Trillium PairBinocular Colour Vision

Home Colour Theory Bicovi Demos Whiter Demos Sky Blue Viewing Artwork FAQ's Links
  Bicovi > Demos > Helmholtz > Text


Helmholtz's "Treatise on Physiological Optics".

This page is the text of Helmholtz's observations that are discussed on:- 
and would normally be in a separate browser window to allow for back-and-forth comparison.

Binocular Colour Fusion pages; 505, 506, 507, 509, 512, 516 and 530.

'Sky Blue' references; Volume III, page 516;  Volume II, page 288.

Copyright Information

The following text is from

"Treatise on Physiological Optics; Volume III" by Hermann von Helmholtz
Translated from the Third German Edition 1910.
Edited by James P. C. Southall
Professor of Physics in Columbia University
Published by The Optical Society of America, 1924
available at

Electronic version of these documents copyright 2001, University of Pennsylvania.
See also the copyright information at the end of this document.

Problems with the OCR (Optical Character Recognition) document are
discussed at the above website.  Not all errors have been corrected.

502 The Perceptions of Vision (409, 410.

[...] It is well known how hard it is to discover a tiny object on an
extensive bright surface unless it has first been noticed in indirect
vision. GOETHE speaks, for instance, of the lark “lost in the azure
depths of space” (“im blauen Raum verloren”). On the other hand,
the gaze is attracted immediately by a somewhat larger object,
especially if it happens to be outlined distinctly enough to get a
glimpse of it by indirect vision; and if a person observes how he considers
an object before he knows what it is, he can easily notice how
his eyes traverse the contours. Thus the result of both habit and
practice must necessarily be to turn the attention to the contours.
In the case of contrast phenomena also, I have pointed out what a
particularly important factor the contours are.
    It might be natural to suppose that the parts of the retina where
white and black border on each other would be strongly stimulated
whenever, as a result of movements of the eye, elements of the retina
pass over from the black into the white. These elements, having had
some relaxation, would undoubtedly be stimulated more strongly
than those which had been already exposed to the white radiation.
However, I do not believe that this circumstance is of any essential
importance here, because it was shown in the experiments described
above that the direction of the ocular movements had no decisive
influence, and because the very first moment the contours in the
double images fall on the eyes their effect is appreciable, before any
after-images have had time to develop.
    On the other hand, PANUM'S theory, that the contours themselves
stimulate the retina more strongly, does not seem to me to rest on any
secure basis of facts; and, besides, it appears to be entirely unnecessary
for the explanation of the phenomena that are observed here. In the
case of contrast phenomena the difference of illumination or colouration
certainly does come out more strongly along a contour where two
fields are juxtaposed than it does when the fields are separated.
Indeed, the difference appears to be greater than would be expected
under the circumstances. But without taking after-images into
account, the phenomena of simultaneous contrast may be explained
as being due to the fact that the illumination at two points on the
retina can be compared much better and more certainly when they
are adjacent to each other than when they are farther apart; because,
as the eyes move, the adjacent points are liable to be exposed to the
same illumination in rapid succession. The fact that such a difference
appears relatively too large, and so leads to errors in our judgment of
the colouring, is in accordance with the general rule by which we are
apt to consider distinctly perceptible differences as being larger than

410.1 of the Two Eyes 503

those which are vaguely distinguishable. Perhaps, a distinctly perceptible
difference of this kind might be regarded as being a more
powerful psychic stimulus, and possibly this may be a partial explanation
of its greater tendency to hold the attention. But if afterimages
are left out, I cannot see any reason for assuming a more
powerful nervous stimulus.’
    When the two eyes are exposed to fields of different colour or
luminosity, similar phenomena of conflict are manifested. If a person
looks through a pair of highly coloured glasses, for instance, through
a red glass with one eye and through a blue glass of about the same
luminosity with the other eye, the external objects will seem to be
spotted with red and blue, the two colours frequently alternating with
each other. Generally this curious, restless fluctuation of colour is
most, lively at first, but presently, as the sensitivity for colours becomes
dulled, the appearance ceases to be so fickle and assumes a grey colour,
although it continues to waver here and there and from time to time
between a more reddish and a more bluish hue. Some observers are
disposed to regard the colour in this case as being a mixture, which
would therefore be pink for this particular combination. I have tried
over and over again in various ways to see the mixed colours, but I must
confess I have never been able to do so with any certainty whatever.
The peculiar characteristics of the objects have something to do with
which one of the two colours is visible. Brighter objects will be more
apt to look red, and darker ones blue. Possibly this may be because
the red sensation generally predominates with higher luminosity, and
the blue sensation with lower luminosity. Naturally, objects that are
red of themselves will also look red, and those that are blue will look
blue; because anything seen through a glass of the same colour as it is
will look brighter than it does seen through a glass of another colour.
Here, too, the matter of paying attention to one field or the other has
much to do with the effect. It is very difficult to fix the attention simply
on the colour of one of the fields, without being aided by contours
belonging to that field; but various observers (for example, FUNKE,
I myself) have been able to concentrate first on one eye and notice

I l/See A. CHAUVEAU, Rivalry between the visual fields in the stereoscope.
C. R.,152 (1911), 659-665.-C. 0. ROELOFFS, tfber Wettstreit und
Schwankungen im Sehfelde.
CRAEFES Arch., 104 (1921), 230-263. (J. P. C. S.)
2 Lehrbuch der Physiologie. 1. AuA. Bd. II, 875.
* M~~LLERS Archiu. 1838, pp. 61, 63.
4 hTeue Beitriige ZUT Physiol. des Gesichls. pp. 97, 99.
6 Programma Colleg. 118.
6 Uber Imdiation. 1852, p. 107.


what is seen by it, and then on the other eye in the same way. And
when this is done, the colour of the particular glass which is in front
of that eye will show up on the objects. FECHNER~ found it harder to
produce this variation by his own voluntary effort; and he concluded
that it was due to an involuntary movement or compression of the
eye, which, according to his observations, was simply conducive to
the change of the colour, without, however, aiding it to take place in
the desired direction. The experiment succeeds very much better
still when the coloured glasses are placed so as to reflect into the eye
images of dimly illuminated objects lying off to one side. Then, as the
attention is turned to one of these reflected images, although it may
be no more than a faintly visible shadow effect, immediately the
colour of that particular mirror will appear on the visual globe at the
proper place. And if an image reflected in the other glass happens
to be visible in the field at the same time and at the same place, and
the attention is concentrated on it, the other colour will come out also.
    I devised a systematic method of making this experiment by
mounting a piece of blue glass B and one of red glass R vertically on a
table, as represented in Fig. 75. A sheet of printed paper turned toward
B is placed against a dark screen C.
On the inner side of a similar screen D some other pattern is attached;
for example, a table of numerals, which cannot be easily confused with the
letters on the other screen. A white screen is placed at A; and the ob-
server’s eyes are shown at 0 and 0’. The illumination is regulated so that
the letters and numerals as seen in the images reflected in the coloured
mirrors will just be visible when the screen A is highly illuminated.
The reflected images of the letters and
numerals will appear to the observer to lie on A. Now when I try to
make out the letters, the ground invariably appears to be blue. On
the other hand, when I try to read the figures, it appears red. And so
by concentrating my attention on the image on one retina, I make
the corresponding coloured ground appear. It may be observed that
the contours, which in this case make the single impression prevail,
are borders between black and white, and yet, as the colour of the
ground becomes visible at these places, its intensity is not affected one
way or the other. Or, when the whole mixed illumination is taken

l Abhund.2. der S&ehs. Gea. d. Wisr. VII (1860), 39wo8.

411,412.1 $32. Rivalry Between the Visual Globes of the Two Eyes 505

together, it will be found that the letters on the left appear pure blue
on pale blue, and the numerals on the right pure red on pale red.
In the case of contrast phenomena, the attention would be directed
simply to the opposition between black and white, and not to the
blue or red; whereas it is just the opposite in the binocular experiments
which have been described here.
    A simpler form of the experiment, which I find works very nicely,
consists in gazing up at the sky with a red glass in front of one eye
and a blue glass in front of the other eye; only, they must be inclined
to the axes of the eyes, as in Fig. 75, so that faint traces can be seen
of the images reflected in the coloured mirrors by objects which are
off to one side. Now move one glass just a little, then the other, and
the reflected images will move likewise. These moving images may
he exceedingly dim and indistinct; yet, by watching one of them closely,
all at once the colour of that particular mirror will be seen to stand
out on the sky. It is really a marvellous sight when suddenly, as by
word of command, the blue sky becomes red all over, or the red sky
blue all over.
    There is absolute divergence of opinion between various observers
as to whether or not the compound colour is visible when two fields of
different colours are superposed binocularly. On the one hand, there
are H. MEYER, VOLKMANN, MEISSNER and FUNKE, who have never
been able to see the compound colour; and I must include myself also
among them. On the other hand, DOVE, REGNAULT, BROKE, LUDWIG,
S PANUM, and HERING are equally positive that they have seen it, not
only when the colours were pale and dim, but when they were satur-
ated. DOVE stated that he saw it when the colours were those of the
prismatic spectrum, which are the most saturated of all colours. A real
spectrum was projected on a screen and viewed binocularly through
two telescopes at once, the image being inverted in one instrument and
erect in the other. According to him, polarisation colours are especially
suited for observations of this kind. The method which he used for
this purpose consisted in adjusting thin sheets of mica or gypsum in
the proper position in front of a plate of black glass, which reflected
the light at the angle of polarisation. A NICOL prism was then held
in front of one eye and rotated in such a position as to transmit the
maximum amount of the polarised light reflected from the glass plate.
Another NICOL prism was placed in front of the other eye, but it was
rotated through an angle of 90” as compared with the other prism, so
that none of the reflected light would traverse it. Under these
circumstances, the sheets of crystal will be seen in colours by both eyes,
but the colours in one eye will be exactly complementary to those

506 The Pe-rcevtims of Vi&on 1412. 413.

in the other eye. Now in an experiment of this nature, both DOVE
and REGNAULT report that they have actually seen these complementary
colours fused binocularly into white. On the other hard, I have
performed this experiment over and over again, and each time it has
consistently and absolutely failed. It makes no difference whether
I use spectral colours or polarisation colours, invariably there is the
same conflict and alternation between the various simple colours,
but I am not able to see the compound colour, such as would be produced
if I had used pigments or the colours of tinted glasses. Incidentally,
I found that a quartz plate cut perpendicular to the axis
made a great improvement in these experiments. When the NICOL
prisms were rotated in front of the eyes, new colours would make their
appearance. But I always see both colours separately, one through
the other, as it were; and I can always tell instantly, without having
to shut one eye, what colours are there. In this case the bright white
background of the reflecting plate, which shows the mixed colour that
is said to make its appearance, is there for comparison with the colours;
and this is one reason why it is easy to notice the great difference in
these experiments between the binocular union of different colours
and their real union.
    It is a hazardous thing, I know, to contradict so many eminent
and reliable observers on a matter about which individual differences
are probably extremely great; and yet I do venture to mention here
some circumstances which in my own experiments, it is true, did
sometimes give the appearance of a mixed colour, although on more
careful investigation it was found that there was really nothing of
the sort, at least so far as my eye was concerned.
    However, the following preliminary statement must be made first.
Suppose a person is gazing at the binocular combination of two colours,
and at the same time can see also each component separately. For
instance, suppose the axes of his eyes are parallel and he is looking
at a blue field which adjoins a red field over to one side, so that there
is a double image of the line of separation between the two, blue being
superposed on blue on one side, and red on red on the other side,
whereas in the middle red and blue are superposed on each other.
Then the blue in the middle can undoubtedly be distinguished from
the pure blue on one side by its having more or less red added to it in
the field of view; and anyone knowing the rules for mixing colours
and being accustomed to see violet or purple compounded from blue
and red, might perhaps take for violet this blue compounded with red.
Even in the monocular field, owing to the contrast between a real violet
and some adjacent blue, or because the blue seems to be a covering
spread over the field or to belong to the total illumination, it may be

413.1 $32. Rivalry Between the Visual Globes of the Two Eyes 507

that the observer will resolve the actually existing violet into blue and
red. Examples of this kind were given in $24 in the preceding volume.
Thus there are some circumstances in which red and blue, that are
actually combined monocularly into violet, yet may appear separated,
just as binocularly superposed red and blue always do appear in my
own case. And so such an observer might possibly be betrayed into
thinking that, when he sees red and blue simultaneously, he really
does see violet or purple. But if the real compound colour made by
the two observed colours is exhibited, the difference between it and the
other will be very striking. The best and most accurate method of
producing this compound colour is as follows. Arrange a pair of blue
squares and a pair of red ones side by side like the squares on a chess
board; for instance, suppose the upper corner on the right and the
lower one on the left are red, and the other two blue. Now take two
double refracting, achromatic prisms of Iceland spar, and place one
in front of one eye’ and the other in front of the other eye. Adjust them
until the two images in each prism are one above the other. The double
images of the coloured fields will partially overlap; and so along the
middle horizontal line between the two upper squares and the two lower
ones, there will be a central strip in the field of view of each eye
consisting of a monocular mixture of red and blue, which will appear
therefore to be pink. Now make the axes of the two eyes parallel, and
gaze at the two fields so that their images will be binocularly superposed.
In this case there will be binocular overlapping between the
upper blue on the right and the upper red on the left, the middle pink
on one side and the middle pink on the other side, and the lower red
on the right and the lower blue on the left. Now when I perform this
experiment, I am positive that not the slightest trace of pink, as it
appears in the central strip, can be detected in the binocular combination
of blue and red, and that nothing can be seen there except
each of the two colours separately.
    PANUM insists that it is very important that the two colours that
are to be binocularly mixed should not be too vivid or too unlike;
otherwise, the rivalry between the two fields will be too intense and too
variable, and so the observer will be hindered from recognizing the
compound colour. Accordingly, I have employed H. MEYER'S method,
which was described previously in connection with contrast phenomena.
The coloured fields which were to be combined were first covered
over with a sheet of fine white paper so that the colours underneath
could only be dimly discerned. At first, when I superposed these very
pale colours, I really thought I did see the mixed colour. But when the
real mixed colour was placed alongside the two fields, I was again
aware of the conflict that existed between the two binocularly superposed


    If we have a varied assortment of papers of all colours, including
the greys, it is sometimes possible to find two colours which when
mixed by a double refracting prism will give a resultant colour precisely
like that of one of the other samples. Under these circumstances
the experiments are easier to make and even more impressive. I placed
a sheet of glazed green paper by the side of a sheet of pink paper, so
that the line where the two colours came in contact was vertical;
and then, horizontally across them, I laid a strip of grey paper, whose
colour was like that obtained by mixing the other two colours. The
whole was covered finally by a sheet of fine white paper. When I
examined these fields through a double refracting prism, in which
the two images were shifted apart horizontally, grey was mixed with
grey along the horizontal strip, but in the centre, both above and below,
pink and green were mixed, producing a grey also, which blended
imperceptibly into the horizontal band of grey. But, when the prism
was removed and binocular double images were produced, the band
where grey was over grey was very distinctly different from the places
where pink and green were superposed. And yet when I withdrew the
grey strip in the middle, I could not continue to detect the binocular
conflict between the two coloured fields; then all I noticed was what
was common to both colours, namely, the white.
    In other cases after-images are responsible for the production of
what appears to be a mixture. The arrangement used above is well
adapted for this purpose, that is, a band of grey paper above, with
pink and green below, the former on the left and the latter on the
right, these colours being such that when mixed by a double refracting
prism they give the effect of the upper grey. When the two lower
fields are superposed binocularly, all I can see at first is a lively
conflict between them. However, after continuing to look at them for some
time, presently the binocularly mixed field gets to appear like the grey
above, differing from it just a little, sometimes on the red side and
sometimes on the green side. But then if the red is covered with green
and I close one eye, the after-image of the green seems to me to be on
green, whereas in that part of the field where pink was before, now
the pure saturated green is visible. Now it is very obvious here that
the green modified by fatigue has really become very much like the
grey in the upper band. The same thing happens in the pink when the
green is covered. Thus the apparent mixing of the colours into white
in this case is because, owing to the appearance of complementary
after-images, the colours themselves, so far as sensation is concerned,
have become much more like the grey, until at last the colours are so
similar that the difference between them and their rivalry with each
other will no longer be noticed as at first when the difference was more

414, 415.1 $32. Between the Visual Globes of the Two Eyes 509

    There are some cases in which the induction of the colour of the
background, mentioned in Vol. II, $24, may produce an apparent
effect of binocular mixing over a small field of another colour. For
instance, I placed a horizontal band of blue on a red background,
and, keeping the fixation steady for a long time by fusing a little black
dot on the blue with a similar dot on the red, I gazed at the binocular
double images. All I observed at first was the conflict between red and
blue in that part of the field where these colours overlapped; but,
finally, I noticed that real violet occurred. However, when I shut one
eye, I could discern with the other eye alone the induced red on the
blue band.
    Lastly, there is a case, mentioned by H. MEYER and PANUM,~ in
which I find the most striking appearance of all of an effect similar to
monocular mixing. On the right there is a yellow field with a horizontal
pink band on it, and on the left a blue field with a vertical band of the
same pink colour. If the yellow and blue are superposed binocularly,
so that the two pink strips appear to form a cross, the arm of the cross
on the left, which falls mainly in the yellow field, will undoubtedly look
much yellower than the opposite arm, which falls mainly in the blue
field. Where the two fields overlap in the centre, pure pink will be
visible, or rather it looks to me here as if the yellowish pink of one
band passed, so to speak, underneath the bluish pink of the other
band, without being blended with it. PANTJM thinks that the yellowish
and bluish tinges of the pink are due to its being binocularly mixed
with the colour of the opposite field in each case. A point to be noted
is that the variation in the two pink bands is most in evidence when
the gaze is allowed to wander; because then the band lying on the
yellow ground gets the blue after-image of the yellow, and the band
lying on the blue ground gets the yellow after-image of the blue.
However, even when the eyes are steadily fixed, the effect is undoubtedly
produced, although not to the same extent. Yet it can be
shown that here also the phenomenon is primarily one of contrast.
Thus, even when one eye is closed so as to preclude anything like
binocular mixing, the change in the colouring of the pink still persists.
The pink band continues to be as yellowish as it was before when the
eye opposite the yellow field is closed. At this instant, it is true, the
yellow which pervades the pink like a kind of yellow mist, does disappear,
but the apparent colouration of the pink itself still persists
without the slightest alteration. Similarly, if the other eye opposite
the blue is closed, the pink band on the yellow appears the same bluish

I Physiologische Untersuchungen tiber das Sehen mit zwe-i Augen.
Kiel 1858. p. 41.
Figs. 27 and 29.

510 Perceptions of Vision [415, 416.

red. It follows, therefore, that the change in the pink cannot be due
to binocular mixing, or at least that it cannot be due to that alone, but
is a constant effect. From the very beginning, even in monocular
vision, the pink on the blue field appears more yellowish by contrast,
while that on the yellow field appears more bluish. Undoubtedly, the
contrast effect is much more vivid at the instant when the two fields
are binocularly superposed; but once it has been brought out in this
vivid way, it does not disappear again, even when one eye is shut,
and binocular coincidence is abolished. We were at much pains to
explain in Vol. II, $24, that in any case of contrast judgment of colour
was unreliable within a certain interval. Owing to secondary considerations,
we are apt to consider the observed colour as lying more
on one side of this interval than on the other. It is possible in the
present instance that the binocular overspreading of the complementary
colour on the ground where the pink band is may be a contributory
cause of this nature. I may add that I shall have occasion presently
to speak of the theory of binocular contrast again.l
    As to the theory of the binocular combination of colours, the only
difference between it and the monocular mixing of colours, on YOUNG’S
theory, is that in the former case the nerve fibres for the three fundamental
colours are distributed over both retinas, whereas in the
latter case they are distributed only over one retina. Either the three
different kinds of fibres at a given point on one retina have the same
local sign, or else, on the supposition that the local signs are different,
there is no possible experience that could enable these fibres to be
stimulated by objects which were in different parts of the field. Hence,
there cannot be any reason for separating the localization of these
sensations with reference to the directions in the visual field of the eye
in question. Accordingly, the various sensations of these fibres are
blended into a resultant sensation, that is, the sensation of a mixed
colour ; and usually this will be the visual token of some definite
property of the locally simple object which happens to be at that place
on the visual globe of the eye. And yet, as we have seen, even when
colours are mixed monocularly, there are certain cases when we imagine
we see one of the combined colours through the other. It may be due
to the irregular distribution of the light, or to the movement of some
image that is limited locally, or to the presence of some portion of the
colour all over the field of view, but it happens whenever we are
induced to separate a coloured illumination or mantle from some
coloured object.

l Concerning the subject of binocular colour mixing, see Note 1 at the
conclusion of this chapter.-K.

416,417.l g32. Rivalry Between the Visual Globes of the Two Eyes 511

    When corresponding portions of the two retinas are illuminated
differently, the impression that is produced is one that can never be
obtained by uniformly illuminating a simple object on all sides. And
yet, perhaps as a result of training and not from any innate mechanism
of the nervous system, both colours will be attributed to the same
region in the common field of view; and so two colours will be seen in
the same field, each being perceived as separate from the other. This
visual picture is certainly very much like those cases of monocular
mixing in which two coloured objects are seen, or appear to be seen,
one behind the other at the same place on the visual globe. Many
observers, including myself, never see this effect any other way. The
attention may waver, being diverted first to one field and then to the
other, making us aware of a conflict. Incidentally, something like
this conflict, only much less pronounced, may be also noticed in the
monocular field by using an unsilvered plate of glass to reflect the
image of an object at the same place where another object as seen
through the glass happens to be. The two images in this case should
both be equally bright and well defined, but entirely different in
pattern. Then we may look at either of the two, and the other one
will retire more or less out of sight, although it may never disappear
completely, as it does when the images are binocularly superposed.
If necessary, the two images can easily be separated by moving the
mirror slightly.
    On YOUNG’S theory, the apperception of mixed colours is invariably
the result of projecting three different sensations of colour at the same
place on the visual globe. Even when colours are monocularly mixed,
this apperception will depend on a mental act, which will vary according
to circumstances; that is, it will depend on the decision we make
as to whether these circumstances are to be considered as a visual
token of some simple quality of one object or of two different qualities
of two objects. While this is the case, yet, on the other hand, it might
conceivably be possible that the difference between the impression
produced by combining two colours binocularly and that produced
by combining them monocularly may be disregarded, and the two
colours considered as being united in the former case in the same way
as they are in the latter case. According to YOUNG’S theory, the mixed
colour is really nothing more than the integration of three different
kinds of impression, which have otherwise no mutual action on each
other, but which all have the same localization. Naturally, therefore,
the mental decisions on which their union or separation depends may
be very different for different observers, according to each individual’s
particular training and variety of experience. Such being the case, it
goes without saying that the union of colours which are very much


alike, and which have therefore much in common and not much that is
different, will be easier to effect than the union of extremely dissimilar
colours. Besides, there may often be minute differences between the
impressions on the two eyes produced by the same real object. For
instance, one eye may be more fatigued than the other, or the light
may be very brilliant or coloured, perhaps entering one eye from the
side and being diffused in it, etc. And so we may get in the habit of
equalizing some of these minor differences unconsciously. Indeed, if a
field producing such an impression is placed close by the side of
another one in which two like colours are superposed, the conflict
between the two impressions will be noticed, even when they are not
very dissimilar.
    Lastly, the binocular combination of two fields which are different
as to colour or illumination is exhibited in the case of stereoscopic
drawings in an extremely remarkable and characteristic way. Thus,
for example, in a stereogram intended to represent some object, suppose
that a certain area in one of the pictures is shown in white and in the
other picture in black, or suppose that this particular place is coloured
differently in the two pictures (although it is better for the colours not
to be too much unlike) ; then when the two views are combined
stereoscopically, this area will shine with a certain Lustre, while all the
other parts of the body where the illumination and colouring in the
two pictures are the same will appear dull by comparison. Incidentally,
this appearance of lustre or dullness has absolutely nothing to do with
whether the surface of the pictures themselves is really dull or lustrous,
provided that in the latter case they do not send any reflected light to
the observer’s eye.
    For example, the outlines of the model of a crystal may be
represented by two drawings on a stereogram, in one of which the lines
are white on black and in the other black on white; and when they are
combined in a stereoscope, the impression will be produced of looking
at an object consisting of some dark shining substance like graphite
lying on a graphite surface. A stereogram of this nature is given in
Fig. Q, Plate IV.
    Similarly, also, places can often be found on photographic
stereograms representing brilliant objects (such as the bright foliage of
plants, satin, etc.) where the reflections of light were unequally bright
in the two views, so that the effect of lustre is produced when they are
fused. One of the most remarkable examples of this kind is afforded
by instantaneous photographs of ripples on a surface of water illuminated
by direct sunlight. On looking at a real object which glitters in
this way, it can often be noticed that more light is reflected by certain
spots in one eye than in the other.


417, 418.1 $32. Rivalry Between the Visual Globes of the Two Eyes 513

    I am inclined to think that this is likewise the explanation of the
appearance of lustre at these places in a stereoscopic view where the
illumination in the two pictures is different. Light falling on a dull
surface is radiated uniformly in all directions, so that the surface looks
just as bright from one place as from another. And so, under normal
conditions of vision, it always appears just as bright in one eye as in
the other. But the reflection of light from a lustrous surface is more or
less regular. Such a surface may exhibit numerous tiny rugosities of
various dimensions; but when it is smoothed and polished so as to
have an approximately definite direction on the whole, the incident
light will be reflected from it mainly in the same direction as all the
light would be reflected by a mirror. Now under such conditions, it is
quite possible for one eye to be in the direction where the light is
reflected, while the other eye is not; the result being that the surface
will look very bright to one eye and very dim to the other eye. And so
in looking through a stereoscope, if the appearance of the image of a
part of a body is very different in one eye from its appearance in the
other eye, the resultant visual impression will be just the same as would
be actually produced by a lustrous surface, but never by a dull surface.
Consequently, this place in the stereoscopic view appears to be lustrous.
    Similarly, also, when a body that sheds lustre is surrounded by
coloured objects, it may reflect light of one colour to one eye and light
of another colour to the other eye, and thus be seen in different colours
by the two eyes; whereas, under normal conditions of vision, the colour
of a dull body will always appear necessarily the same for both eyes.
And so in a stereoscopic view, if the colour of the same surface is
different in the two pictures, the result will be a visual impression such
as only a lustrous object can produce. As a rule, the colour of the
lustrous body itself will be mingled with that of the light of the two
reflections, and the latter seldom contain just the one pure colour only.
And so the differences of colouration in these reflections from lustrous
I bodies for the two eyes are not apt to be very great; and that is why the
effect of lustre can be produced better by combining colours which are
not very much unlike than by combining very brilliant colours which
are very far apart. The latter will exhibit conflict rather than lustre.
    According to WUNDT’S experiments, the best way of combining
two coloured fields so as to get the effect of lustre is when there is
about the same contrast between each colour and the background. If,
however, the contrast is much greater for one colour than it is for the
other, this effect will be weaker, because then the former colour will
prevail in the conflict between the two fields and subdue the latter.
For instance, if two coloured squares of the same size, one bright yellow
and the other dark blue, are laid together on a white or black back-

514 Perceptions of Vision

ground, and then binocularly superposed, the contrast between yellow
and white in one case, or between blue and black in the other case, will
not be great enough; and so the lustre will be much weaker than it would
be if the two coloured squares were laid on a grey ground, where the
contrast was the same for both.
    Moreover, the effect of drawing some pattern on one of the coloured
squares would be to give it an advantage in the conflict, and so to
impair the effect of lustre.
    Without using stereoscopic pictures at all, binocular lustre may
also be produced by simply looking at variegated objects through
glasses of two different colours. The object, for instance, might be
some pattern executed in blue and red, which was viewed with one
eye through a blue glass and with the other eye through a red glass.
As seen through the blue glass, the blue portions will appear bright
and the red portions dark; whereas with the other glass it will be just
the reverse. And so the pattern viewed in this way will show lustre
to a high degree. A remark which DOVE makes in this connection is
worth noting. He states that when one colour or the other happens to
prevail entirely in the conflict between the two eyes, the lustre
disappears; but at the moment of transition, when both colours are seen
side by side, the lustre shows.
    A characteristic thing about metallic lustre is that frequently the
regularly reflected light itself is coloured already, and not white like
the light of transparent substances. Thus bodies that exhibit the
iridescent colours of thin films, such as the brilliant plumage of birds
and certain highly coloured refrangible substances like indigo, are apt
to show metallic lustre.
    The phenomenon of stereoscopic lustre is particularly important in
connection with the theory of the activity of the retinas of the two
eyes. The statements of various observers as to the result of the
binocular fusion of unlike images are so different, that, were it not for
this phenomenon, doubtless, we never should have known positively
that the visual impression produced by the action of two different kinds
of light on corresponding places on the two retinas was absolutely
different from that produced by the action of two homogeneous kinds
of light on the same retinal places. If one eye sees black, and the other
eye sees white on the corresponding part of the visual globe, the
impression will be that of a surface shedding a pale lustre. But if the
white light, which fell previously on one side only, is distributed
uniformly over both sides, that is, if grey is combined with grey, the
impression will be that of a dull grey, absolutely different from the
lustrous white effect in the first instance. The same thing is true
with respect to the lustre produced by binocular union of different

$32. Rivalry Between the Visual Globes of the Two Eyes 515

    The same conclusion, indeed, may be inferred from the fact that
the impression obtained by the binocular fusion of two stereoscopic
pictures is that of a body, and not as if all the lines were on the same
sheet of paper. But undoubtedly in this case the movements of the
eyes have an important influence, which we never do get rid of entirely
except in the case of instantaneous,illumination by the electric spark.
    I might add that I have taken stereograms that show stereoscopic
lustre and viewed them by the illumination of the electric spark, and
that then also the impression of lustre is perfectly produced. This is an
important fact, because it shows that the lustre does not depend on
the alternation of colouring and illumination that is responsible for
rivalry. My experience is that, when the attention is relaxed, there
are never more than about eight alternations per second in the conflict,
and generally the frequency is much less than this. On the supposition
that the luminous impression on the retina lasts a small fraction of a
second, no appreciable change can occur during this time as the result
of rivalry between the two fields. And yet in this brief interval we can
notice that the two different impressions on the two visual globes are
seen at the same time and in the same place in the common field of view.
    Incidentally, the impression of lustre may be produced by images
and objects even in monocular vision. This happens, for example,
when the illumination changes rapidly in consequence of movements
of the observer. Then the elements that constitute stereoscopic lustre
are not observed simultaneously, but in quick succession. So also,
objects in motion may give the appearance of lustre, provided the
illumination at particular places on them varies in quick succession.
This is the explanation of the glitter of ripples on the surface of
water. Even when the variation of the illumination of the parts of the
surface is simply an imitation of the way light is scattered by diffused
reflection, it is sufficient to give the effect of lustre. WUNDT produced
monocular lustre by looking through a plate of glass at a dark square
on a dark ground, where an image of a brighter square on a bright
ground was very nearly superposed by reflection in the first surface
of the glass. If the reflected image was apparently exactly at
the place where the dark square was, the lustre disappeared, and then
only the mixed colour was visible. But when the reflected image
appeared to be behind the other one, the lustre showed. If the reflected
image was in front of the other, it seemed to shine better. The idea
obtained in this case was as if one were beholding another square which
was beyond the first one, and were seeing it through the latter as if it
were a reflected image of it. It was this that gave the appearance of
lustre These experiments indicate very clearly that the special
qualities of the colouring do not matter so much, and that the im-

516 The Perceptions of Vision 1420.

portant thing is to produce the illusion that another image is reflected
in the observed surface.
    Sometimes the appearance of transparency is produced also by the
binocular superposition of two fields of unlike colour. WUNDT called
attention to this effect. For instance, if a bright yellow square and a
dark blue one, both lying on a white ground, are binocularly superposed,
but not exactly coincident, the blue appears to be transparent
where it is superposed on the border between yellow and white. But
where the yellow is superposed on the border between blue and white,
this transparent effect will be lacking. On the other hand, when
the ground is black, it is the yellow, and not the blue, that looks
transparent. The general rule seems to be that the field which looks
transparent is the one for which the contrast with the ground is the
greater. This is in accordance with the objective law, that anything
seen through a translucent medium, which is itself distinctly perceptible,
is always seen indistinctly; whereas the border of this medium,
not being concealed by some other translucent substance, will usually
be well defined.

    Lastly, some phenomena have yet to be discussed, which should
be, or at any rate may be, interpreted as being contrast between the
sensations in the two eyes.
    Let us mention first a matter which FECHNER noticed especially;
and that is the extraordinary acuteness of perception of minute
differences in the instantaneous colour-tuning (Farbenstimmung) of the
two eyes or mode in which the eyes react to colours, when the binocular
image of a tiny luminous object seen against a black ground is resolved
into separate double images by changing the adjustment of the eyes.
Suppose, for example, that one eye has been closed for some time while
the other eye was exposed to luminous white surfaces; immediately
after opening the closed eye, there will be two double images of a
white band on a black ground, and the one belonging to the fatigued
eye will appear darker and at the same time more violet than the other
one belonging to the dark-adapted eye. But if the surface exposed to
the open eye had been coloured, the colour of the image in this eye
afterwards would be complementary to that of the inducing field, but
the colour of the image in the other eye would be like that of the
inducing field. While the two images in this case are being compared,
the complementary colour in the fatigued eye will continue visible
very much longer than it would do if both eyes had been “colourtuned”
alike by having both been exposed to the same colour in the
same way. For example, without the help of double images in this
fashion, it is extremely hard to perceive that there is a bluish tinge in


the after-image of a white surface of moderate brightness; and yet this
bluish tinge will be evident at once, as soon as it can be compared with
the apparently bright orange-yellow image seen by the eye that has
been resting. When there is too much difference between the bright-
ness of the two images, the comparison may be greatly facilitated by
proportionately reducing the brightness of the image in the exposed
eye, either by looking at it through a tiny hole in a piece of black paper,
or by viewing it through a double refracting prism which will resolve
the image of the bright band into two, each half as bright as the original
one. Or the image may be viewed through a grey glass, provided we
are certain beforehand that the glass itself is absolutely colourless.
    These experiments prove that a very accurate comparison can be
made between the sensations of colour at approximately corresponding
places on the two retinas. Apparently, indeed, the comparison can be
made more accurately in this way and for a longer time than when the
colours have to be compared at the same place on the retina of one eye
alone. Thus, suppose it is desired to compare the colour which cor-
responds to the sensation of the retina for white, say, with the colour
which it seems to resemble in the eye which has not been fatigued; then
it will be necessary to develop a good after-image by gazing steadily
at a white object on a black ground, which must then be projected on
a uniform white ground. In this method the necessity of keeping the
fixation steady involves considerable strain, and that may have some
effect on the course of the process. And, besides, there is the further
disadvantage of not being able to reduce the intensity of the bright
image as desirable. But, worst of all is the fact that the limited
afterimages on one retina quickly disappear and therefore cannot be per-
ceived except for a brief space; because it is hard anyway to notice
constant differences of luminosity or colour between two different
places on the retina which have not been revived by change.
    We saw in Vol. II, $24, that the tendency always is to regard
differences of luminosity or colour which can be perceived distinctly
as larger than those which are just vaguely perceptible, and that, in
fact, most contrast phenomena were due to this peculiarity. In this
particular case, the fact that the unchanged image always assumes
the opposite phase of colour and brightness from that of the changed
image, is a manifestation of a contrast effect of this kind. Thus the
pure white in the unfatigued eye is made to look yellow by the side of
the violet-grey in the eye which has been exposed to white; or if the
latter is coloured pink by the after-image of green, the former will
he made to look green, etc.
    Instead of one of the double-images being coloured by an afterimage,
it may also be coloured directly by viewing it through a piece

518 The Perceptions of Vision [421, 422.

of coloured glass. But here, too, as is characteristic of contrast
phenomena, we find that a faint colour is apt to produce a much more marked
contrast effect than one that is highly saturated. A piece of greenish
window glass or yellowish bottle glass will enable us to see the
complementary colour on the image in that eye much more distinctly
than it can be seen by looking through a piece of highly coloured glass,
although in the latter case the image in the other eye can be reduced
to the same luminosity as that of the coloured image by viewing it
through a suitable piece of grey glass.
    As a matter of fact it is possible to have a contrast between colours
lying on corresponding places on the two retinas. Place a strip of black
on a white ground, and after separating it in double images, insert a
blue glass in front of one eye and a grey one in front of the other eye,
the two glasses being about equally dark. Then one of the images of
the black band will appear to be surrounded by prominent blue, and
the other image by prominent white, while over the rest of the ground
blue and white will be superposed more or less uniformly. In this
case the white that comes out around the edge of the black strip will
be decidedly yellowish. On removing the two pieces of glass, yellowish
white will be found to appear where blue prevailed before, and bluish
white where it was previously yellowish.
    The effect of substituting a yellow glass for the blue glass in this
experiment will be to interchange yellow and blue in the two images
wherever they occur.
    No doubt it must seem very curious that the effect of the border
around the black band is to attract the attention to the adjacent
white and separate it so completely from the overlying blue in the
common field of view that this white does, in fact, look yellowish.
This yellowish white, by the way, exhibits also the characteristic of
a contrast colour by persisting for a brief time even after the eye behind
the blue glass has been shut tight. It may be recalled, in connection
with the phenomena of coloured shadows (Vol. II, §24), that, when
once the judgment has decided about the nature of the colour, this
impression persisted even after the contrasting colour, whose presence
had been responsible for the mistake, was removed from the field of
    In the preceding experiments the contrast was developed by comparing
two colours in the rival visual fields. But the effect of monocular
contrast may be enhanced also by binocular comparison with the
complementary contrast. Place a piece of pink paper by the side of
a piece of green paper so that they touch in the middle; and on each
of them lay a strip of white paper near the border between them. Then
gaze at the two strips with both eyes; as a rule, no contrast colour-

Rivalry Between the Visual Globes of the Two Eyes 519

ation will be noticed on either of them, unless after-images have already
been developed of the two colours. Now if one eye is closed, while
the other looks at one of the white strips through a black tube, a faint
complementary colouration will, indeed, be observed. But if a black
tube is held in front of each eye, so that the right eye sees one of the
white strips and part of the pink ground, and the left eye sees the other
white strip and part of the green ground, but without the two strips
being binocularly superposed, the complementary colourations on the
white bands will come out to an extent which can scarcely be seen by
any other method. If, without keeping the gaze riveted on any one
place, the experiment should be continued for some time, the contrast
effect goes on increasing; and then, of course, the after-image of the
ground will be more and more intense. But the right eye sees only red
ground and the left eye sees only green ground, and hence, no matter
how the eyes move, the background can develop nothing but green in
the right eye and red in the left eye; and so the effect of contrast must
needs be heightened.
    The above would be a successive contrast depending on after-images.
If at the beginning of the experiment the eyes were focused as soon as
possible where the white strips should be, the contrast colours will
be seen too, but they will be much less intense. However, in the
way this experiment was performed, it was particularly easy to see
after-images of the ground by comparing the colouring in the two visual
fields; and so I thought it necessary to devise some method by which it
could be certain that there was no after-image of the ground. Accord-
ingly, I attached two vertical parallel strips of paper to a plate of
glass, the one on the right being black above and grey below, and the one
on the left grey above and black below. The plate of glass itself was
placed over a flat surface which was covered with red paper on the
right and with green paper on the left. Thus the strip of paper on the
right was on a red ground and the strip of paper on the left on a green
ground. However, before beginning the experiment, a sheet of white
paper was inserted between the glass plate and the coloured ground so
as to cover the latter entirely. Then I gazed at the two grey-black
strips and binocularly superposed them on each other so that the
upper and lower halves of the resultant image consisted of the black
half of one strip and the white half of the other strip mutually
overlapping. A white mark was made in the centre of each strip which
served for a fixation-mark, so that when these two marks were fused
binocularly I could be absolutely sure of keeping the common image
of the grey-black strips. After these preliminary arrangements had
been made, the sheet of white paper was withdrawn so as to leave
the coloured areas behind it exposed; and then I could undoubtedly


detect some traces of contrast colouring, but they were exceedingly
faint. The grey on green had a reddish look, and the grey on red was
rather greenish. However, all that was necessary to make the colours
both come out in full intensity was simply to move the eyes a little from
right to left and back again. The faint contrast colourings which were
detected at first were fainter than they would be in monocular contrast;
and when white was substituted for grey, they were fainter still.
    Accordingly, the pure effects of simultaneous contrast on the two
grey strips were diminished by binocular comparison. By bringing
the grey in the visual field of one eye close to that in the visual field
of the other eye, binocularly, the two greys could be compared more
accurately than was possible before in the monocular field, where the
two strips were separated from each other by wide intervals of green
and red. Therefore the phenomena of successive contrast depending on
variation of the sensation by after-images are entirely different in
behaviour in this respect from those of simultaneous contrast, which
were regarded as mistakes of judgment, In binocular comparison the
former show up better still, whereas in the case of the latter the effect
of this comparison is to correct the errors of judgment.
    In the form of experiment which was described above, the grey
strips were not allowed to be binocularly superposed on the coloured
ground, but were, so to speak, fused with the black. However, by a
change of convergence of the eyes, the images of these strips can be
shifted so far apart that they merely touch without overlapping.
Adjust them in this way, with the sheet of white paper interposed at
first, and notice how the grey looks alike on the two strips. Then
withdraw the sheet of white paper and expose the coloured ground
which was behind it. The strip surrounded by red, which is binocularly
superposed on green, will appear decidedly green; the other one
surrounded by green, and superposed on red, will appear decidedly red.
The impression of a binocular mixture of the grey with each of the
two colours is really quite startling. Now if the sheet of white paper
is again interposed between the glass plate and the coloured background,
the colourations instantly disappear, as they would necessarily
do if the colours of the ground had been mixed with the grey.
    But another experiment proves that what we obtain here is not
a mixture. When the strips are seen in their complementary colours,
suppose I close my right eye, so that only the strip surrounded by
green remains visible. Then, although a kind of red veil seems to
extend over it, due to having the red binocularly superposed on it, its
own natural colour is left, that is, grey, and yet as reddish as it was
before; which would not be possible if the reddishness of the grey
were simply the effect of its being (binocularly) mixed with red; for

Rivalry Between the Visual Globes of the Two Eyes 5 2 1

as soon as the red disappeared from the mixture, the original colour
would be obliged to assert itself, more likely becoming greenish by
contrast. I myself am disposed rather to think that the explanation of
these experiments is as follows. We know already that when grey is
binocularly superposed on black in the visual field of each eye, the hues
of the two greys can be compared with much precision, and that the
result of this direct comparison will be to diminish effects of monocular
contrast that might have a tendency to make us suppose that the two
greys were different. On the other hand, in the experiment just de-
scribed, grey, which was surrounded by red and which it would be
natural therefore to regard as greenish, was binocularly superposed on
green; while the other grey, coloured reddish by contrast with the
surrounding green, was binocularly superposed on red. The mere fact
that the two greys were binocularly superposed on two different vivid
colours may make any comparison between them very unreliable, and
therefore heighten the contrast.
    If a white surface is afterwards interposed, enabling the eyes to
revise their judgment as to the white, the contrast vanishes immediately.
The contrast between the two grey areas will also be made to
disappear at once by interposing a black surface, and then they may
be accurately compared without danger of mistake. On the other
hand, when one of the eyes. is closed, there is nothing left by which
the judgment might be corrected, and so the contrast persists.
    The results of these experiments thus far may be summarized as
follows :
    Let a and b denote the two images side by side near each other
in the binocular field, as seen by eye A and eye B, respectively; these
images being superposed on the backgrounds a and b, respectively.
Then a very accurate comparison may be made between the objective
colourings of a and b, or between their colourings as modified by
afterimages, whenever the grounds a and b are both of one colour. But
when the colours or illuminations of a and b are different from each
other, the comparison will be very unreliable. When the former
condition exists, it interferes with monocular simultaneous contrast,
whereas the latter condition is conducive to it.
    Just as in the case of a number of experiments on monocular
contrasts, there are some other experiments on binocular contrast where
the fact of our being accustomed to separate the objective colours of
bodies from the colour of a surrounding illumination constitutes a

522 The Perceptions Vision [424, 425.

    FECHNER'S so-called paradoxical experiment1 should be described
here first. Look at a white surface and open and close your right eye
alternately; then, at the moment you close your eye, so that the white
surface is exposed only to your left eye, it will look a little darker than
it did when both eyes were open. It is natural to suppose that by
cutting off the light from one eye the image would be darkened, as it
actually is; but the effect is so extremely slight that many people can
scarcely see it at all. Now vary the conditions by interposing a rather
dark piece of grey glass in front of the right eye. Then, on opening the
right eye, the white surface will look darker than it did before, and on
closing this eye, it will look brighter-exactly opposite to the previous
case. That is, the luminous area appears to be darker when more light
falls on the eyes, and to be brighter when less light falls on them! If the
experiment is repeated, each time using a brighter piece of grey glass,
presently this paradoxical effect ceases and changes to the first effect
which was obtained without using any glass at all; that is, the surface
begins to appear brighter when the eye that was closed is opened. On
the other hand, when the experiment is tried with darker shades of
grey glass, a limit is reached finally when it ceases to matter whether
the eye behind the glass is open or shut, because the light that enters
this eye is too slight to have any effect. The maximum effect is obtained,
therefore, with a glass of medium darkness. The glasses which
FECHNER himself used in this experiment transmitted between 3 and
5 percent of the incident light. Instead of having an assortment of
various shades of grey glass, AUBERT'S instrument called an episcotister
answers the purpose and is very convenient.2
    In order to be sure that changes in the size of the pupil had no
effect in this experiment, the observer should gaze with his free eye
through an opening of smaller diameter than the pupil. In all these
experiments small openings in pieces of black paper may be used for
darkening the image instead of dark glasses.
    One interpretation that might be given to this paradoxical experiment
is to suppose that, under certain conditions, the sensation of light in one
eye tended to lower that in the other eye, as if there were some

1 Abhandl. der Sachs. Ges. d. Wiss. VII (1860), 416463.
2 (This note was added by HELMHOLTZ on page 856 of the first edition.) The
episcotister is composed of two black discs made of brass, which are
mounted together, one in front of the other, and in each of which four 45”
sectors are cut out. The discs can be adjusted relatively to each other so
as to leave four slits open whose angular widths may be anywhere between 0”
and 45”. By rotating them rapidly the same appearance and effect can be
produced as is obtained by using a piece of grey glass, and the amount by
which the light is reduced can be computed easily and exactly. The
instrument was described by AUBERT in his Physiologic der Netzhaut, pp. 30,
34, 283. A similar device had been previously used by TALBOT. See POGG.
Ann., XXXV (1835), p. 459.

Rivalry Hrtween the Visual Glohes o.f the TWO Eyes 523

antagonism between the two retinas. However, by making a slight
modification in the experiment, I have been able to show that some-
thing of an entirely different nature is involved here.
    The observer should take a position where there is some white
object before him clearly outlined in the field of view. A white door
opposite the windows will answer the purpose. He must look at the
door and select a piece of dark glass that is found to be suitable for
performing FECHNER'S experiment nicely. This piece of glass is held
in front of the eye and a sheet of white paper interposed just behind
the glass between it and the door, so that it will hide the door and
occupy the whole field so far as this particular eye is concerned. By
turning the paper more or less obliquely to the incident light, it is
easy to regulate the illumination so that it will look just as bright as
the door beyond it. Now if the experiment described above is repeated,
the result will be just opposite to what it was there. The effect of
opening the closed eye behind the glass and the sheet of paper will be to
make the door appear a trifle brighter, as if a sort of luminous haze
had descended on it. This is the binocular image of the white paper
superposed on the door. Having verified this, the observer should
then take the paper away and expose both eyes to the door, in which
case the door will seem to be considerably darker than it was, although
the brightness has not changed at all at those places in the two visual
fields where the door appears to be.l
    This variation of FECHNER'S experiment shows that there is no
question here of a change in the sensation of the light, but that it
is simply a matter of changing our opinion as to the real colour of the
white object. If one field is dark all over (as is the case when this
eye is closed), or if it is dimly and uniformly luminous all over (as
is the case when the sheet of white paper is seen through the dark
glass), this uniform illumination, extending far beyond the confines
of the field of view that corresponds to the door, is not attributed to
the actual colour of the door, but our judgment of this colour is derived
entirely from the information received through the other eye which
perceives the outlines of the door. Variations of illumination in the
first eye can do no more than produce the effect of a dark or bright mist
settling down on the door and the other objects. But if the outlines
of the door can also be perceived by the shaded eye and appears to it
to be dark grey, this grey will seem to belong to the door just as much
as does the white in the other eye; and the result is then that the door
itself looks darker, like a grey body sparkling with white. But, of course,
this shading of the door will not be shown if the darkening produced

* Mr. HERING has also noticed that it all depends in this experiment on
whether the surfaces seen by the shaded eye are limited or unlimited. See
Reitriige zur Physiologie , pp.311,312.


by the glass is so slight that the additional light entering the other eye
is merely noticeable as light; or if, on the other hand, the darkening
due to the glass is so great that the objects can scarcely any longer
be seen through it.l
    Similar results are obtained in the case of monocular vision in the
experiment made by SMITH and BR~CKE (see Vol. II, p. ZSS), which
FECHNER calls the “side-window experiment.” I found that this
experiment could be made in another way, by which the conditions
for obtaining the effect could be better regulated than they could be
in the original method. I had a plane parallel plate of uranium glass
cut in half. This glass does not seem to have any colour at all by
candle light, because it absorbs only the violet and some of the blue
rays, and there is not much radiation of this kind from the light of
a candle. Unless the substance of the glass itself happens to be brightly
illuminated, white objects seen through it by daylight appear to be
slightly yellowish. But when the glass itself is exposed to the direct
rays of the sun, an intensely green fluorescence will be radiated from
all parts of it. Now suppose that I place one of the plates in front
of one eye and the other in front of the other eye, and that no light can
reach the eyes except that coming from the object. If this is a white
area on a black ground, and if I see it, in separate double images, the
two images of the white area, of course, will be seen in the same yellowish
white colour. But then if I let the direct rays from the sun fall on
one piece of glass, the visual field of that eye will be filled with the
green fluorescent light; and after my eye has moved a little, the image
of the white area in it will look pink, although it is still flooded with
green light, while its image in the other eye will appear brighter and
greenish, although objectively it is pure white. Thus in this case,
for the eye that is looking through the fluorescent piece of glass, where
the whole background is uniformly overspread with faint green light,
the limited white area is so completely separated from the diffused
green, that even this white assumes the pink colour that is produced
when the eye is fatigued with green. In contrast, with it the other
image, which is not green, appears greenish.
    In SMITH'S original experiment, as described in Vol. II, page 289,
it was the red light penetrating through the sclera that made the image
on that side look darker and blue-green, and the other one red. This
red light can be rendered visible by illuminating the eye from one side,
and then gazing at black letters on a white page, when the former will
often look bright red. At the same time, on resolving the image of a
black spot on a white ground into separate double images, the image

1 See Note 2 at the end of this chapter.-I<.

Rivalry Between the Visual Globes of the Two Eyes 525

belonging to the eye that is illuminated from the side will, of course,
look reddish by comparison with the one in the other eye. On the
other hand, if green or blue light is concentrated by a lens at a place
on the sclera, the white image in this eye will be pink or yellow. As
-there has been some controversy about the explanation of this experiment,
l this modification of it with the plates of uranium glass,
which enables us to survey all the concomitant conditions more
clearly, may prove to be more convincing.
    Accordingly, the phenomena of binocular contrast admit of easy
explanation from our point of view. But if contrast colours are considered
as being caused by changes of sensation due to the spread of
the stimulation at one place on the retina to the adjacent places, as was
a very prevalent theory at one time, we are obliged to infer that
binocular contrast also is due to some action of the sensations in one
retina on those in the other retina. Accordingly, this was supposed
to be an argument in favour of the innate anatomical connection
between corresponding fibres of the two optic nerves.

    DOVE, who discovered the phenomenon of stereoscopic lustre likewise
offered an explanation of it, which needs to be mentioned. He
draws a distinction between the shining white light reflected from the
outer surface of a body and the coloured light that is radiated from
the superficial layers of the substance. His idea is that the effect of
lustre is caused by our seeing the illuminated substance behind the
illuminated surface; in other words, it is due to light of two kinds, one
shining through the other. When two colours are combined, say, red
in one field and blue in the other, DOVE believes that we imagine that,
they are at different distances, because it is necessary to focus the
eyes differently for each colour. I have not adopted this explanation,
because subsequent experiments on judgment of distance by accommodation,
particularly in a case like this where the convergence of the
two eyes must be kept constant, appear to me to indicate that it is
extremely unlikely that any such apparent difference between the
distance of the colours can be perceived. Another difficulty about this
theory is the fact that lustre can also be produced by combining white
and black. DOVE’S explanation of this effect is that the tendency of
white illumination is to cause a contraction of the pupil, which is
always a concomitant result of higher degrees of accommodation,
whereas the tendency of black is to cause a dilatation of the pupil;
and so he infers that different feelings of accommodation are produced

* FECHNER iiber den seitlichen Fenster- und Kerzenversuch. Berichte der
Kdnigl. S&sischen Ges. d. Wiss. 1861, pp. 27-56.
z See Note 3 at the conclusion of this chapter.-K.

526 oj Vision (428.

by looking at white and black. But in these particular experiments
the peculiarity consists in the fact that one eye is gazing at white
while the other eye is gazing at black, the consequence being that the
two pupils would be of the same average size; and, secondly, any
accommodation here would relate only to the edges of the coloured
areas, and not to the central portions, and it is difficult to see how a
difference in the feeling of the accommodation can arise from the fact
that in one of the images white is on the right and black is on the left
of the border between them, or white is above and black is below this
border, while in the other image it is just the reverse. And so I have
ventured to propose the explanation which I have given above and
which I think is simpler than the one which was originally offered by
the celebrated discoverer of this phenomenon.
    Historical.-Some of the earlier investigators were aware of this
conflict or rivalry between the visual fields of the two eyes. Du TOUR used
it as an argument in favour of his theory that only one eye sees at a time,
and that therefore things are seen single in spite of both eyes being used.
HALDAT insisted, on the other hand, that he had seen binocular mixing of
colours, and he tried to establish a connection between this phenomenon and
the hypothesis of an anatomical union of corresponding fibres of the two
optic nerves, which had been accepted by NEWTON and afterwards by WOLLASTON
and J. M~~LLEIZ.
WALTHEH HALDAT’S view was adopted by M~NNICH, JANIN, and on the other hand,
J. MULLER himself, who had been mainly instrumental in developing the
theory of identical places on the two retinas and the consequences of this
theory, and who certainly would have been primarily interested in the
phenomenon of binocular mixing of colours, never once alludes to it.
All he could see was the binocular conflict between them. The wide
divergence of views on this subject on the part of more recent observers
has been discussed previously. There seem to be great differences between
individuals in this respect. As long as the sensation of a compound colour
was considered as being a simple effect of two concomitant causes,
apparently a sensation of this nature could occur only in one and the same
fibre of the optic nerve: and so the observation of actual binocular
mixture of colours seemed tantamount to proving that there must be an
anatomical fusion of each pair of corresponding nerve fibres. Besides, on
such an hypothesis binocular mixture of colours was a necessary
consequence. It is true that this particular point loses much of its
importance because, as stated above, it comes in conflict with YOUNG'S
colour theory.
    A distinct advance was achieved when DOVE found out the objective
significance of binocular fusion of different luminosities or colours by
discovering the phenomenon of stereoscopic lustre. Although BIIEWSTER
adopted DOVE’S theory of this effect, he seems to have been under some
misapprehension about it because he really argues against it,. But the
simpler theory given above was first suggested afterwards by J. J. &TEL.
Without knowing of his explanation, I myself reached the same view of the
matter, and emphasized the importance of the phenomenon on account of its
bearing on the theory of the sensations at corresponding places on the two
    The phenomena of binocular contrast were not studied until more
recently. FECIINER, especially, has discussed them very fully. Some
previous scattered observations had been made by E. BR~~CKE, H. MEYER, and
527 [...]


Notes on $32 by v. Kries
1. According to SCHENCK' and STIRLING,~ a very satisfactory
method of obtaining binocular mixture of colours (see page 505) consists
in exposing two objects of different colours but of the same rather
complicated form; one being visible to one eye and the other to the
other eye. For instance, postage stamps, which often have the same
design printed in different colours, are particularly adapted for this

1 SCHENCK, Sitzungsber. de7 phys.-med. Gesellsch. zu WtiTzbuTg, 1898.
* STIRLING, An experiment on binocular vision with halfpenny postage
stamps. Journnl of Physiology. XXIII (1901), 27.



purpose. The exact correspondence in the forms of the two images
makes it easier, of course, to fuse them stereoscopically, and at the
same time it aids us in mixing the colours binocularly. As SCHESCK
says, some persons who may be totally unable to perceive a mixed
colour in any other way can succeed in doing it by this method. A long
time ago, after numerous vain attempts, which had made me very
sceptical about the whole phenomenon, I happened to use some
painted coins in a way quite similar to SCHENCK’S method, and did
succeed in getting a binocular mixture of colours which was thoroughly
convincing. In my own case the method works better when the lines
of fixation are crossed. The colour mixture which I obtain in this way
is perfectly uniform and stable. After watching it for some time and
then screening one eye without changing the fixation of the other one,
I am persuaded that the colour which I saw in the mixed image is very
different from that which appears in monocular vision merely as the
result of the long-continued fixation and modulation (Umstimmung)
of the eye.
    Consequently, as far as I am personally concerned, I must admit
the possibility of mixing colours binocularly, although I cannot succeed
in accomplishing it except under very special conditions. I think it
has to to be acknowledged, without any reservations, that there is
such an effect. This being the case, and without disputing such
possibilities of illusion as are mentioned by HELINHOLTZ, it would seem
natural to suppose that some individuals have much less difficulty in
obtaining this effect than others, and that the explanation of the
conflicting statements on the subjects is due primarily to real

2. FECHNER’S paradoxical experiment (page 522) was subsequently
repeated by SCH~~N and ,MOSSO~ and more carefully studied in some
of its details. They discovered a new effect consisting in a periodic
fluctuation of the relations of conflict or mixing, such that a luminous
area viewed with one eye screened appeared to get alternately brighter
and darker.
    By the way, PIPER’S observations, in which he found that the
sensations of brightness in monocular and binocular vision were not

1 BFV. TREICCELENBURG, Versuche iiber binolculare Mischung von
Spektralfarben zft. f. S~WZ&&O~., 48, (1913), 199-210. Also, PFL~~GERS
Arch., 201 (1923), 2x,-249.- S. DAWSON, The experimental study of binocular
colour mixture. &it. J. of Psychol. 8 (1917), 510-551; and The theory of
binocular colour mixture. Ibid., 9 (1917), 1-22.-G. F. ROCHAT, Etude
quantitative du fusionnement binoculaire des couleurs ComplCmentaires
Arch. nderl. de physiol., 7 (1922), 263-267. (J. P. C. S.)
1 SCH~N und MOSSO, Eine Beobachtung iiber den Kettstreit der Sehfelder.
\Archiv f. Ophlh. XX. 2. 1874. p. 289.


connected in the same way in twilight vision and daylight vision,’
ought to be included here also. In the first case (dim light, darkadapted
eye), the brightness was found to be decidedly greater in
binocular vision than it was in monocular vision. Consequently, the
threshold values in the former case are only about half as large as
they are in the latter case.
    As a special form of the rivalry in binocular vision, some reference
should be made here also to the case when an after-image is produced
by illuminating one eye only, the appearance of which is then watched
by closing this eye and opening the other one. BOCCI,~ especially, has
made observations of this kind, which show that under these conditions
the after-image is often visible. In other words, there is a
combination of impressions in which the after-image in the closed eye
is involved on one side, and the sensations mediated by the seeing eye
are involved on the other side. Obviously, we cannot expect these
observations to reveal any information as to the (retinal or cerebral)
seat of the after-images.

    3. This is the place (see page 525) to mention another observation
relating to the connection between the two eyes or their central
counterparts. SHERRINGTON~ tested the frequency of an intermittent
light needed to obtain a steady sensation or to get rid of the flicker;
and in doing so, he varied the phase-relations between the illuminations
of the two eyes in a number of ways. It developed that the essential
question here was how these relations were adjusted for each eye
separately, and that, so far as fusion or time-discrimination was
concerned, each eye apparently operates independently by itself.

4. Supplementary Note.4-The phenomena of binocular colourmixing
have been carefully studied of late, especially by TRENDELENBURG.’
These experiments show beyond doubt that with suitable
objects of observation a real mixture of colours does take place; that
is, there is actually a binocular mixing of colours. As was also found
by the earlier investigators, the mixed colours obtained by binocular
fusion are practically such as would be obtained by mixing the colours

1 PIPER, Zeitschrijt j. Psychologie etc. 31. p. 161; 32. p. 161 (1903).
2 BOCCI, Annali di ottalm. XXV. 1896. p. 445. 8 SHERRINGTON, On binocular
flicker and the correlation of activity of corresponding retinal points.
British journal of psychology. I. 1905. p. 26.
* BPrepared by Professor v. KRIES for insertion in the English Translation
at this place, and communicated to the Editor in January 1924. (J.P.C.S.)
6 W. TRENDELENBTJRG, Versuche tiber binokulare Mischung von Spektralfarben.
Zft. f. Sinnesphysiologie 48. 1913. p. 199.-Idem, Weitere Versuche tiber
binokulare Mischung von Spektralfarben. Archiv j. d. ges. Physiologic.
201. 1923. p. 235.

    Review of the Theories 531

monocularly (that is, by letting the two kinds of light fall on the
same place on the retina). However, it ought to be added that this
agreement is not perfect. There are some differences between the
sensations produced by mixed colours when the observations are made
with both eyes and those obtained when only one eye is used. It is true,
they are not large but still they are very positive and very uniform.
Thus, if a mixture is to look the same way in binocular vision as it
does in monocular vision, the lights must have the same wave-lengths
in both instances (for example, in yellow-purple and white mixtures),
but they must be mixed in different proportions. This result was
found also by ROCHAT . 1 Everything connected with the interaction
of the two eyes is of so much interest that these differences are very
important to consider. However, any attempt at a theoretical explanation
would perhaps be premature at this time.-K.

    $33. Review of the Theories

    Having thus presented the facts that have been ascertained in
regard to the perceptions of vision, we ought now to examine their
theoretical bearings once more, in order to decide between the several
theories that have been proposed and to see which of them are consistent
with these facts, and which are not and therefore less likely
to be true.
    It ought to be said in the beginning that our knowledge of the
relevant phenomena is still too limited to justify us in accepting any
one theory to the exclusion of all the others. This being the case, it
seems to me that in trying to decide between the various theories the
tendency heretofore has been to yield too much to a predilection for
certain metaphysical modes of thought, instead of being guided simply
by the facts themselves; especially as fundamental questions continue
to arise in the realm of psychology which have long since been completely
barred from the domain of the phenomena of inorganic nature.
    In my judgment, many natural philosophers have been far too
ready to presuppose all kinds of anatomical structures in the theory of
the perceptions of vision and also to postulate new qualities of the
nervous substance that are contrary to what we actually know about
the physical and chemical properties of bodies in general and about
the nerves in particular.

1 G. F. ROCHAT, Etude quantitative sur le fusionncmcnt binoculaire des
couleurs compiementaires. Arch. nccrlandnists de physiologic de I’hommc
et des animaux. 7. 1922 p. 263. c


    Consider, for instance, the case of two double images of one and
the same object, both situated on the same side of the median plane.
On HERING’S theory, one of them excites a positive, the other a
negative, sensation of depth, not of any slight amount either, but, as
his theory of stereoscopic phenomena assumes, of considerable and very
clearly perceptible magnitude. But because we know that the double
images belong together and are images of one object, of whose distance
we are more or less well aware, we are supposed ordinarily not to
perceive the difference between their sensations of depth, even if we
try to see whether one looks nearer to us than the other. But now
suppose that a slight difference of colour is produced in the two images,
either by previously fatiguing one eye for one colour or by illuminating
it from the side; then there will be a real difference of sensation between
the two double images. But even when this difference is of the minutest
possible sort and may not be perceptible at all except by the aid
of binocular contrast, it comes out in spite of our knowing distinctly
that the two images are images of the same thing and must therefore be
of the same colour, and notwithstanding that the colouring is no real
colouring, but only a subjective appearance, and that we are aware of
this likewise.

The following text is from

"Helmholtz’s Treatise on Physiological Optics: Volume II"
The Sensations of Vision


288 The Sensations of Vision [244, 245.

    Some other illustrations of our faculty of distinguishing the colours
apart of two objects placed one behind the other will also be added
here. The first one is connected with VOLKMANN’S experiment alluded
to above. He held two small strips of coloured paper in front of his
eye, one quite close and the other at the distance of distinct vision; and
noticed that, instead of seeing the mixed colour, he saw one colour
through the other. Hold a green veil close in front of the eyes, which
is so highly illuminated that the entire field of view has a green tinge,
whereas the pattern and creases in the veil are seen merely as a very
faint blurred image. Then there will be no difficulty in recognizing
correctly the colours of objects seen through the veil, although on the
retina some of the green light of the veil is mixed in with all colours. It
is even more striking still when presently the retina becomes fatigued
for the green light; and then the objects seen through the veil will even
be pink-red, although green light is mixed in their retinal images. The
best way to see this is to close the left eye and look through the green
veil with the other eye. Presently a white paper seen through the veil
will look not simply white but even reddish-white. Then if the right eye
is closed and the uncovered left eye opened, the paper will look green to
this eye by contrast. When the eyes are opened alternately, the paper
looks reddish with the right eye where the retinal image of the paper is
greenish white; and, conversely, it looks greenish with the left eye
where the retinal image is white.
    The same result is obtained in the experiment described by SMITH
of Fochabers (Scotland),1 which was afterwards modified and theoretically
explained by BRUCKE.2 When a bright flame is placed close
by the side of the right eye, or when the eye is illuminated from the right

1 Edinb. JOUTX of Science, V. 52.-POGG. Ann. XXVII. 494.
2 Denksciw. der k. k. Akad. zu Wien. III. Bd.-Poco. Ann. LXXXIV. 418.

245.1 $24. Contrast 289

side by the sun, so that no light goes directly into the pupil, the other
eye meantime being shaded, white objects will look greenish to the
right eye and reddish to the left eye. This is seen distinctly by opening
the two eyes in succession, sometimes the right eye and sometimes the
left eye; or by looking steadily with both eyes at a white sheet of paper
and holding a little black rod vertically midway between the paper and
the eyes. Then two images of the rod will be seen projected on the
paper, one for each eye. The image on the left, where the surface of the
paper is seen by the left eye, but not by the right eye, will look red, and
the other image will look green. On the other hand, when a person
looks steadily at a black plate and holds a white object in front of it
some distance away, so that there are two images of it, the right image,
which now is the one seen by the left eye, will be red, and the left image
will be green. Thus, white looks greener to the eye that is illuminated
from one side than it does to the eye that is not illuminated. Now
under these circumstances, light penetrates through the sclerotica and
eyelids into the illuminated eye, and this light is red, as we already
know from previous experiments (Vol. I, p. 213). If sunlight is allowed
to shine on the eye from one side, the red colour will be recognized on
dark objects too. For example, on looking at a printed page, the black
letters appear a beautiful red and the white paper green. This red
light coming in from the side is diffused over most of the fundus of the
eye, and the places on the retina of the illuminated eye where the image
of a white object is formed are therefore simultaneously illuminated
by white and red light, but the sensation is greenish white. The greenish
colouring gets more and more distinct as the experiment goes on,
because it depends on the eye’s being fatigued for red. But with excessive
red illumination of the retina the only way this can happen
is by the illumination already diffused over the ground getting separated
from the additional light coming from the objects; and thus
this latter light looks greenish because the eye is fatigued for red. In
contrast therewith pure white looks reddish in the eye that has not
been affected.



This digital document copyright 2001 University of Pennsylvania.
Permission granted for all nonprofit educational and nonprofit scientific
uses, provided that
(1) the electronic version is properly cited, and
(2) this page or a reasonable facsimile prominently accompanies any excerpt
(i.e.any collection in one place of one or more files) that contains more
than 1000 total words or more than 4 figures.
All other uses, including any commercial use, requires written permission
from the University of Pennsylvania.
Citation of this work should include the following information (in addition
to information about the original 1925 work from which this edition was

Electronic edition (2001): University of Pennsylvania
URL: http//

Please contact Prof. Benjamin Backus ( at
the University of Pennsylvania if you have any questions about permissible

[Home] [Colour Theory] [Bicovi Demos] ["Whiter" Demos] [Sky Blue Viewing] [Artwork] [FAQ's] [Links]