between the CW

and CCW values has clearly increased at the inverted positions

(150, 180 ). A more refined picture is acquired from the exponent!\1

regression analyses (see formula in fig. 4); only the values up t<~ 90 sec after stop were used. In order to give a passable fit of the calculated curves an additional constant (d) had to be added for

each position; it represents the CW-CCW-difference remaining 1-2 :in

„after stop (cp. inset fig« 4)• As the main diagram of fig« 4 indicates, the initial value of the canal effect (c) increased

clearly with increasing degree of tilt at stop; it reached a maximum at 150°.

Although for an accurate calculation of the timeconstant the time

intervals of 15 sec arc too long, nevertheless useful results could be obtained« The timeconstant values of the inverted positions were

greater than those obtained in the normal range of tilt: for the

positions 0-120 the average tirneconstant was 10 sec, for 150 and

Pig. 2. Nystagmus responses as a function of stopposition; the four subjects

are presented separately,

a) Number of fast nystagmus beats. The mean of the values at c< = O-t-30

differs significantly from the mean at cA = 150+18>0 ; (p <^.01 for every subject; Mann-Whitney U-test after Siegel, 1956).

b) Nystagmus, duration. The mean of the values at &( = 0+30 differs

significantly from the mean at <X = 150+18Q ; ( p < f . 0 5 for every subject; Mann-Whitney U-test)«

Fig. 4» c, the effect of the canals on the AV at t (i.e. 15 sec after stopping), as a function of stopposition. c is calculated by means of exponential regression analyses, using the formula in the figure, of the

2a 2 0

-g .

UI

ca

w

U.

O

K 5 LU

m

30 60 90 120 150 180 15

2b

u O) M

l»

oc

O

(O

2

s

30 60 90 120 150 180 60- 50-•O CC 20 C 30 60 90 120 150

STOPPOSITION (cx)(degrees)

180

7

-A special note should be made about the 15O position. Here for some Ss two distinctly different ways of adjusting were possible; one following the same trend as that of a smaller degree of tilt, the

other involving adjustments in approximately the symmetry plane of

the body (cp. also Fischer, 1930a). The choice between these two was influenced by the direction of the stop impulse? stopping after CW rotation resulted more in the one type, stopping after CCW rotation more in the other, thus producing the high remaining difference (d) at this position (cp. fig. 3b)»

DISCUSSION

Nystagmus » The results of the nystagmus experiments resemble those on rabbits of Janeke (1968), who also found an increase in the number of beats and of nystagmus duration from 0 to the 180 position.

Apparent vertical« The canal effect on the AV increased with in-creasing degree of tilt at stop. This confirms the finding of von Hoist and Grisebach (1951) of a difference in the effect in the 0° and 180 position. Our results further indicate that the effect is

strongest at the 150 position; the two adjustment types of the AV

may have contributed to the large effect in this position,

General, The timcconstant of nystagmus and AV showed about the same

range; the values are in agreement with those measured by Melvill

Jones o.a. (19^4) and Benson and Bodin (l966b). The duration of

both responses differed much more: the canal-induced change of the

AV lasted up to 45 sec (cp, fig. 3b), whereas nystagmus lasted only

5 - 1 5 sec (cp.fig. 3b). The canal-induced change of the AV therefore

seems to have a lower threshold than nystagmus. An analogue is fou^.d

in turning chair experiments with stop after yaw rotation in a

horizontal plane: the nystagmus lasts shorter and has a higher

threshold than the apparent movement of a point of light (the

"oculogyral illusion"; for review see Howard and Templeton, 1966).

The data indicate that with increasing degree of tilt 1) the

sup-pression of nystagmus decreases and 2) the canal-induced change

of the AV increases. The results fit the hypothesis of the declining

effectiveness of the statolith organs (cp, introduction).

As to the origin of the declining effectiveness we may consider

th-idea of the "blind spot" of the statolith organs (Quix and Werndl

1924; Quix and Eijsvogel, 1929). These authors assumed that the

pressure of the statoconial membranes is the effective stimulus for

the receptors. The "blind spot" concerns the stimulus situation ir

those inverted positions, in which no macula undergoes pressure atv

therefore the statolith organs would be out of function.

However, as it is now generally accepted, (cp, Trinckor, 1962), that the macular receptors are not stimulated by pressure tut by

shear force we must drop the "blind spot" idea«

Beyond that any peripheral explanation has to be doubted. Recent measurements of Trincker (pers. connu.) indicatet that the discharge

rate of single utricular afférents is almost as steady in the 18O as in the 0 position.

Consequently the effectiveness of the statolith organs is not

likely to depend on the steadiness of the peripheral affcrence.

Further research is therefore needed to ascertain which physiological

proooooes underly the decline in the effectiveness of the statolith

organs.

It might be worth noting that the "break down" of the effectiveness

occurs in positions taken up only occasionally by humans, in vihich

therefore accurate functioning of the space orientation system is

not required.

ACKNOWLEDGEÎÎENTS

We wish to thank Dr Walter Heiligenberg for helpful discussions of

the computer analysis, Dr David Hughes for his help in the

translat-ion and Miss Brigite Bey for technical assistance.

STABILITY OP APPARENT VERTICAL AND OCULAR COUNT^HTORSION

AS A JUNCTION OF LATERAL TILT

HLLIAS A. UDO DE HAES

Max—Planck—Institut für Verhaltensphysiologie Seewiesen—Germany

Abstract.

In experiment 1 the inter- and intra-individual variance of the

apparent vertical was established as a function of lateral bodytilt.

Both quantities increased with increasing degree of tilt, with a maximum at the inverted positions (15O-18O°). Because the apparent vertical depends mainly on the stimulation of the statolith organs, the findings imply that these organs become less effective in indicating the vertical. In experiment 2 the ocular countertorsion reflex was

photographically recorded, simultaneously with the measurement of the

apparent vertical. This eye reflex, which also depends on the

stimul-ation of the statolith organs, showed no change in variance level at bodypositions from 30 to 180 . The result is discussed with respect to the origin of the change in effectiveness of the statolith organs.

Introduction«

The position of a luminous line, that is perceived as vertical, is referred to as the "apparent vertical" (AV); we will deal here with

the AV in the absence of a visual framework. Two aspects can be measured: its direction and its stability (variance). For a physiological approach

it is appropriate to measure the direction, not relative to gravity, but relative to the subject (see fig. 1). The angle between the AV and the subject is a function of lateral bodytilt. It has been shown that this

depends mainly on the stimulation of the statolith ( otolith ) organs (Schöne, 1962, 1964; Correia et al., 1965; Graybiel & Clark, 1965; Miller & Graybiel, 1966b; Schone et al., 1967; Graybiel et al., 1968).

Also the stability of the AV is a function of the degree of tilt. It is highest in the normal upright position, and declines with increasing

degree of tilt (Nagel, 1898; Quix & Eijsvogel, 1929; Brown, 1961; Miller & Graybiel, 1963a; Schöne, 1964; Colenbrander, 1964; Schone &

Udo de Haes, 1968). It has been inferred that this decrease of stability

might indicate a decline of the effectiveness of the statolith organs

(Quix & Eijsvogel, 1929; Graybiel & Patterson, 1955; Brown, 1961; Graybiel & Clark, 1962; Schöne, 1962, 1964).

This hypothesis has been supported by investigations on the influence

of non-statolithic stimulation: somesthetic (Schöne & Udo de Haes, 196?),

semicircular canal (von Hoist & Grisebach, 1951; Udo de Haes & Schöne,

in press) and visual (Bischof & Scheerer, in press; Udo de Haes, in press,

stimulation all excerted a stronger influence on the AV in the range of

inverted, as compared to the range of normal bodypositions.

2

-The aim of the present experiments was to investigate the origin of the 'e changes in effectiveness. Therefore more precise measurements were per-formed of the stability of two responses, both dependent on the statolith organs, but involving different levels of neural organisation: the AV -rid the ocular countert ors ion reflex; the-'-former involving complex perception processes, the latter being a more direct reflection of statolith organ activity (Woellner & Graybiel, 1959? Schöne, 1962j Miller & Graybiel,

1963b; Colenbrander, 19645 Kellogg, 1965).

EXPERIMENT 1

METHOD.

Apparatus. The equipment for controlling lateral tilt (see fig. 3a) was the same as that described earlier (Udo de Haes & Schone, in press;,,

Subjects. Thirteen Ss were used (6 men and 7 women); all but one, aged 47 were between 16 and 30 years old.

Procedure. S was rotated with eyes closed, in a dark environment, to the required position. He then adjusted a luminous line to apparent vertical!' ;; at 15 sec intervals for 8 min. Between the adjustments the line was moved to and fro. After each 8—min—test S was released from the apparatus and waited up to 5 ™in before the next test started. All 3s underwent at

least two (in the average six to eight) tests in the following positions; 0 (head-up), 30, 60, 90, 120, 150 and 180° (head-down). The material from different experimental series was used, resulting in the varying number of tests per S; however, the attention is directed mainly to the analysis of the data from the individual Ss.

Variance calculation. The inter- as vieil as intra-individual variance of the AV has been established. The latter concerns the variance within the single 8-min-tests. It has been shown, that the AV during prolonged tilt changes (Müller, 1916; Miller & Graybiel, 1963a; IlcFarland & Clarkson,

1966; Schöne & Udo de Haes, 1968). In order to overcome the influence of this factor on the variance calculation, the variance of the differences between successive adjustments was computed.

RESULTS.

In fig. 1 the AV is plotted as a function of bodyposition separately fo:-the 13 Ss. The variability between fo:-the curves increases with increasing degree of tilt; it is obviously highest at 150°. The inter-subject varir,nc. values, as calculated from these data, are presented in Table 1.

The intra-subject (intra-test) variance is plotted in fig. 2 as a funct-.o:. of bodyposition for the 13 Ss. It also increases with increasing degree

Fig. 1. Inset figure: Bodyposition as stimulus and apparent vertical (AV) as

response, measured in relation to the subject; GV: gravitational vertical. Main f igure : Apparent vertical as a function oC body-position for 13 Ss ; each entry represents the mean of at least two (in the average seven) test-means (at least 64 single values). Peints under the dotted line represent A-phenomenon, points over this line E-phenomenon.

Fig. 2. Intra-test variance of apparent vertical as a function of bodyposition

for 13 Ss. The curves are based on the same material as presented in

fig. 1b; each entry is the mean of in the average 7 variance values.

Fig. 3a. Apparatus, adapted for countertorsion photography simultaneous with the

adjustment éf a line to the apparent vertical. The eye is photographed with

the camera (C) over a «ne way mirror (M) through which the luminous line

can be seen; the flashes are not shown in the figure.

I). Angle between apparent vertical and median bodyplane (MBP) as the sum

of ocular countertorsion (MBP-eye) and angle eye—apparent vertical.

300 30 60 90 120 ISO BODYPOSITION (degrm) 3.) 180 \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ 3 b) EYE MBP 60 90 120 150 BODYPOSITION (degrees) 180