Acta Ophthalmologica 2014

Historical Article

The entoptic view of the retinal

vessels
Harry H. Mark
Yale-New Haven Hospital, New Haven, CT, USA

ABSTRACT. interested observers – named by Michael

The first time the retinal vessels were seen in man in vivo was reported in 1819 Polanyi ‘mutual authority’ (Polanyi &
by Purkinje as an entoptic view. This was understood to show the shadow of Prosch 1975) and by John Ziman ‘maxi-
the vessels, an interpretation objected to in 1834 by Brewster. Müller in 1855 mal consesuality’ (Ziman 1978).
(Über die entoptische wahrnehmung der netzhautgefässe, insbesondere als be- Early History
weismittel für die lichtperception durch die nach hinten gelegenen netzhautele-
The first to observe the retinal vessels
mente, Stahel, Würzburg) used the phenomenon to deduce the location of the
in vivo was Johann Purkinje (1787 –
photoreceptive layer of the retina, and his conclusion is accepted as true today.
1869), published in 1819 as his medi-
Because the phenomenon has some characteristics of an afterimage, it touches cal doctorate dissertation at the
on the question of what is subjective and what is objective physical reality. It University of Prag (Purkinje 1819)
was recently used clinically to measure potential visual acuity and in the diag- Purkinje reported using five different
noses of diabetic retinopathy and macular degeneration. stimuli to observe his vessels. (i) He
slowly moved a lit candle from the
Key words: afterimage – angioscotometry – entoptic phenomena – H. Müller – photoreception one side of the visual field to the other
Purkinje – retinal vessels – visual neuroscience side. A black image of the retinal vas-
culature appeared (Fig. 1). (ii) The eye
Acta Ophthalmol. 2014: 92: e237–e240 was illuminated from the side while
ª 2013 Acta Ophthalmologica Scandinavica Foundation. Published by John Wiley & Sons Ltd. turned nasally. (iii) By electric (‘gal-
vanic’) light production. (iv) By strong
doi: 10.1111/aos.12192
pressure on the eye. (v) Shutting the
eye after looking at a large white
The beautiful entoptic view of the reti- the nature of the visual process. The plane. Friedman more recently further
nal vessels can be easily elicited in name itself [msóV Greek for ‘inside’ added: ‘They may be seen by rapidly
one’s own eye or another’s. Recent and ópsijóV for ‘light or vision’] was shutting and opening the eye a num-
studies show that details of this image introduced by Johann Benedict List- ber of times in a brightly lighted room
are as good or better than those ing (1808 – 1882) in 1845 (Listing while the vision is directed towards a
obtained by fundus photography or 1845), although is sometimes errone- white wall’. And ‘If the eye looks into
fluorescein angiography (Bradley et al. ously called ‘entopic’ [Greek for a bright light and the light is suddenly
1998). The phenomenon is usually ‘inside place’] (Albert 1966). extinguished’ (Friedman 1942).
brought about by shining a focused The entoptic view of the retinal vessels ‘Purkinje did not say unambigu-
light (such as from a penlight or a cannot be readily verified by an outside ously whether his dark vascular figure
hand-held ophthalmoscope) on the observer, in contrast to the entoptic view was a shadow of the vessels or gener-
sclera, or in front of the pupil, and of vitreous opacities, for instance. It ated in some other manner. Sir David
then rapidly moving it back-and-forth. may therefore be confused with halluci- Brewster in 1832 (Brewster 1834)
When coming from in front, this nations, dreams or optical illusions. vehemently denied that the figure rep-
motion is not always necessary, and Purkinje’s own paper was titled ‘Contri- resented retinal blood-vessels, and
the light may also be replaced by a butions to the knowledge of vision in a proposed a hypothesis …’for an
small opening in a black screen while subjective perspective’. His discovery explanation of the remarkable experi-
looking at a bright white wall. It has conforms, however, to the now generally ment on the eye, first made known by
been historically used to determine the accepted definition of a true physical Dr. Purkinje of Breslau, ‘though I had
photoreceptive layer of the retina and fact as one which appears in only one in vain tried to see this phenomenon’.
raised some intriguing questions about particular way to the largest number of He suggested that the candle’s light

e237
Acta Ophthalmologica 2014
Fig. 1. Image of Purkinje’s own retinal vessels, right eye (21) and left (22).
was composed of ‘direct’ and ‘propa-
gated light’ [reflected from the lens
and cornea] and that the retinal ves-
sels were ‘insensible to propagated
light, and therefore the blood-vessels
must be delineated in obscure lines’.
Sir Charles Wheatstone ‘dissented
from this ingenious explanation’ by
describing his ways of producing the
phenomenon: First, in a dark room a
candle light by the side of the eye is
moved up and down (Wade 1983).
‘The blood-vessels of the retina, with
all their ramifications, are distinctly
seen, apparently projected on a plane
before the eye, and greatly magnified’.
Secondly, the candle is viewed
through a small hole in a glass plate
fixed to a perforated metal plate
which is kept in motion, ‘the blood-
vessels will be seen distributed as
before’. ‘Mr. W. observed, that there
was no difficulty in accounting for the
image; it evidently was a shadow
resulting from the obstruction of light
by the blood-vessels spread over the
retina; the real difficulty was to
explain why this shadow is not always
visible’. He maintained that an object
‘when continuously presented to the
same point of the retina becomes
invisible… but by continually shifting
the place of the image of the object
on the retina, or by making it act
intermittently on the same point, the
object may be rendered permanently
visible’. Helmholtz called this invisibil-
ity ‘tiredness’ of the retina, or in
today’s term ‘adaptation’.
Brewster, however, was not moved
by Wheatstone’s views and repeated
his explanation of the light proceeding
from the candle indirectly. ‘Unless the
blood-vessel is placed at a certain dis-
tance in front of the retina [his empha-
sis], and consequently in the vitreous
humour, it can have no moving sha-
dow’. Apparently, he positioned the
e238
vessels in the retina, whereas Wheat-
stone wrote ‘over the retina’. ‘There
are two facts which appear to be con-
clusive against the explanation
[Wheatstone’s]: the one is, that the
blood-vessels of the retina are not at a
distance from it; and the other, that
the ramifications [the vessels’] may be
seen distinctly when the candle is not
in motion’. The idea that the form of
Purkinje’s figure was that of the reti-
nal blood vessels was confirmed by
comparing it with anatomical prepara-
tions of the retina as described, for
instance, by Soemmering. Its nature
as a shadow of the vessels was simply
noted as fact in Johannes Müller’s
classical textbook on physiology, and
thereafter generally believed (Müller
1838).
Physiological
Application
The entoptic view of the retinal vessels
has many characteristics of an after-
image. It remains visible after the
intermittent short stimulus has
stopped; in this manner, it also has
the nature of palinopsia (Wikipedia/
Palinopsia; accessed May 5, 2013). It
remains visible when only one eye is
stimulated while the other is open. In
contrast to real images, the entoptic
vascular tree does not move when the
globe with its retina is passively
moved, for instance by digital pressure
or pull with forceps (vom Hofe 1941).
Passive motion of the globe normally
moves the perception of the external
world because the final location of the
ordinary real image of the world is in
the cortex, which conceives it as sta-
ble, based on inputs from the extra-
ocular muscles and all other sensory
organs (including the other, stable,
eye). On the other hand, when one
commands to oneself to intentionally
look right or left, after-images, and
the entoptic vasculature, also move in
the same direction (Meissner 1854;
Göthlin 1927). The extent of this
movement is, however, not at all pro-
portional to that of the globe, which
may even remain completely at rest.
The apparent size of the vascular tree,
similar to the size of after-images,
depends on the distance judgment of
the background unto which it is pro-
jected rather than the actual size on
the retinal vasculature (Duke-Elder
1968; Dwyer et al. 1990). And, above
all, microelectric recordings from the
retina (elctroretinograms) show no
evidence of activity relating to after-
images, whereas recordings from the
cortex (Electroencephalogram, visual
evoked potentials) do so indeed
(Grüsser & Grützner 1958; Burr et al.
1987; Fishman et al. 2001).
That retinal vessels may throw a bit
of shadow on the retina (relative sco-
toma) is demonstrated under very
strict testing conditions by angiosco-
tometry with minute object sizes (0.5–
1.5 mm diameter) seen at close range
(19 cm) through a collimating lens.
One of the scotomas’ contrasts to the
entoptic view is their enlargement by
pressure on the globe, pressure on the
other eye or compression of the jugu-
lar veins (Evans 1938, Goldmann
1947; Duke-Elder 1962a,b).
A most thorough investigation of
the phenomenon was carried out by
Heinrich Müller (1820–1864), whose
name is attached to retinal cells and
an orbital muscle, published in a
paper 37 pages long titled in transla-
tion: ‘On the entoptic awareness of
the retinal vessels, especially as evi-
dence for light’s perception by the
retinal elements which lie in the rear’.
With the help of two assistants, Mül-
ler shone light on his globe at loca-
tion a (Fig. 2) and marked the
projection of the visible blood vessel
v’ on a screen in front of the eye (A).
A second area (b) was then illumi-
nated, and this projection marked
(B). On a true diagram of the globe,
about an inch in size, he drew the
lines aa, aA, and bb, bB through the
nodal point. He thus geometrically
deduced location v’ of the blood ves-
sel and was able to calculated the
distance between this point and the
retina (R, where a and b are also
located). This distance was then
transposed on a fixated anatomical

Fig. 2. The candle’s light is assumed to origi-
nate at a, the vessel’s image is projected to A.
Fig. 3. Vessels’ image remains centred no mat-
ter whether light arrives from the right or left;
it moves to the left when the light moved to
the right – in opposite direction – unless light
is assumed to reflect from the retina at b b.
specimen of a globe and measured
from a retinal blood vessel back-
wards. It ended at the cellular layer
of the rods and cones. Therefore,
Müller concluded, this area must be
the photoreceptive one.
Müller’s judicious mode of reason-
ing and conclusion (which Helmholtz
called theory) (Helmholtz 1896; Duke-
Elder 1962a,b) seems to contain,
nonetheless, several untenable presup-
positions forming perhaps insur-
mountable obstacles to accepting their
Fig. 4. Typical retinal vessel scotoma according to Evans (1938).
validity. First, the light source was
strangely presupposed to emanate
from a single point on the retina (a or
b, Fig. 2), not spreading out from its
real origin. Even granted this assump-
tion, the blood vessels on this side of
the globe (right) (i) must cast their
shadows across to the other side (left)
(a), thus being superimposed on the
vessels’ shadows on this side (v), for a
shadow is never on the same side
from whence light arrives. In reality,
the projection of the total vascular
tree (AB) does not appear this way.
Secondly, when the light is moved
from point a to b, the vessel’s projec-
tion always moves in the same direc-
tion from A to B (right to left). The
theory assumes that these distances ab
and AB are directly proportional, and
therefore, the position of point v (the
vessel) may be geometrically calcu-
lated. In reality, although no direct
fixed ratio exists, that is, small move-
ments of the light (a–b) may cause large
movements of the projection (A–B) or
small ones, and large movements may
result in small displacements of the
image, or none at all, as Purkinje him-
self and Brewster already remarked. A
variable ratio cannot simply be submit-
ted to geometrical deductions.
Thirdly, the seen projection of the
vascular tree is always inverted (ab is
seen as AB, the nasally located blind
Acta Ophthalmologica 2014
spot, for instance, is seen laterally).
When the light is shaken in place in
front and to one side of the eye (at
location A or at B), the projected
image of the vessels moves in the same
direction as the light, as noted above
when light falls on the sclera, but sur-
prisingly, when the light originates at
A (aa Fig. 3), the image is located
across the midline at B (dd Fig. 3),
and vice versa, when the light source
is at B, the image moves to A, that is,
it moved in the opposite direction.
When the light source is at aa (left
Fig 3), the image of vessels is to the
right of it, and when it moves to cd,
the image moved to the left of it (Pur-
kinje 1819; Duke-Elder 1968). Follow-
ing Meissner’s hint (Meissner 1854).
Müller procrustean fit his theory to
accommodate this contradiction by
suggesting that the light source does
not directly illuminate the blood ves-
sels (v) but is first reflected off the ret-
ina on the opposite side (bb and cc).
In this way, when the light from aa
projects the vessel to dd, and then
moves to dd, the image of the vessel is
projected to aa in the opposite direc-
tion, as observed (Fig. 3).
Clinical Application
Entoptic perception of the vascular
tree was sometimes used clinically to
e239
Acta Ophthalmologica 2014
determine potential visual acuity,
where nonperception of the vessels
strongly suggested poor acuity (Mu-
rillo-Lopez et al. 2000). The test cor-
rectly identified 91% of eyes with
good macular function, and if no
vascular pattern was seen, it was
probable that the eye had poor mac-
ular function (Talbot et al. 1992).
More than half the untrained patients
with diabetes were able to visualize
entoptically their own parafoveal reti-
nopathy (Applegate et al. 1997).
Some advocated the method for regu-
lar examination of the retina (Eber
1922).
Under strictly controlled artificial
circumstances, it is possible to show
objectively that retinal blood vessels
do in fact prevent some light from
reaching the photoreceptive layer of
the retina, and in this manner made
to cast shadows. The method was
variously termed skiaskotometry
(Greek: Vjia- shadow; VjósoV – dark-
ness) (Fig. 4), scotometry (Evans
1938), angioscotometry (Goldmann
1947). It seems of little clinical use
(Harms & Aulhorne 1959), and Evans
lamented ‘the erroneous impression
that the mapping of retinal-vessel
shadows is advocated for clinical
study’.
References
Albert DM (1966): The history of ophthla-
mology. Cambridge, MA: Blackwell Sci-
ence 84–85, 130, 368
Applegate RA, Bradley A, Van Huven WA
et al. (1997): Entopic evaluation of diabetic
retinopathy. Invest Ophthalmol Vis Sci 38:
783–791.
Bradley A, Zhang H, Applegate RA et al.
(1998): Entoptic image quality of the reti-
nal vasculature. Vision Res 38: 2685–
2696.
e240
Brewster D (1834): Observations on the sup-
posed vision of the blood-vessels of the
eye. Philos Mag 4: 115–120.
Burr DC, Fiorentini A & Morrone MC
(1987): Electrophysiological correlates of
positive and negative afterimages. Vision
Res 27: 201–207.
Duke-Elder S (1962a): Angioscotometry. In:
Duke-Elder S (ed.). System of ophthalmol-
ogy, Vol. 7. St Louis, MO: CV Mosby
415–516.
Duke-Elder S (1962b): System of ophthalmol-
ogy, Vol. 7. St Louis, MO: CV Mosby
454–455.
Duke-Elder S (1968): System of ophthalmol-
ogy, Vol. 4. St Louis, MO: CV Mosby 441,
663, 671–672.
Dwyer J, Ashton R & Broese J (1990): Emmert’s
law in the Ames room. Perception 19: 35–41.
Eber SI (1922): Autoophthalmoscopy, subjec-
tive examination of the retina. Am J Oph-
thalmol 5: 973–974.
Evans JN (1938): Clinical scotometry. Lon-
don: Oxford University Press.
Fishman GA, Birch DG, Holder GE et al.
(2001): Electrophysiologic testing. Am
Acad Ophthalm 2001: 2–10.
Friedman J (1942): Observations on entoptic
phenomena. Arch Ophthalmol 28: 285–312.
Goldmann H (1947): Beitrag zur Angioskoto-
metrie. Ophthalmologica 114: 147–158.
Göthlin GF (1927): Die Bewegungen und die
physiologischen Kosequenzen der Bewe-
gung eines zentralen optischen Nachbildes.
Zentralbl f d ges Ophth 19: 289–290.
Grüsser OJ & Grützner A (1958): Neurophys-
iologische Grundlagen der periodiscchen
Nachbildphasen nach kurzen Lichtblitzen.
Albrecht Von Graefe Arch Klin Exp Oph-
thalmol 160: 65–93.
Harms H, Aulhorne E (1959): Comparative
studies on the value of quantitative perime-
try, skiaskotometry and fusion frequency
for the recognition of beginning visual field
disorders in glaucoma. Doc Ophthalmol
13: 303–356.
Helmholtz H (1896): Handbuch der physio-
logischen optik. Hamburg & Leipzig: Voss
192–202.
vom Hofe K (1941): Untersuchungen über
das Verhalten eines zentrales optischen
Nachbildes bei und nach unwillkürlichen
Bewegungen sowie mechanischen Verlager-
ungen des Auges. Albrecht Von Graefe
Arch Klin Exp Ophthalmol 144: 164–169.
Listing JB (1845): Beitrag zur physiologischen
optic. Göttingen: Vanderhoeck & Ruprecht
35.
Meissner G (1854): Beiträge zur physiologie
des sehorgans. Leipzig: Engelmann 76–
84.
Müller J (1838): Handbuch der physiologie
des menschen. Coblenz: Hölscher 50.
Müller H (1855) Über die entoptische wahr-
nehmung der netzhautgefässe, insbesondere
als beweismittel für die lichtperception dur-
ch die nach hinten gelegenen netzhautele-
mente. Würzburg: Stahel.
Murillo-Lopez F, Maumenee AE & Guyton
DL (2000): Perception of Purkinje vessel
shadows and foveal granular pattern as
measure of potential visual acuity. J Cata-
ract Refract Surg 26: 260–265.
Polanyi M & Prosch H (1975): Meaning. Chi-
cago, IL: University of Chicago 182.
Purkinje J (1819): Beiträge zur kenntniss des
sehens in subjectiver Hinsicht. Prag: J.G
Calve.
Talbot EM, Murdoch JR & Keating D
(1992): The Purkinje vascular entopic test.
Eye 6: 322–325.
Wade NJ (ed.) (1983): Brewster and wheat-
stone on vision. New York, NY: Academic
Press 265–273.
Ziman J (1978): Reliable knowledge. Cam-
bridge, MA: Cambridge University Press 6.
Received on February 18th, 2013.
Accepted on April 28th, 2013.
Correspondence:
Harry H. Mark
Yale-New Haven Hospital
16 Broadway
North Haven
CT 06473
USA
Tel: 203 234 2212
Fax: 203 651 1015
Email: iimd@aol.com