A FURTHER INVESTlfiATiON OF THE 00838.5
FLASH PHENOMENON

Thain éar flu Dogma of Ph; D.
MICHIGAN STATE {EN‘EVERSITY
WEISiam R. Mackavey
1959

SSSSSS

lHlIUHIHHIIIUIUHI’IHHIh”(WWIIf'lHlllJHlll

193 01591 3993

A FURTHER INVESTIGATION OF THE DOUBLE FLASH PHEKOMENON

By
w ILL 1AM R. mommy

A THESIS

Submitted to the School for Advanced Graduate Studies of
Michigan State University of Agriculture and Applied
Science in partial fulfillment of the requirements
for the decree of

DOCTOR OF PHILOSOPHY

Department of Psychology

logo

ABSTRACT

Several invesfigators have reported an elaborate series
of perceptual events in response to an isolated pulse of
light. Under specifiable conditions, two flashes are re-
ported to each photic pulse. The nature of these conditions
has led to the speculation, both here and elsewhere, that
the phenomenon requires an explanation w ich emphasizes the
role of the inplex system of retinal photoreceptors. In an

to clarify the basis of tie double flash, an experi-

\

r'f'

, 4- .1
a t \I all

in!

ment was performed to de‘ervine the effect of (l}pulse
duration and (2)5eparation between successive pulses on the
range of intensities within which the double flash could be
obtained. The target was an opal glass disk subt nding a
visual angle of 16: The disk was illuminated from the rear
by a projector whose beam was periodically interrupted By
an episcotister. The duration and separation of the pulses
was controlled through the episcotister. Quantitative data
from five of the six observers revealed that as the interval
between pulses was reduced, the strongest intensities at
which the double flash could he obtained tended to decline.
As the pulse duration was lenghtened beyond 50 ms., the
'npression of duality becane progressively weaker until at

pulse durations longer than about 150 ms. it was no longer

present.

[—10
H9

Collateral oTServations revealed that the dorble Clash
failed to appear when the tarfieb sas imaged entirely withiL
the fevea, but could he rade to appear with eccentric fix-

lso observed to be marfedly

5‘
D
U)

l?

etion. The double flash

LL (I .'

,igh a filter which

‘V
~ .

J
"S
’3
.1
—l

(1
U)

weakened when the target V.- viewed t-
passed only the deep reds.

1ese results has heen attempted in

)wt

An explanation of t

C

‘titorv influences at the level of the retina.

i; ‘\ { .
V ~ ‘ '_| i ; a...“
APPRO ‘J'r-ED: V-L.-’L~‘_:}AAD_¢ :5- ‘5...) ‘r-.J-‘I-.} . '5‘: .
\

 

x: 1‘ L .n r a. M
DATE: f la, Vi L L: \ .
.v J-

"V ,‘V v-‘f r1 fis .. | 'F\
. I . I ‘V 1' ‘
ACXVtJIsQunlsl;

vv
\

The discuSSiols to which Dr. S. n. Bartley has

g.)

V.

Generously devoted his time have provided very reTreshi“r

I
.'-.' .~ .., .a 'J

4 \- '.- M‘s-L. ' O 5-, , 4 ('..~‘ ‘ ‘3 'h
insuar s and mare always neon re¢.rded as a source hi
personal encouragewcnt.
3.7 .. - 1 , ,. .n: j - ,~° - :- .' .2
For the nawy o LClth cerlJed iron an aSSOCJation

O"

with a first ra e scientist and scholar ghe author is deeply

g r a t e 371.? -- .

TABLE OF CCITEITS

ACI‘Z“O\3’18CI{§OYIQVtS ooooooooooooooooooooooococo
LiSt Of Fifiures oooooooooooooooooooooooooooo

LiSt Of T3.b].cs OOOOOOCOOOOOOOOOOOO...0......

UTRODUCTIOH ...............................
APPKRATLS AID PROCEDURE ....................
RESFLTS ....................................

Introspective hate ial ...............
Eccentric Fixa'ion ...................
Viewing Through Colo ed Material .....

Tn ‘TCQ‘Q ‘\ _,
DJ-‘DCL'UJION 00.0000000000000000000000000....o

SITIJ'E.’\RY OOOOOOOOOOOOOOOOOOOO0.000000000COOOO

APPET'YDIX 0.0.0.0...0.0.0.0....OOOOOOOOOOOOOO

m
l

Raw Data For Observers Showing

F1!

hresholi

..

m

Both lfie Ascerding And Descending Series.

REFERET-CES .QOOOOOOOOOOOOOOOOOOOOOOOO0.0.000...

.0...

For

\n
[\J

7.

LIST OF FIGTRES

TVpr‘er and lower threshold curves for 0—1 at each
of the pulse drret.t.s and separations at which
75110 (3.0717319 {1.51941 76.8.3 £59701”; (3d 0.0000000000000000.

Upper and lower threshold curves for 0—2 at each
of the pulse Eurations aTd separations at wtich
the double flash was “CPFFT. ed 0.0000000000000000.

ppe and l we: thres‘old curves for 0-3 at each
of the p wl e d rations ami scpeim.tiozs at w ich
the 7.07.: -7319 7 gb V3.3 reported 0.0000000000000000.

1

Upper and lower threshold curves for O-h at each
T the pulse durati ions and operations at which
the dOLlIDle flag 11 'WQ 3 Tef‘Oroed ooooooooooooooooooo

Pier and lower t reshold e“rves “or 0—5 a.t each
of ”he “uls e durations and segm irations at wiich
th IG’ (7.797.77338 J..]..C'Sh \I'EES 2.7;0 ”1'th 00000000000000.0000

Upper and lower threshold curves for 0-6 at each
f the pulse dwir tions aTd separations at which
' ‘OTOLC 2138h was refor"ed 00.000000000000000.

Data of Figure 1 presented in an alternative
.. (‘7‘.

-—-L)L--'-0n .0...OOOOOOOOOOOOOOOOOOCOO0.0.0....0.0...

3O

31

Table

l.

2.

Raw Data for Observer 1
Both the Ascending and

Raw Data for Observer 2
Both the Ascending and

Raw Data for Observer 3
Both the Ascending and

Raw Data for Observer h
Both the Ascending and

Raw Data for Observer

T5 LL. .v,‘ . 3.“ ~ -
no»: the ASCGDJIAE aha

.71 \ .‘1

4'

Raw Data for Observer 6
Both the Ascend mg and

OF IAVIES

W?
{D
l 7‘"
(D

Showing Thresholds for
Descehling Series ......... %6

showing Thresholds for
Descending Series 000000000 h?

Showing Thresholds for
Descending Series ......... #8

howisg Thresholds for
DESCCHaifiq Series 000000000 #9

shclfis for

.03 000000000 50

Showihg Thresholfis for
Descendifir Series ......... 51

g)

vii

INTRODUCTION

The neural'response to an isolated photic pulse has
several distinct features. Under specifiable conditions,
the neural response will consist of a duplex discharge at
stimulus onset and a discharge at its termination. This
neural response can be observed at the gross retinal level
using the electroretinogram (1), along the optic nerve '
(1, 9), probably the Optic tract (31), and the visual cor-
tex (9, 23, 26). In one of his earlier papers Bartley (6)
called attention to what he believed to be some perceptual
correlates of certain features of the optic nerve discharge
pattern. Specifically he noted that midrange stimulus
intensities and durations were capable of evoking an extra
or secondary wave in the Optic nerve record. Presenting a
human observer with an isolated pulse having these intensity
and duration characteristics resulted in the perception of
a double flash. The temptation is obvious. Bartley took
the secondary wave to be the neural basis of the additional
flash. The following figure will aid in the visualization
of what has been said.

on OFF SECONDAR'j

{V‘ q WAVE

l Shana-IRS —l_

 

 

Having demonstrated that both the double flash and
secondary wave could be observed only at those photopic
levels at which rods were yet in action and that the extra
flash could not be seen foveally, Bartley concluded, "What-
ever the second flash represents, it is dependent upon,

(a) the activity of both systems, and (b) a peculiar tempo-
ral relation in the behavior of the two systems."(lO) .

It could hardly be claimed that the secondary wave repre-
sented an initial reSponse of either rods or cones in view
of the fact that it could trail the beginning of the ON
reaponse by as much as 250 ms. Psychophysical and physiolog-
ical investigations are in agreement on this point. Psy-
chophysical studies have indicated that although the latency
of the rod system is greater than that of the cone system,
the discrepancy falls far short of 250 ms (26, 28, 3o).
Physiological studies indicate that the total spread of
discharge latencies at the level of the retinal ganglion
cell is only one quarter of the requisite value (1%).

Mention is found of the double flash phenomenon well
into the last century. Indeed, some of the earlier inves-
tigators have reported at least two brightness peaks or
flashes following a momentary photic stimulus. C. A. Young
in 1872 reported that when a room was illuminated by a

momentary spark:

The appearance is precisely as if the object had
been suddenly illuminated by a light, at first bright,
but rapidly fading to extinction, and as if while the
illumination lasted, the observer were winking as fast
as possible...I have ventured to call the phenomenon
recurrent vision.1 (Quoted in 26).

Judd (22) reports that as many as four complete light-
dark cycles have been enumerated in response to brief visual
stimulation, with all light phases except the first his-
torically being regarded as after images. These are as
follows:

(1) The primary image (not an after image).

(2) A very short dark interval (negative phase).

(3) A positive after image (bright); of the same
hue as the primary image (Phase 1) though usually less
bright (Hering after image).

(h) A short dark interval darker than the sur-
roundings, a negative phase. It is, however, longer
than the second phase (Phase 2).

(5) A second positive after image, which appears
most plainly in the hue complementary to that of the
primary after image (Purkinje after image). It is
definitely less bright than Phase 3.

(6) A long, dark interval, a negative image.

(7) A third positive after image of long duration;
it is of the same hue as the primary image, but un-
saturated. It is less bright than Phase (5).

(8) The final phase, a long lingering dark image
negative. This final-phase does not appear immediately
after the disappearance of (7) but only after a
(usually long) blank interval.

Thus there exist 8 separate phases in the image
following a momentary retinal illumination, 7 of which
may be called after image phases. It must not be as-
sumed, however, that these eight phases can always be
observed in every experiment, for such is not the case.
The conditions which make it easy to observe the first
part of this after image series, make it difficult to
observe the last and contrariwise. (22)

 

 

1It has been pointed out to the author that a spark
is itself an oscillatory phenomenon. This may be crucial to
the effect described by Young.

ikzDougall sought to explain these observations by ap-
pealing to the duplicity theory (26). On the basis of cer-
tain controlled observations he concluded that there was a
significant difference in the latencies of the rod and cone
systems. Two of his experiments are most pertinent. In
the first experiment, a radial slit was made in a rotatable
disk and then covered with red gelatin. The intensities
of the outer and inner halves of the slit were indefendently
manipulated and as might be expected, upon rotation the
dimmer half trailed. As the intensity was lowered even
more, the dimmer portion became colorless and the lag be-
came more abrupt. The temporal magnitude of the lag jumped
from about eight to forty-—eight ms. In a second experi-
ment, two slits were displaced from each other by 10° in
the stationary position. The intensity of the leading slit,
objectively green, was reduced until it appeared colorless.
The trailing slit, red, was undimmed. The observer's task
was to find the rate of rotation at which the "colorless"
green and "colored" red slits appeared to fall on the same
radius. The latency difference was then computed.

The amount of this delay seems to be about the
same for different individuals for some dozen persons,
to whom I have demonstrated this experiment, see the
two images fall into line at the same rate of revolu-
tion of the disc. We may conclude then, that the rate

of development of the cone sensations exceeds that of
the rod sensations by about 55 ms. or 1/18 second. (26)

With regard to the after image sequence he is then led
to conclude, "The initial pulses of such a series are pre-
dominantly due to processes initiated in the cones of the
retina- while the terminal pulses are processes initiated
in the rods...” (25). It is unlikely that the latency
discrepancy can be attributed to the different intensity
levels of the two slits in view of McDougall's first ex-
periment discussed above. Alpern has also shown that the
maximal latency variation within the foveal cone system
due to intensity differences runs to only about 25'ms. (h).

Bartley believes that the extra flash which he observed
using pulses of midrange intensity and duration corresponds
with the Hering after image or Phase 3 of the after image
sequence. Apparently the conditions under which his ob-
.servations were made did not permit the emergence of the
other positive phases. Linking this flash with the secondary
wave-of the Optic nerve record is not the only alternative
available. Recording from cortical surface electrodes,
Lennox—and Madsen (23, 25) give recent evidence that the
duplex 0N discharge behaves in a manner very like that

reported for the secondary wave of the Optic nerve.2 Not

2The reader should not be led to assume that Lennox
and Madsen make any attempt to draw parallels between their
own work and Bartley's work on the secondary wave, nor
between their own work and any aspect of the after image_
sequence. The responsibility for this analysis rests en-
tirely with the present author.

all of this evidence is new (8), but a greater range of
conditions capable of delivering a greater amount of infor-
mation has been employed. Lennox andlhrkmmi show that at
the intensity extremes there is but a single cortical ON
response whereas in the midrange of intensities a duplex
response-is obtained. The shorter latency of the first
member of the duplex ON response, which can lead the second
component by as much as #5 ms., is consistent with the sup-
position that its origin is in photoreceptors of rather high
threshold and short latency, viz., the cones. Their further
demonstration that the duplex ON discharge cannot be observed
in those regions of the cortex receiving projections from the
retinal periphery, where the rods are numerically dominant,
lends additional support to this conjecture. The single ON
response which is recorded possesses an amplitude peak

which is temporally commensurate with that of the second ON
response in other regions of the visual cortex. The course
of emergence of the additional flash should be helpful in
deciding the merits of this explanation of the double flash.
If the two flashes were attributed to direct cone and rod

ON discharges, a la Lennox and Madsen, the first flash
should be less bright than its companion at the lower limit
of their joint intensity range. Unless all cone elements
have identical thresholds, only a small subset would be con-

tributing to the experience at this point. If to the con-

trary, the second member of the pair were always less bright
than its mate the liklihood of this explanation would per-
force be reduced.

Other alternatives are not wanting. There is ample
evidence indicating latency differences in the peripheral
message which is fed to the visual cortex. This evidence
is not limited to rods vs. cones. Cone latencies appear
to vary in accordance with their absorption spectrum:(5)
and rods are probably dichotomized in their latency reactions
into "ideal" and "cone-like" rods (18). Any combination
of these could serve as the basis for one or more phases
of the after image train. An assertion of this type could
be considered a modern recasting of KcDougall's explanation
given above. A danger exists however. The latency differ-
ences involved here are sufficiently small to cause skep-
ticism regarding their capabilities for furnishing a basis
for discrimination.

Thus Bartley‘s demonstration that the additional flash
can be observed only at those intensities at which both
rods and cones are Contributing to the visual experience
poses a problem. How do these receptor types operate to
produce the effect? Perhaps as he suggests, the extra
flash is produced by elements of both systems converging
upon common ganglion cells. To account for the dim interval

between the first and second flashes it would then be neces-

sary to assume that the synaptic delays in the retinal inter-
nuncial network are at least as great as the duration of the
dim interval. Although this might be the case, it seems
more likely that the ganglion cells would be discharging
continuously. Cgteris pagibug, the latency of the discharge
would be dependent upon the number of synaptic delays en-
countered. t is difficult to see why these should be or-
ganized in such a way that :12 discharge groupings result.
The magnitude of the average latency difference between
rods and cones reported by various investigators does make
it reasonable to suppose that the rod and cone discharges
could have a separation in experience. This would be es-
pecially so for very short photic presentations, i.e., those
whose duration is less than the average latency difference
between the receptor populations. The after discharge from
the cones would then be at a low'value when the rods began
their discharge. Perhaps the dual receptor involvement is
somewhat different than has been supposed. The necessity of
rod involvement need not be to produce a direct discharge
down the optic nerve (Lennox and Madsen) nor to converge
upon a number of the same ganglion cells as the cones
(Bartley) but rather to inhibit the ggng'discharg . The
rods upon discharging would briefly but markedly reduce the
ganglion cell discharge originating in the cones- hence the

dim interval between flashes. One way to put it would be to

say that there is "really" but a single flash which is
briefly interrupted at the moment the rod ON discharge
reaches the synapses across which the cone discharges are
flowing. Inhibitory influences of this variety have re-
ceived frequent mention in both the psychological and neuro-
physiological literature. Simultaneous and successive con-
trast phenomena have often led investigators to entertain
the possibility of inhibitory influences at the retinal
level C3, 16, 17, 27 p. 233). Hartline and his colleagues
have repeatedly demonstrated inhibition among ommatidia in
the lateral eye of Limulus:

The inhibitory influence exerted upon an ommatid-
ium that is discharging impulses at a steady rate be-
gins shortly after the onset of illumination of neigh-
boring ommatidia, with a sudden deep minimum in the
frequency of discharge....The mediation of the inhi-
bitory influences appears to depend upon the integrity
of nervous connections in the plexus, cutting the
lateral connections to an ommatidium abolishes the
inhibition exerted upon it. (20)

When only two ommatidia are illuminated, the
magnitude of the inhibition...depends only on the
degree of activity of the other... (21)

Pursuing an explanation of the double flash in terms
of inhibitory effects, it would be expected that pulse
durations could be made "too short" or "too long" to yield
a strong impression of a doubled flash rate. Pulse dura-
tions too brief to allow the cone after discharge to con-
tinue for a sufficient time beyond the rod inhibition would

be "too short". Pulses prolonged beyond the time at which

10

the rod and cone discharges met at a common synapse would
be "too long". Presumably the rod interruption would be-
come less effective as the cone discharge rate is accelera-
ted.

At the lowest intensities, only the rods are active-
hence a single flash. At somewhat higher intensities both
populations are discharging, leading to the experience of
a double flash. Why only a single flash emerges at upper
intensity levels requires a knowledge of rod behavior at
these intensity levels, and the necessary information is
not available. It might be that (l) the rods cease to
function at these levels or (2) the rod response becomes
submerged within that of the cones and cannot be discrimina-
ted from it. The mode of action of visual purple on the rod
cell and its role in light and dark adaptation is no longer
the straightforward matter supposed a few years ago. Many'
of the problems which have arisen in the attempt to relate
adaptation phenomena to the course of visual purple degener-
ation and regeneration are neatly summarized by Granit in
his recent book, Rggeptors and Sensory Perception (19,
pp. lh3-150). Suffice it to say that some workers have
found it profitable to entertain neural as well as photo-
chemical explanations.

It is important to realize that this line of reasoning

is entirely consistent with the view that the secondary

11

wave is the basis of the extra flash. The problem is to
relate this wave and the extra flash to retinal happenings.
The reader will have noticed by this time a strong
tendency on the part Of the author to consider as possible
explanations of the after image sequence, especially the
initial phases thereof, those which emphasize the role Of
the retinal photoreceptors. In part, this stems from two-
of Bartley's demonstrations. He has shown that the addi-
tional flash could only be observed at those intensity'
levels at which both the rods and cones were contributing
to the visual experience. Secondly, it could not be Ob-
served when the retinal projection Of the target was limi-
ted to the fovea. This tendency also stems from the demon-
strated complexity of the retinal neuroplexus.(29). The
neuroretina possesses an internuncial network unmatched in
complexity outside the central nervous system. Given such
an intricate synaptic network, the possible variations in
the peripheral frequency code approach infinity. It is
worth considering then, that at least some phases of the
after image train described by-Judd have their origin in
temporally staggered discharge groupings along the Optic
nerve. The requirements which such groupings must meet,
and how they must be acted upon by the higher centers to al-
low them to be experienced in just this way is not at all

clear. Discharge groupings which have greater prominence

12

than the secondary wave hay; bggn demonstrated, yet there
is little evidence that they contribute to the visual axe
perience in a way similar to that proposed for the secondary
wave. Specific reference can be made to the OFF discharge.
Whatever the origin, there is no doubt that discharges
occur along the axons Of certain retinal ganglion cells
when stimulation is terminated. A limited view of the mat-
ter suggests that this discharge grouping would be inter-
preted as a brightness experience, e.g., a flash, a bright-
ness peak, or the extension in time of an ongoing visual
experience. Although something like this might be demon-
strated under suitable laboratory conditions, it seems more
reasonable to suppose that this discharge acts to signal
the termination of stimulation more effectively'than might
otherwise be the case. The effective contrast between the
presence and absence of stimulation would thereby be en-
hanced. Kept active by physiological nystagmus, combined
ON and OFF discharges would be important in the formation
of sharp subjective contours (27, pp. 229-230).

It should be clear that what is being done here is to
take a well documented experiential phenomenon, the double
flash.reported by Bartley, and make a post hoc entry into
'Uae neurophysiological literature in an attempt to uncover
relevant evidence. The pitfalls of this approach are many.

3h1 ordering the data it is generally necessary to review

13

results Obtained from separate species, using different
techniques, at diverse levels of the visual system, and with
distinct sets of variables under manipulation and control.
In the interpretation of studies utilizing microelectrode
isolation Of individual retinal or optic nerve elements,

the information yield pertains to that element alone. In
the absence of sampling data indicating its representative-
ness, errors of extrapolation must be guarded against. In
single unit analysis, the sensory physiologist is likely to
focus his attention on those units which are most responsive
in terms Of threshold, latency, and discharge behavior.

This tends to preclude a selection Of elements whose dis-
charge is initiated via the lateral internuncial network

of the retina. In this vein, Donner's finding reported
earlier (1H), regarding the spread of latencies at the level
of the ganglion cell may be too conservative- an artifact

of selection.

In interpreting the electroretinogram (ERG), it must
be remembered that the EEG is a mass response which averages
retinal happenings. Because of this, the cone response of
a rod dominated eye, such as man's, may be entirely masked
in the record. Early ERG flicker-fusion studies obtained
fusion at frequencies in the neighborhood of 25 cycles per
second. Subjectively determined fusion thresholds could

exceed this value more than twice over. Using stimulus

1h

intensities of extreme strength, thereby inhibiting the
rod response, investigators have recently been able to drive
ERG fusion thresholds up to values obtained by subjective
methods (13).
Just how the double flash would behave if the pulses
‘were presented serially Offers Opportunity for interesting
speculation. The rate of presentation could be varied
from a considerable separation of pulses to subjective
fusion. As the rate Of presentation is increased, succes-
sive double flash pairs would have progressively shorter
separations. Similarly, the neural activity associated
‘with the successive pulses would have progressively shorter
separations. ‘What would be the perceptual outcome? In
traditional flicker-fusion studies, the separation between
successive flashes becomes gradually less distinct until
the observer is unable to distinguish the intermittent
:stimulation from stimulation which is continuous. In the
IJresent instance it might be supposed that the second mem—
ber of the double flash pair would fuse with the leading
Inenhber of the succeeding double flash pair. Pilot Observa-
13i43ns in the laboratory have shown that this is not at all
thecase. Although by shortening the pulse separations
"tile; second flash can be made less distinct, it is still set
apart from the subsequent flash by a p_I;ominent dag}; interval.

131143 result can be obtained at very low "flicker" rates, e.g.,

15

two pulses per second. Thus whatever the nature of the
neural activity responsible for the second flash, it appears
to be curtailed by the activity aroused by the subsequent
pulse.

Certain features in the behavior of isolated ganglion
cell elements, as revealed by microelectrode techniques,
vprovide the basis for a tempting analogy. Contrary to
naive expectations, the axonal discharge of a ganglion cell

does not attain a steady rate as fusion is approached. At

 

the shortest inter pulse intervals, the element a§,a rule
becomes silent, i.e., it ceases to discharge (12, 15). Here,
one might reasonably expect the absence of a brightness exe
perience rather than fusion. Those elements which do not
‘become silent above "fusion" maintain a slow asynchronous
(iischarge. Apparently when taken as an aggregate these ele-
znents maintain a sufficiently steady discharge rate to allow
tune experience of brightness continuity. Presumably, a
.ffilller understanding of the EQQEé operandi of these discharge
IDertterns will aid in the elucidation of some traditional
IIINDblems encountered in traditional psychophysical flicker
and fusion studies (7). It seems unlikely however, that
the curtailment of activity mentioned with regard to the
double flash at low flicker rates is of‘ this variety. The
Second flash undergoes a brightness decay while still "well

removed" from the following double flash pair and therefore

16

probably too removed from the type of inhibitory influences
suggested by these studies.

The reduction in the prominence of the second flash
can be-accounted for in terms of some inhibitory'action
which is able to produce a perceptual effect across an
interval of approximately 200 ms. Although this interval is
quite-long, it would be even longer were the secondary'
wave not taken as the basis of the second flash. The length
of the interval is obtained by subtracting the latency of
the secondary wave from the time elapsing between pulse
onsets. The rod ON response is probably too removed from
the discharge initiated by the trailing pulse to be influ-
enced by it.

The partial or complete inhibition of a potential
visual experience by another following close behind is well
documented. The effect, almost always limited to a giggle
stimulug pair, has been reported under a wide range of con-
ditions including:

(1) the adjacent presentation of targets (3, 16).
(2) when the second target forms a frame for the

first ( 2).

(3 when the first target falls entirely within

the second (11).

(h) when the second target falls entirely within

the firSt (21").

(5) and even when the second member of the pair
possesses a luminance value less than that of the

first (2%).

For present purposes it is important to note that certain of

these inhibitory effects have been demonstrated at target

l7

intervals in excess of 200 ms. Presumably the separation
between the secondary wave and the trailing activity falls
within this interval. Indeed, an exposure asynchrony of
100 to 150 ms often maximizes the effect (2, 32).

The striking feature of the present observation is that
the brightness decay of the second flash takes place across
a series of pulses. A little reflection will indicate an
apparent paradox. Each double flash pair is inhibited,
yet retains the ability to exert an inhibitory action on
the preceding double flash pair. One claim might be that:
the conditions which allow the emergence of the second
flash are the same conditions which permit the inhibition
1 to be effective across a series of pulses. The activity
responsible for the first flash is the inhibiting agent
and it is the second flash which_is inhibited.

In a somewhat different context, Alpern (3) has also
commented on the paradox which results frOm positing tem-
porally extended inhibitory effects;

....Frbhlich noted that a slit of light moving on

a track was not seen immediately after it emerged from

behind a screen but only after passing through a cer-

tain distance. Rubin showed that by placing a narrow
screen in the path of the slit it could be made to ap-
pear in a region where formerly it had been invisible.
Piéran pointed out that these results were ex-
plicable on the basis of the theory that the slit was
prevented from being seen because of the inhibition of
itself that it produced in a subsequent position along
its track....by placing a narrow screen along the path
...inhibition was removed, and the slit now appeared

in a position where formerly it had been invisible.
This explanation, however, raises the question

18

as to how the slit eve; b om visible at all....one

would expect that, if Pieron's explanation were cor-

rect, the farther the slit moved along its track, the
less chance would it hgve 9; becomming visible.(3,
emphasis added

The present study has two primary aims:

(1) To clarify the basis of the double’flash.

(2) To determine the course of the-perceptual
obliteration of the second flash as the separation
between successive pulses is shortened.

The absence of the double flash when the target is
imaged entirely within the fovea has been taken as evidence
that both rods and cones are involved in the experience.

A control is needed for this observation, viz., a control
for area gu§_area. It is possible that the double flash
was not seen foveally simply because the visual subtense
was too small. One of the observations to be reported will
indicate whether or not a target subtending a visual angle
of 1.60 when presented to the retinal periphery allows the
experience of a double flash.

The effect on the double flash of viewing the target
through filters of various colors, including one which
passes only the longest wavelengths, will also be determined.
This will serve a twofold purpose:

(1) To more accurately match the additional flash
with one of the phases of the after image sequence
described by Judd.

(2) To obtain additional information regarding the
necessity of rod involvement in the double flash. The

double flash should not emerge when the target is
viewed through a filter passing only the deep reds.

19

The roles of pulse duration and intensity will be ex-
amined with regard to their effect on the double flash as
the separation between pulse onsets is reduced. Whatever
the discharge basis of the successive double flash pairs,
any experimental manipulation which acts to maintain their
discreteness should partially compensate for a shortened
ON-ON pulse interval. High stimulus intensities would be.
effective in reducing the latency and prolonging the dura- .
tion of the neural discharges. As the interval between suc-
cessive pulses is shortened, it would then be expected that
the second flash would be weakened at upper (midrange)
intensity levels. The neural activity associated with ad—
jacent double flash pairs would be brought nearer in time.
This same conclusion would follow if it could be shown that
the first flash were strengthened more than the second flash
by the increase in the intensity of the target, i.e., the
inhibiting agent would be made more effective. Among other
considerations, this requires that the discharge basis of
the second flash decay to the same asymptote at different
intensity levels. '

Although the neural separation between double flash
pairs can also be decreased by extending the pulse duration,
it is likely that this would obscure the necessary separa-
tion of activity within a given pulse rather than between
pulses. The double flash would then be eliminated for

20

reasons other than those involved in decreasing the separa-

tion between pulse onsets (see pages 9-10).

APPARATUS AKD PROCEDURE

Observers were obtained from among the students in a
perception course where a portion of the laboratory require-
ment was to be satisfied by participating in an experiment
of direct relevance. Fivo observers (0's) were obtained in
this manner, the author serving as a sixth. The nature of
the required observation was explained to them and they
were also referred to a section in their text, Bartley's
Principles gf Perception, where the phenomenon was briefly
elaborated. Usually it was only after the phenomenon had
been demonstrated by the experimenter that the 0 had any
clear i”ea of what was expected of him.

After a 15-20 minute period of dark adaptation, the O
was made secure in the observational position by attaching
a chin rest to his chair. Attention was then directed to
a six inch opal glass disL located at a distance of two
feet and subtending a visual angle of 15? The only light
source present was that of a Balopticon Projector illumi-
nating the disk from the rear. Fixation was monocular and
aided by a small dot of luminous paint at the center of the
disk. Stray radiation from the projector was materially
reduced by directing the beam into a box at one end of
which the Opal glass was mounted. The projected beam was

interrupted by an episcotister whose Open sector could be

22

varied from 0-1351 The rate of rotation was controlled by
adjusting the drive shaft speed of a motor to which the
episcotister shaft was attached by a leather pulley. Each
of these manipulations was made by the experimenter.. The
O thus found himself confronted with a disk which was alter-
nately illuminated and darkened. He was required to ma-
nipulate a Variac controlling the intensity of the projected
beam according to the method of limits, reporting the range
of intensities within which the double flash could be seen.
A region of certainty was demanded, i.e., every pulse had
to be accompanied by a double flash between the upper and_
lower limits. A shielded red lamp was turned on momentar-
ily to enable the experimenter to observe and record the
setting of the Variac.

Pulse durations of 5, 15, 25, 50, 75, 100, 150, and
_200 ms. were employed, the onsets of which were separated
by 1000, 900, 800, 700, 600, 500, and #00 ms. A.maximum
of 56 conditions were possible, each condition being repre-
sented by h readings, viz., the ascending and descending
limits. Even for a well practiced observer, the number of
readings necessary to tap the limits of the effect required
about 6 hours in the laboratory. The practice sessions
revealed that although an O was quite consistent at any
given session, his day to day fluctuations were such as to

make the interpretation of a completely randomized order of

23

presentation an unnecessarily difficult task. Because of
this the data were collected in blocks and the order of
presentation within blocks randomized. Thus for any given
observational session two or three adjacent pulse durations
were selected, e.g., SO, 75, 100 ms and all pulse sepa-
rations explored. The final data were collected in three
separate one and one half to two hour sessions for each 0.
The total time spent by an O in the laboratory ranged from
ten to fifteen hours of which approximately half was spent
in practice.

It became apparent that if a pulse duration was too
long or the pulse separation too short to allow the e-
mergence of a double flash, any more extreme condition
would have the same consequence. Because of this the total
number of conditions presented was some number less than
56.

The 0's were encouraged to report and were duestioned
on auxilliary features of the experience such as apparent
movement, relative intensity of the flashes, color differ-

ences, and the region of the disk occupied by the flashes.

RESULTS

The data from the individual O's are given as Figures
1-6. Tb eliminate overlap in the plotting of the curves,
a separate pair is given for each of the pulse durations
at which an 0 reported the double flash. The settings of the
variac, the dependent variable, have been converted to a
suitable photometric unit of luminance, candles/footz, and
the ordinates are labelled in terms of this. The intervals
between successive pulse onsets appear on the abscissa.
The upper curves represent the upper thresholds and have
been obtained by averaging the upper limits on the ascending
and descending series. The lower curves represent the lower
thresholds and have been obtained in a like manner. It
should be noted that the data of Figure l are those of the
author and also that the data were collected at a single
prOlonged observational session. Separate data for-the ,
ascending and descending series are given in the Appendix.

The upper threshold data of 0-6 are readily seen to be ,
unlike those of the other O's. To permit a clearer presenta-
tion and interpretation of the results, the upper threshold
data of 0's 1-5 will be treated apart from those of O-d.
An inspection of Figures 1-5 reveals the following:

(1) As the pulse duration is lengthened, the
intensity levels at which the double flash is obtained

decreases, e.g., the upper and lower thresholds are
lower for a pulse duration of 100 ms. than for a pulse

 

 

 

 

 

 

 

 

 

 

 

 

2W ’0[ .l
. P .Ir— -
57n5. 9
1-7 ,2v- A
/‘6 _7~ /
I-5 /- .6' /
Id] 5.
.9 '5’"
8 \———J/ '3. .
7 _L l .21 I J I V ' '
g .3 6° -0 "‘ °‘ w \
as o s s s :2 é‘é‘ 3 E 2%" g s
to
750%
/5V?75. _9
,7
L3 :7
I .
n1 .// .5
Iv “” .4
.6 3
at. A . i i i . 4i. \—
J: U: a~ q o. Q ‘ .t q 6‘ xi as & x
2 z 8 2. a 8 "3, :3 g s s :2 s g
‘3' 25/975 /00m5-
1.2.
l-l' ,9,
ba[ '9‘
.9 ~7~
07F '6‘
t
,7L 'fr ‘/
.6 .yL
A
garages; .ldhaé§s\
° ° ° 0 ° ° 0 3 6 Q g ,2 §
r /5c~'5

1., [ c/{tzx/ooJ' °°°"

 

a? s e

INfé’rt/A/ Br/wefd
Pee/5e Owed: '0':

Figure 1. Upper and lower threshold curves for 0-1

at each of the pulse durations and separations at which
the double flesh was reported.

D\Nw~&"fi0‘\l
\ :

00hr-
005-
00"
UOIT

.5UIH5

 

 

 

 

 

 

 

 

 

 

 

 

5/775 .
I?» 26 /,?p
/'.";’> ”9’
/-7 /-7-'
/-6» A‘-
/|5— /.5_
I'Yk Nil
/-3’ "
/.Z- /'2’
Iv/’ lo,“
’00. /'0~
7' '9P /
34 1 Ox I L 1 _J .SL A o 1 “1° 4 j
t “I \J ‘0 Q \ ‘fi W \l \D ‘
88 8 3 8 2 § 22 3 s S a 3
K' .
11'
/5n75_ 7v .5
n:-
l~7~ l‘3[
W6» \_/ 1.2 /————
/-5> //" /
l-V- [6"
/-3' 3*
['2’ gr-
lol” ‘7
/.o~ \/ .6~
.7” if \
,8 A n A n I _1_ ‘b/ A a l _‘
*2 0‘ 6‘ ‘1 on '0 3 "i 0‘ " \l °‘ ‘§ ‘-~
8 2 2 2: 8 '3 8 8 :3 8 3 8 «>- g
25ms. wows.
1.9- L2?
N 13* ‘ “I 14
g I-7' 3149*
\ \ .
5: I." x .9
~\ ’5 N P
. ’ \Lj .
\Q“ ’04) i 07!
K
K: [.3h :1 _6r
?"2P I a 5’)
\1 "IF \0 ‘4‘ \V/\
I-O‘ A 1 I 1 A ‘1 .3 A ‘; ‘1, g; 4; _‘L
§ :9 a: 2 2° 2? g e :2 g g 2 3“ s:
O 0 ¢ . o c O o o D
.2 MCK’VA/ [30/14/534] IN {(k'VA/ BF/Wf'f-‘J

Pu/sc 0N5€{5-m5. fill/56 05/56/5-5105'

Figure 2. Upper and lower threshold curves for 0-2
at each of the pulse durations and separations at which
the double flash was reported.

27 7511.15

 

 

 

 

 

 

 

3.5- If/V‘s /}'r

3.3 /\// ‘ x [

3.2 x’

/-6 3i"

/5 \ ,3;

/ “A

I V /.0r W

, L i L L 1 . #4 «9. 1 3’ A J. if A

/Z'§§§§§°§§ §§§§§§§

30 r 2 firms. /.é~ ”0””

79" /j"

29- F—_///—-‘““‘~ ’$rw/,,y/’/”\\\\J///

2.7' / ,3. .

I“; 1.2’

“/5 ,,/.\\

I-E r \-1/——\ 1.0 ~ \\_l_\‘

"fitted; 5”? ‘iicfieseg g
g; 5’ Z 2 3 g E g g C) 3 9 o g

30 _ 50ms 0’0!st

29 \\\\

28" ,/

2.7" /_5r

2.6? /.l/'

 

/-0’

g
Y

 

 

 

 

NZ ’
M-
OW .
MIL
(’05 t
009 ~
00z~
60X .
dog’p
000/ .

00/1'

 

 

N 200 1175
m9 [.y.
N" "3 ’
\a
% Ll ’ /
K. /./ ' _../
Vs ?5\»__4/
\7 /'0 ’ \ /"
.q -h—-~——- * * ‘3 “i
(v ~q
§ e 9 § 3 g

I” {(5 v :3/ 6:0{Weé’f
2§bc Ciw4s-Kws
Figure 3. Upper and lower threshold curves for 0-3

at each of the pulse durations and separations at which
the double flash was reported.

 

 

 

 

 

 

75hfi4
1" Jins, 20-
&5t 28 [9,
39' /3»
2.3? L7»
2.2 [.‘L
24* 1-5,
20} my.
['9’ [,3 ..
[I ' 1.), .
I'IL . /.I y-
"! [.0
/..5’ [
.9
2? 5 8 8 g g '7
:.9. 15M. ‘ i
I //-a//f——““‘t\- .4r
1. .5.
'4 l l 1 4 L i ‘L
" ‘§ ‘é’ S 5: ‘8' a a
.9 '5
1 f/ [-90 I ”OHS.
.7 (.9 .
.6 -: Afi—A I71»
8 §‘ § § § x6»
I-5 P
”7 Zflmg L¢P
l6 3
£5 ’ '
/.2.*
I
I- r/ ML»
[0'
.7
" /____"’\ '9 r
.57 3 C ' t o;— ‘ t '3 '
6‘ \
§ § 5? § 8 '§ § .7»
finhs "r
I" '5 c (8 95- \3 3 §
[.7 8 o 8 3 8 o
I" /'/ '- /5'ah,$.
[-5 Ao-
'? W ,4 -
Q
"X "1 N8 18'.
S LI" ~§ .7!
3‘ /'0 ' Q: 1‘ k
35.? V b “5}
V I? ’ Kl ,¢ P
h
‘7 P \01 3 r-
.‘ J 1 . ,2 r-
§§ S § g g g 'lgén 5 B“ é T 1
Inflmwfl Bufluwv 8:3 ° ° 8
Fiéazlgaé-n§. -1%£MMu/dkfimm~
«4.6%afiynmu,
Figure h. Upper and lower threshold curves for O—h

at each of the pulse durations
the double flesh was reported.

 

 

 

 

 

 

and separations at which

.7

5 ms. 29

 

 

 

 

 

 

 

 

 

 

 

 

[cafe/{(3 mo

 

 

 

[T q, 50/125.
Lg' -7
L7. 7f
Lé' 6’
/.5i— .5.
Mt ”'
/._{> .3’
/-Z’ .ZL' ._._—/\
I-I ’ / ,l 3 £3 3:; g 3: 2353 JE
Q U o O 0
L0?
.4 r ———-/-\, q . 75/275.
8 A . . 1 g ’1 .
4; N ‘° ‘
:33 35 § 2 :3 3 .3“ ’//
Q o 7'
15- /503‘ 4_
t4- 5
1.2
/ .5.
1.,-
/ .1-
1-0' -———
qr I,
. 0‘ . A A ‘ L 9..
8* é 2: if E E? E? 8
L
.7 ' ’COVMS
.6; . 8
.5’ \\/—’* 7
.4 H /\\\‘
«3 f A A L ‘1 j
g 0c g 3' g g 8 q» /
O 2 ° c O o Cb
‘ 3
25/125 2F
/.‘/. /
l-3P 4 L
0 A L A 9° -
1.2- / sf ‘é‘ E 3 8 3 §
L! ' 9 /50H75
1 [-0L '
.7'
.9“ if; 6
.5." g; ,5.
'6!’ b‘
J \, .3'
.'r V/‘ g '2’
3 L: a: :5 a A _: \’ /
"7 . P
£3 2 2 e 8 o A A T A A
’ ' 5n o 00 lb \
I/v/(l‘JI1/fl‘/"”ee” 2? 0 3’ § ‘3 ° §
Wu/se Cunt/5m»: Jufezu/ Ke/weo/

2( ./5c Cns6{$- M5

Figure 5. Upper and lower threshold curves for 0-5
at each of the pulse durations and separations at which
the double flash was reported.

 

 

/'i~ 75,775

/ fr
2.2- 5/76 a,

\ 30 r

ZJ~ , A}
zoP/ IV}
“I? /\ /jr
’8} / Ir}
17+ /\/ //
/.6‘ ' [,0 L
1.;

 

 

 

 

 

 

L s» so < u: s" : s ~=‘ —‘
S 3 3 § :3 c; .1? d 3 it
’5m5 /-7 /[[7 m5
56H
x: "
z‘ /-"1
\“\ K3
\ /2+ /
l-l
’lf
4
b . . L w.
(a
.2.
2.3": I -

 

 

 

 

 

 

 

 

. / \ (I,
6* o]? i /' “ //A/
\ '1
5L\/\ 6L \ I
.v -~ii- 3 \V
3 4 4 .‘ ‘
é a S ? l? e g 2 \\___////
o 0 0 Cr 9
,/
l7 \fflw'hS o A A g A 1 4L
1. Ur 0‘ N M .0 r
'5 \ g ‘3, 8 8 8 g 5;
N
8 5 ,,,, 200mm
X 7 " \
"' I3 "/ r
(2' ., ,9
\ f. N 8T
\l " Q I
.’ \\ _q
\I ‘1 \’,/' \x.\ \\\ ‘.
3 r 1 . A L A ._~i- w' “\
-L «A 6‘ d t? 5“ :5- \J
8 C’ 3 ‘v *> 3 3 5y2' \\N‘
1U (chi/4/ l'. 7"? UK!” N .I'
A OW+~-.4—— -———-— - —:
PU/SC plug/Q V1446. g :9 § § fig §

é
a o
LIfl’Hi/A/ (Eff/wt (21

%(/SC O/‘St'{$~ "75

Figure 6. Upper and lower threshold curves for 0-6
at each of the pulse durations and separations at which
the double flash was reported.

 

' ~ /5’ £300 H75

 

 

 

 

 

/«5
5 ’5 25 n 75‘ me 150/;
/.>‘
A3 .
/.7[ 900 "-1 /.1
/.0»
/.5 L
/%
13

 

 

 

°\;\'ua'~qo\&\'\léo'~q

 

 

 

 

 

 

 

 

A; 6
A,[ 00vxi
, - . M
5 .f0 75' /00 lia"
/./
/Jv . N _
' fji'- 3x0
11/. Q: .9
,
h. M“ S .‘o
§ 9r Q3 .7
"S? 'X ‘ :36
‘3 .7 ' \1 ,5‘.
N I
g-éf 9
\J .37 .3
‘/ ‘ o - . + - .2 9L , T 4 , g
‘ of 15 2? 5'0 75 No we 5‘ 0' J5 , 5‘ 75‘ mo /5‘('
¢64/.5’3 Wuflaflok/ -m5 Pu?5€ Dueehou— 1473

Figure 7.
fashion.”

Data of Figure 1 presented in an alternative

32

duration of 25 ms. The data of Figure l are presented
in an alternative fashion, as Figure 7, to show this
more clearly.

(2) As the interval between pulse onsets is short-
ened there are fewer pulse durations at which the
double flash is reported. If the data are collected
to reduce day to day variability, Figure 1, there is
the further suggestion that the elimination of the
double flash occurs at progressively shorter pulse
durations as the separation between pulses is reduced.

(3) The double flash is generally not reported
at the longest pulse durations. The only 0 to report

adouble flash at a pulse duration of 200 ms., 0-3,
is the only 0 who does not report the double flash
at a pulse duration of five ms.

(h) If the extreme points of each of the upper
threshold curves are compared, a clear trend is re-

' vealed. The upper threshold at the shortest OI -ON
interval at which the double flash is reported is
lower than that of the 1000 ms. ON-ON interval. Al-
though there are individual trends in the lower thresh-
old curves, there is no grou p trend, i. e., seventeen
of the curves lepe up, fourteen slope down, and three
remain level.

Introspectgzg Material

All 0's agreed that the additional flash which emerged
at midrange intensity levels was the second member of the
double flash pair. As the lower threshold was approached
on the ascending series, the 0's reported that the single
flash appeared to hesitate and then "peak" in brightness
before fading. At any combination of pulse duration and
ON-ON interval, the second flash appeared most clearly soon
after this preliminary experience was reported. The se-
quence was orderly and allowed the O to "set himself" to
report the emergence of the additional flash. The transi-
tion in the region of the lower threshold appeared quite

abrupt when compared to the gradual transition in the re-

33

gion of the upper threshold. The abrupt transition was ac-
companied by a feeling of confidence that the Variac setting
was indeed a "true" setting.

In he ascending series, the second flash was most
distinct shortly after its emergence and at pulse durations
of about 50 ms. Even under these optimum conditions how»
tever, it was never reported to be as bright nor as distinct
as the first flash. Although the second flash was generally
most prominent at the borders of the disk, in these cases
it seemed to occupy the disk more completely with a more
pronounced dark interval separating it from its mate.

These circumstances often led the O to report an illusion
of abient movement in the line of sight. The dimmer sec-
ond flash was interpreted as the terminus of the spatially
receding initial flash.

As the pulse duration was lengthened to about 100 ms.,
the separateness of the two flashes was lost. What was
observed could be better described by speaking of a single
prolonged flash with two brightness peaks. These peaks
became less distinct as the pulse duration was lengthened
further, i.e., the second flash appeared after only a minor
decay in the brightness of the first.

Fbur of the 0's reported three flashes per pulse under
certain circumstances. These circumstances were also most

favorable to the clear emergence of the second flash, viz.,

3%

pulse durations of intermediate length and long intervals
between successive pulse onsets. The third flash appeared
only in the low and middle range intensities of the double
flash. Thus in the complete ascending series an 0 might
report the number of flashes as "one"-”two"-"three"-"two"-
"one" or as "one"s"three"-"two"-"one". The descending
series would always begin as "one"-"two". The triple flash
was not reported until five to ten hours of observation
had elapsed. In their first reports, the 0's did not speak
of a third flash, but of a "flutter" between the first and
second flashes. It was as if ingiving a rhythm to the
visual experience, it would be more suitable to count
"UNE-two-three, ONEetwo—three" than to count "ONE-two,
ONE,two". With additional observation, the 0's became con-
fidant that a third flash was being viewed.
Eccentric Fixation V

When the target was limited to a visual subtense of
l.6°and imaged foveally, it was impossible to obtain the
double flash at any pulse duration or intensity level.
Fixating the border of the disk instead of its center yield-
ed a prompt and clear emergence of the double flash if the
conditions were appropriate. The nature of the conditions,
in terms of the experimental variables, was not explored.

The capacity of the peripheral retina to sustain the

double flash also became apparent if fixation was not main-

35

tained on the luminous dot as the thresholds were being
determined. If the intensity of the target was set just
below the lower threshold or just above the upper thresh-
old, a slight shift in fixation from the center of the disk
resulted in the appearance of the second flash.

Viewing Through Colored Material

 

TWO additional 0's, both highly practiced, were asked
to view the disk through blue and green cellophane. Both
reported two flashes and two color experiences. The first
color experience took on the color of the cellophane, blue
or green, with the second color experience tending toward
its complement. One 0 felt sure that the two colors were
a property of the two flashes.‘ The other 0, although re-
maining uncommitted, thought that the complement may have
appeared only as the second flash faded. The 0's indicated
that the color experiences were transient and presented a
difficult observational task. When viewing the target
through a filter which passed only the longest wavelengths,
the author and one of the above 0's agreed thatthe second
flash was severely weakened and on occasion completely
absent. For all of these observations a 50 ms. pulse was

used at an ON-ON interval of lOOOms.

36

DISCUSSION

A successful explanation of the double flash phenomenon
will have to do more than show that retinal discharges are
spread out in time or that discharge latencies among ele-
ments of the neuroretina vary. It will also have to show
how these discharges assume a temporal organization which
allows them to be experienced as a double flash. The Len-
nox and Kadsen type explanation falls short on other counts
and can be dispensed with apart from this consideration.
This explanation asserts that the first and second flashes
are to be attributed to the cone and rod ON effects respec-
tively. By this time it will be clear to the reader that
(l) a sizable body of evidence exists indicating cone la-
tencies are shorter than those of the rods and (2) if the
extra flash is to be the result of a cone OH discharge,
it must emerge ahead of the single flash present at scotOpic
levels and (3) this does not occur.

The myriad neural interconnections which are known to
exist within the retina form a tempting dgug g; machine
for the worker in vision. Although it may not be possible
to predict the results of a given experiment based on a
knowledge of this structure, it may be possible to conserva-
tively conclude that experimental results "are consistent

with" such knowledge. Thus asserting that the second flash

37

may be the result of joint rod and cone activity made pos-

sible by convergence on common retinal elements may be true
(10), but the assertion fails to indicate why there should

be twp flashes, i.e., why the discharge activity should be

bimodal rather than prolonged.

To account for the "silent" interval between flashes,
an inhibitory effect is suggested. Since at midrange in-
tensity levels both rods and cones are dischargi.g, and the
cones have the shorter latency, it must be the cone dis-
charge which is momentarily interrupted by the rods. The
anatomical mechanism by which this is accomplished can be
tentatively speculated upon as being the lateral association
neurons (20, 21). The physiological mechanism is more
obscure. It may prove to be conceptually help ul to look
upon the vertically oriented pathways of the retina (photo-
receptors, bipolars, and ganglion cells) as contributing
primarily to the summation of excitatory influences and the
lateral association network as functioning primarily in the
communication of inhibitory influences.

Several lines of evidence, reported in the last chap-
ter, converge upon a common conclusion, viz., both rods and
cones are involved in the double flash experience. Much of
the evidence has a prima.facie duality about it. It has
been shown that:

(1) viewing the target through a filter which pas-
ses only the deep reds severely reduces the effect.

38

(2) although the double flash cannot be seen
foveally, it can be clearly seen when a target which
subtends a visual angle of l.6° is imaged on the
peripheral retina.

(3) the second flash becomes less prominent as the
pulse duration exceeds the average latency difference
reported between rods and cones.

(h) the double flash does not emerge at the in-
tensity"extremes.

The nature of the'second flash at various intensity
levels lends further support to the notion of a duplex
receptor involvement. At low midrange intensity levels,
where supposedly only those cones having the lowest thresh-
old would be activated, the double flash is weak. Percep-
tion at this point is rod dominated. As the intensity of
the target is raised, presumably still within the rod range,
the number of cones involved becomes considerable and the
second flash is most noticeable. A further increase in
the intensity of the target once more results in perception
being dominated by a single sense cell population. The
result is a single flash experience.

The results reveal that the interval of intensities
within which the double flash is reported becomes smaller
as the separation between pulse onsets is reduced. Fbr
every 0, at each pulSe duration, the upper threshold was
higher at the longest ON-ON interval used than at the short-
est. The verbal reports of the 0's leave little doubt that

the reduction in the upper threshold reflects a decay in

the brightness and duration of the second flash. The rea-

39

der will recall the instructions given to the 0's: piggy
pulse must be accompanied by a doubling in the flash rate
for the double flash to be reported. As the separation
between pulses is shortened, the second flash becomes less
distinct, the experiences more compressed, and the 0 un-
certain as to whether or not every pulse has two flashes
associated with it. To be sure, extra brightness "pips"
are noted, but assigning one to each pulse is another mat—
ter- hence the double flash is not reported. If the author
were to advise another worker in a further study of the
perceptual inhibition of the second flash, hindsight would
recommend that the method of indexing this inhibition be
changed. The deterioration of the second flash follows a
tapered course. The "all or none" criterion employed in
the present investigation Egg seems to be an inefficient
way to reveal this gradual alteration. In a replication,
a matching technique would be preferred. The 0 would be
reouired to match another target with the second flash by
manipulating its brightness and duration. The settings of
the matching target under the conditions of the experiment
would compose the raw data.

The author in his advisory capacity would suggest
other modifications as well. One of these concerns the
instrumentation. The pulse durations and separations em-

ployed were obtained by adjusting the aperture and speed

1+0

of rotation of the episcotister. As the open sector of
the episcotister swept across the beam from the Balopticon
Projector, the illumination first appeared on the disk at
about "7:00" and left the disk at about “1:00". Although
this sweep, as such, was never reported by any 0, a com-
pelling illusion of rocking or pendular movement between
these two positions was often reported. In addition to
the difficulty this introduced for the 0 in reporting on
the double flash, it indicates a methodological flaw. If
the illumination swept across the target in this manner,
and if fixation remained at the center of the target, then
it follows that the peripheral retina was stimulated both
before and after the foveal region. To some extent then,
the rod elements had an opportunity to discharge before
the cones and to continue their discharge beyond that of
the cones. The earlier stimulation of the rods acts in
the direction of obscuring any perceptual effect which re-
quires the separation of the rod and cone discharges. A
stimulator capable of producing square wave forms would
eliminate this difficulty.

It will be recalled that each of the 0's claimed dif-
ficulty in trying to break down the double flash into com-
ponent colors. In the light of this difficulty, one is
led to wonder how Judd and some of the earlier investiga-

tors were able to describe color nuances among the phases

#1

of the after image sequence with such assurance. Although
it might be assumed that the second flash corresponds with
the Hering after image or third phase of the after image
sequence, some caution is required in this statement. One
of the 0's felt that the second flash tended toward the
complement of the first. If the report is accepted, this
would link the second flash with the Purkinje after image.
Indeed, if the third flash reported by the majority of the
0's in the present study is regarded as a member of this
after image sequence, then the second flash cannot be
equated with the Hering after image. This follows from the
observation that the third flash appears betgggn flashes
one and two. As it stands, a definite assignment of the
second flash to one or another of the after image phases
is not possible.

What about 0-6? Why should his upper threshold data
be in such marked contrast to that of the other 0's? It
may be an instance of individual differences in perceptual
research (33), it may be that the 0 was attending to other
aspects of the situation, it may be that the O was inatten-
tive, or any number of factors. Two features in the results,
unique to this 0, stand out most clearly. First, he reports
the double flash at all combinations of pulse duration and
separation. Even when the interval between successive

pulse onsets is #00 ms. and each pulse is one half this

h2

interval, the double flash is reported. Secondly, although
the upper threshold data are erratic, the curves do show a
tendency to rise at the shorter pulse separations- a rever-
sal of the usual trend. If forced into an explanation of
this outcome, it could be claimed that as the second flash
was weakened by bringing the pulses closer together, the

0 in an overzealous effort to furnish as much information
as possible allowed his fixation to wander over the disk
”in search of the second flash." It has already been pointed
out that the second flash will reappear under these condi-
tions. Thus the rise in the upper threshold and the sus-
tained ability to view the second flash at the shortest
pulse intervals and longest pulse durations.

The ability of the peripheral retina to sustain the
double flash can be accounted for in terms of its histology.
If the rods temporarily inhibit the cone 0N discharge, they
should do this more effectively in the periphery. Here
they are in numerical superiority, the lateral internun-
cials are more plentiful (29), and their average separation
from the cones smallest. It must be supposed that what-
ever weakening occurs in the double flash due to a decrease
in the number of cones mediating the initial flash is more
than compensated for by these considerations and the absence
of a rod free fovea in the center of the visual field where

the double flash is weakest.

SUMMARY

Several investigators have reported an elaborate series
of perceptual events in response to an isolated pulse of
light. Under specifiable conditions, two flashes are ree
ported to each photic pulse. The nature of these conditions
has led to the speculation, both here and elsewhere, that
che phenomenon requires an explanation which emphasizes the
role of the duplex system of retinal photoreceptors. In an
attempt to clarify the basis of the double flash, an experi-
ment was performed to determine the effect of (1)pulse
duration and (2)5eparation between successive pulses on the
range of intensities within which the double flash could be
obtained. The target was an opal glass disk subtending a
visual angle of 15: The disk was illuminated from the rear
by a projector whose beam was periodically interrupted by
an episcotister. The duration and separation of the pulses
was controlled through the episcotister. Quantitative data
from five of the six observers revealed that as the interval
between pulses was reduced, the strongest intensities at
which the double flash could be obtained tended to decline.
As the pulse duration was lenghtened beyond 50 ms., the I
impression of duality became progressively weaker until at
pulse durations longer than about 150 ms. it was no longer

present.

h)...

Collateral observations revealed that the double flash
failed to appear when the target was imaged entirely within
the fovea, but could be made to appear with eccentric fix-
ation. The double flash was also observed to be markedly
weakened when the target was viewed through a filter which
passed only the deep reds.

An explanation of these results has been attempted in

terms of inhibitory influences at the level of the retina.

APPENDIX

1+6

TABLE 1

RAW DATA* FOR OBSERVER l SEEO'W'IISG THRESHOLDS

FOR

PULSE
DURATION

5

100

1570

2'70

7"

—

_:;OTI'Z T733 ASCENDING AND DESCENDIIIG SERIES

If“.

It"
*2]

 

 

 

 

 

 

 

 

 

ON-ON INTERVAL

#50 500 600 700 800 900 1000
A D D A__D..L_A__.D._.A._D._.L_.D__A__Jl.
.003 .079 .05 .055 .05 .030.01.3 .055 -063 .043

.17 .21 .1? J! .37 .57 .44 am .44 ~47
.003 .043 .034 .030 r034 .72.: .01: .034, .072! .033 752§‘.02§.023A mil
.07: .07: .1/ .095 .12. -/I .15 .13 ml .20 .16 46 .14 .22
.02: .025 .00; 015.01: .02; .010. .0101 .0252 .022 .025

4503.05 00 -07 .094. .088 .01 oil .12, ./3 ,/l./5‘./3

.L.....m... w “a...
.01‘ .010 oOI6 '0“ .0/6 0020.0“ .01‘ .0/6 .020
.04 .04 .030 .050 .038 0073.088 088.096 '033
'0"- ‘W‘ '0” ~0’3 101.2 .016 .012. .on/ .012. .01;
.019 .031 003‘ '018 005 004 '07 '06.; 007 '063

L021 .018 .004 me .012. :0“! .004 .0"! .0/2. .0/1/
.03; warms; .oadhm .030. .032 .051 .05 .003
«2.20 .015 .0"! .0". .004 .074 .002, Mai
'01; Ipa%oon-8 .025 00” '0” 1032 0032.

 

 

 

 

 

 

 

 

 

* in c/ft2

1+7

TABLE 2

RAW DATA“ FOR OJSSERTER 2 SLICW'JIIG TE'IRESEIOLDS
FOR BOTH TIE] ASCETIDI. G AI-ID DESCEEjDL-IG SERIES

ON - ON IIJTERVAL

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MOO Soc 600 700 800 900 1000
PULSE IA Ammr... “.1 - -.
DURATION A D A D A D A D A D A j) 'A D
_LT .13 .0/090 073.1: .07?
5
UT .3/ .3100 .3912 .0;
LT -/2 J1 .096 07 .072 .072.” .090
15
UT - iii-37 -iltfl-.. .z/ 0 _-0 47.0.3-
LT }.// ,// .13 ./1 JV ./.5‘.22 J?”
25’ t ?
UTIL J“ 4 0-4 .......--E}.?L/i ”4/9 '0‘l/_ 274;;- .5’/ '5/ {3"
LT l 3.07 0723.07: 073.07! .096 .096 .096
so ! S
' !
UT 1 ...... ML -.__ L07 .3 .37 37.47 .9 .01 .50:
LT 1 £063 firm . .032 036.022 .023
75 U 3 §
T - - . 51.2. -0 .0: _ -02 L0- -_-../_é...._/0.._. 0
LT 53-023 $25.02? 031.032 025.02! .020
l
100 I I ';
UT i
L_..........1L,. . g . .. 20.6.5. ioj‘ilzqq.".1459.L§£8-...-;££LL2§. '049
‘ |
150 f f i
‘ UT- ' 3 . F
LT“) i .7 , i 3
200 '5 E ' 3 l T
UT .I i ; 1 _ §
1 ‘ aw ....-.~....a.— 4 w“

 

 

 

#8

TABLE 3

RA '1' DATA* FOR OIEQERVER 3 SHOWII‘G THRESHOLDS
FOR BOTH TIE ASCEEJDIITG AIVD DESCETTDING SERIES

ON - OI! II‘.‘ ‘I‘ERVAL

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A00 500 600 200 800 300 1000 _
PULSE -_A_ D A _D A .D A D A n A D.
DURATION fl "
LT
6
LT ,02? .021 2'2; :07 .0257 .036 .020 70202010 .025 055“" .025"
is ”
UT
-00....09-../'3_ .21..-.0.22.--.20.213, 2:2 .21...2-.4.21 .
LT r.“ 012/ .020 .012 ,02 .012 .020 .020 023.020 020.012 .516"
27 ,
UT L-~ ~04 .37 .57 .Q‘J/ .084 .U .0: .03 .59 ,0; .5
LT I {,0/7 ‘023 lwz ,0/6.0/0 0181.008 .014 .012 .0/6 .0/6 .0/2
F
30 ’
“T 1.0 3/20 ...-£3._r.3z..-.;s»£ .12... "-311.20... 0:
LT .003 .012 .012 .014 .012 .016 .010 .0/4/ .00? .012 .0/0 .011 .0/2 012
75 I _
. t
Ur .flfl...:93.°y0f/.’.. '43)} '47,” '.”.’?‘.'”‘:3. soft”? 205.295 :93“! '0“
L3? .012 0’6 [012 ,0]; .0/0 -0/2 .009 .0/2 22003 .012 .008 .020 .007 .010
i S-
100 z
[-1 é
”1 311.4921}? 0.2.0.0? “-123. 250......0111-423 _....:43.2.1.e'€.- -;e.2..2..-.0.3.3.- -022.
LT -0// .012 L012 .012 20/0 0103.010 $150100? .0/2 .007 .009 -007 .008
i 1
s ~ .
1’0 UT i i
I ;
” ”CO 3 ‘ ‘10}? 1.0/3 013/010. "0’1_'_0H’./‘__w-~0’I?j.0fllflh .012 .012 .012 -0l6 __,Q20
LT .0/2 .01 .008 0102003 .010 .008 .012 .012 .003
200 * 3
~~ 1
T l -l .0/25‘ .OIZ 0H: ,olzLa/o __,A,‘z__.0u/ 01% 0,3. .0”,

 

* c/ft2

TABLE 1+

RAW DATM‘ FOR O‘SER‘ ER 1+ SHO‘JIII-LG THRESHOLDS
FOR BOTH THE ASCEE‘IDIIIG A’TD DEFSCEIIDIZCG SERIES

ON... OI. II‘ITERVAL

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L100 500 600 200 800 99m Jenn---
PULSE A D A D A D A D 11 / 1L D A. D
DURATION LT ,3, .ff ,1: .37 .37 .37 .‘/‘/ .3/ .57 .37 46 15/
57
UP «9.? 06 1.65 2.10 jzA 1,1,0 2.2! 20/ 2.? -971 3.2 “(.0
LT 'bgf‘I‘EEI-Hw .7133 -/2 .0 .12 .01 .070 .05’ .05’ .011
15
UT .
«57 .5061 .11 .61 0’9 .6? 4321”]? .62 .6;
LT .02/ .025 .05 .022 .05' .0zEF05‘ .032 .02/ . 36
25’ 3
UT ‘-
‘r-—- _L”_‘“__‘J_.__2_3_m../1 .27 _ :3’ .37 .24.}, 341.07 3-7..
LT .08? .07 20727 .029 .0: .113 .015" 177 .071 010;.12 .010 .12 .15
)xfi ‘
UT .24 .31 .30 .22 .37 .31 .0-1 «10.0“: .31LVV-__._;Z%_1_§Z__ 03
LT .00 .022 .063 032 .12 .063 .011 .056 .11 .071?! .13 .22 , 11
75 i .
U50 . ,1
.Ffl-LZQ1LZ7 ...:.C.".‘ 'Li?....--:..2fl.;f.’.".'.,251.337- .. 773'.“ 1‘ t '. 7/. 7276
LT .05 .4 1.0V 4321.053 , oy‘ .035 ,0323,M5‘ 1022!' If (’7
100 i I 1
UT I 3‘ 31
.096 .m/JJb .028 .1; .096 J! W .1141; $.07 2"; “:30
‘LT I" 020 0201102; 0101.032 .022. .020 .008 I032 .016 {05' .07
1530 l . . (
UT l 5 3 i
1.020 .236 41-....- '2 0.4!..-00720122..--:22. 020.....022... 42..--..-/2.-
L3? 1 1 9
200 i ; 1‘ } :
UT '
J- +____ mmJ—muuym-jL—~.~.a,wvv .. . can. ’4

 

 

 

RAW DATA“ FOR OBSERVER 5 Si-iOWII'O THRESHOLDS
FOR DOT}: THE ASCEIADIG AND DESCEICDIIG SERIES

035- O; I I.» .TERVAL
I000 509 600 7 00 800 BOO lOOO

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

931.10.. LL D A D A D A D A D A D A D
“‘4‘ “LT .01: .07} .078 .078 .000 .073 .096 .07! .070 .063—
c:
/
LT
.096 .12 .fi .43 .‘1/ .W .68 _;.Z_‘;.§L..4--“
LT '03‘ 2032 '03: '036 .025‘ .0zs’r.032 .032 .036 .028
1:
U2
"77 fiLMfl -/6 .d 20 .22 .27 ’37“
LT .02; 012:.020 .022'022 .025’.02)" .025"
25
1:10 I
101/ .011/3 .12 13 .121: .18 1’3
LII ’ ' {019/ '50; 1'0”] .0" 10/3 .0/6 .0/9/ .0/6
i i 3'
, L1,; 50.. 41‘ a
a 2' .0 .50 063.07! .251 ._0_;_0___
LT ”011/ .016 .0191 .0/6 .012 .012 0,2 .012
75’ i
UT g
902: .011 .06’ .05’ 063 .05 01! __.07
L11“! ?L012 .012 .01; .012 .011 .012 012 ,01;
100 1 *
UT. i
T 402$ 021.0% 0759.036 .036 .053 .05;
4 LT! "3 3'0’2 '0’1'0’2 20/4417 .012
130 1 ‘
.1 I *
U‘J.‘ 5 3 ‘
‘ 1* r .. 1 A ;2020L.az{%0zz ,01C 1 .0:
LT } 5 ,1 l
200 _‘ 5 ‘ g 3 g ’
UI' , i i 3 ‘

 

* C/ft2

PULSE
DURATION LT

5
UT
LT

15
UT
LT

2?
1.7T
LT
UT
LT

75
CT
100 I
UT
LT

l ‘50
{IT
LT

ZOO
713-3

RAW
FOR

DAAT
BOT

TABLE 6

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

* FOR OBSERVER 6 SHOWIICG TI-ARESIEOLDS
H THE ASCEEIDIIG ATFD DESCEIQ'DIZ'QG SERIES
ON-OII ILITERVAL
- ADD. 700i - 600 700 800. _990. 1000
A D 1A DDIA. D A ' D A D A D A D_
.096 .07 3.07: .OIZI,038 .076 .053 .12 .063 .12 17. 421.17. 12.
f v
1 !
,96 /./ éfl/ 20L3] .‘I‘Il.37 ,6! .‘l4 a‘l; 17‘ 0’ ‘ 63 _
.032 ,ozzfmg 0231.056 .04 .0: .0; .0? MFIWK ,036 ,0? .0‘15'
' t
a i it
My Alt/3.3L .72_L.2j 15 .22 .ZL/ .37 +121qu .37 . .27 . 2] -
.032 0221.01: .0230: .022 0‘1: 025' .032 -020§.02r.02r[.02r .02?
I ". ‘ 2
! . E g
i f g g
.37 37937 3501122 171.13 1/8 127 -22;.#21k 122. L131 . 2.2.
‘ fir
.035 .018 591? 42200015" .ozf 1232. .052. .030 .02: .032 .02; 023’ .011,
A l .
E 2 . 1
.37 £171.22 .17 L 7.0 .27 {.3/ .27 .27 .Bygu 2110,, 231
, T
2015 :018‘022 .011. I 032. 016 'Ph’ .0150”. o0’6[02{.011§.0w .020
i
1 ‘ f 1 i ’
.97 15’ 027 22' 22 18/14: a .12 db} [6 .IZj 22. -22..
3% 021.027. 001.025.1916 .01? .OIbioz‘r .0133“; 020 02? 014
I
. * . 1 f 5
./g 157 0’20 1’2 1°96 107 ‘,/6 a“,]q§'l12-0 Ill ’.27 1’, ;-/Z all.“
.011, .008 1017 ~0’Z 0:018 .012. {.016 .02210" '0’:- 30’5 ‘0" '0” ’ 0Q
- z I i .
"- I I
”33 2°" .07! .Ws’!.07 077T”, .0720?! .12. 205’ ,053‘1.073 .01.;
T . t "7"
on. 012 L012 .012 .010 0203.01? 07.40;; ,018 30/! ,Olep/V ,0”,
' ' E 3 j' t
1 i E ' f
.039 100343.12. .01)“ .121i72 .063 £073 ml *0“ 001-me 4051?

9.

10.

ll.

12.

13.

52
REFERENCES

Adrian, E.D. Rod and cone components in the electric
response of the eye. J, Physiol., 19H6, 105, 29-37.

Algfirn, M. Metacontrast. Am..J.thom., 1952, 29, 631-
.6.

Alpern M. Metacontrast. J 0 t .80 Am ., 1953, 113,
6h8-é57.

Alpern, M. Relation of visual latencZ to intensity.
A.M.A. Arch, Opthal., 195%, 369-37 .

Armington, 3.0. A component of the human electroretino-
gram associated with red color vision. J. Opt Sgg.
Amer., 1952, #2, 393-h01.

Bartley S.H. The features of the optic nerve discharge
underiying recurrent vision. J. exp. P§19h91., l9h2,
30, 1254135. 1

Bartley, S.H. Some factors influencing critical flicker
frequency. J._Psychol., 1958, #6, 107-115.

Bartley, S.H., & BishOp G.H. Factors determining the-
form of the electricai response from the Optic cor-

tgfi of the rabbit. Am, J. Physiol., 1933, 103, 173-
1 .

Bartley, S.H., & Bishop, G.H. Some features of the Optic
nerve discharge in the rabbit and.cat. J, Cell, &
Cemp. Physio;,, 19h2, 19, 79-93.

Bartley, S.H. & Wilkinson, F.B. Certain factors in
producing complexity of response to a single pulse

Cheatham, P.G. Visual perceptual latency as a function
of stimulus brightness and contour shape. J, gap.

Dodt, E., & Enroth, Christina. Retinal flicker response
in cat. Acta physiolI scand., 195%, 30, 375-390.

Dodt, E., & Wirth, E. Differentiation between rods and
cones by flicker electroretinography in pigeon and
guinea pig. Acta physiol, sggnd., 1952, 27, Suppl. 100.

11.. .

16.

17.

21.

22.

23.

2%.

'J
.1 3
/

Donner, K. The Spike frequencies of mammalian retinal
elements as a function of fave length of light. Acta
physiOI. scand., 1930, 21, Suppl. 72.

Enroth Christina. The mechazisn of flicker and fusion
studied on single retinal elements in the dark adapted
eye of the cat. A t - sio can , 1952, 27, Suppl.
100.

Fry G. A. Depression of the aomivity aroused by a
flash of light by applying a second flash immediately
afterwards to adjacent areas of the retina. Am, J.

thsiol,, 193M, 108, 701-707.

Graham, C. H. Vision. III. Some neural correlations.
In C. Murchison (Ed.), Handbook of general psychology,
Worchester: Clark University Press, 193 .

Granit, B. The dark aiaptation of mammalian visual
receptors. Acta phzsiol, scand. 19h4, 7, 216-220.

Granit,_R. Receitors and sensor *ion New Haven:
Yale Lniversity Press, 19 ;.

  

Fartline, H. R., &'Wagner, H. G. & Ratliff, F. Inhi-
bition in the eye of Limulus. 3. Gen. Physicl,, 1956,

39 , 651’673 o

Hartline, H. E., & Ratliff, F. Inhibitory interaction
of receptor units in the eye of Limulus. J. Gen,

Judd D. B. A quantitative investigation of the Purkinje

Lennox, Margaret A., & Madsen, Annelise. Cortical and
retinal responses to colored light flash in anesthetized
cat. J h uro h sio , 1955, 18, h12-H23.

Levitt, M. Some parameters of visual apparent a ient
movement. Unpublished Ph. D. dissertation, Michigan
State University, 1957.

Madsen, Annelise, & Lennox, Margaret A. Response to
colored light flash from different areas of optic
cortex and from retina in anesthetized cat. J, Neuro-
physiol., 1955, 18, 57h-582.

26.

27.

29.

30.

31.

32.

33.

McDougall, A. The senset1one excited by a single
m01entary SW11] tion of the eye. Brit J P5 cho
1999 1’ 78-113,

   

Osgood, C. E. Math and theori in e nerimental
psychology. flew York: Oxford university Press, 1953.

Poffenbe ger, H. T. React:Lon ti me to retinal stimulation.
Arch :sychol., 1912, #23.

 

 

Polyal The vertebrate vi§QQ;_;1Q£gQ, H. Kluver
(Ed.3, SChicago: University of Chicago Press, 19 57.

Sweet, A. Temporal discrimination >y the human eye.
Am. J. Psychol,, 1953, 60, l< 5-199.

Talbot, 8. A. ., & Marshall, W. A. Multiple response in

the eptic tract and 08% ccortex of the cat. Am, J,
thsiol,, 19’0 129,

 

Werier Studies on cowtour. I.Q1e1itative analysis.
__._.5 HPSYCllOl £4, 1/35, ’47, ”'0" 61:.
Within, H. A. The nature an -d importance of indivi ual

L:r\

differences in perce1t1,1. J, Pers s,, 1999 , 18,1L5-170.

a: ESE my

 

 

 

 

 

 

1111111111111