Content uploaded by Charles G Gross
Author content
All content in this area was uploaded by Charles G Gross
Content may be subject to copyright.
Available via license: CC BY-NC-SA 4.0
Content may be subject to copyright.
The Journal of Neuroscience, April 1990, m(4): 1154-i 164
Afferent Basis of Visual Response Properties in Area MT of the
Macaque. II. Effects of Superior Colliculus Removal
Hillary R. Rodman,” Charles G. Gross, and Thomas D. Albrightb
Department of Psychology, Princeton University, Princeton, New Jersey 08544
In a previous study (Rodman et al., 1989), we found that
many neurons in the middle temporal area (MT) of the ma-
caque monkey remain visually responsive and directionally
selective after striate cortex lesions or cooling. In the present
study, we examined the effects of superior colliculus (SC)
lesions and combined lesions of striate cortex and the SC
on the visual properties of MT neurons. Removal of the SC
alone had no effect on the proportion of visually responsive
cells, strength of direction selectivity and direction tuning,
orientation tuning, receptive field size, or binocularity in MT.
There was, however, a slight increase in response strength
to both stationary and moving slit stimuli. In contrast to the
minor effects of SC lesions alone, addition of an SC lesion
to striate cortex damage abolished all visual responsiveness
in area MT. The results indicate that pathways damaged by
the SC lesion are not necessary for most of the properties
of MT neurons found in the intact animal, although these
pathways are capable of sustaining considerable visual re-
sponsiveness and direction selectivity when striate input is
removed.
The middle temporal area (MT) of the monkey is an important
locus of cortical visual motion processing and has been exten-
sively characterized both anatomically and physiologically (e.g.,
Allman and Kaas, 197 1; Dubner and Zeki, 197 1; Zeki, 1974;
Gattass and Gross, 198 1; Van Essen et al., 198 1; Maunsell and
Van Essen, 1983a, b, c; Albright, 1984; Albright et al., 1984;
Ungerleider et al., 1984; Allman et al., 1985; Ungerleider and
Desimone, 1986; Rodman and Albright, 1987, 1989). In an
earlier report (Rodman et al., 1989), we showed that consid-
erable responsiveness and direction selectivity remains in MT
after removal of its striate cortex input. The finding of residual
responsiveness indicates an input
to
MT from a portion of the
visual system that receives visual signals in parallel with those
reaching striate cortex. Anatomical data indicate that such an
Received June 27, 1989: revised Oct. 17, 1989; accepted Oct. 30, 1989.
We wish to thank L. Kuseryk and M. Hess for work on the histology and figures
and M. Rosengarten for word processing. We are also grateful to D. Bender for
assistance with the superior colliculus lesions and to J. Maunsell for helpful dis-
cussions. Support was provided by N.I.H. grant MH-194120 and N.S.F. grant
BNS-8200806.
Correspondence should be addressed to Dr. Hillary R. Rodman, Department
ofNeurosciences, M-008, University OfCalifomia-San Diego, La Jolla, CA 92093.
Reprint requests should be addressed to Dr. Charles G. Gross, Department of
Psychology, Princeton University, Princeton, NJ 08544-1010.
a Present address: Department of Neurosciences, M-008, University of Cali-
fomia-San Diego, La Jolla, CA 92093.
b Present address: The Salk Institute for Biological Studies, P.O. Box 85800,
San Diego, CA 92 138.
Copyright 0 1990 Society for Neuroscience 0270-6474/90/041154-l 1$02.00/O
input pathway might involve the superior colliculus (SC). The
superficial layers of the SC receive direct retinal inputs (Hubel
et al., 1975; Pollack and Hickey, 1979), which potentially gain
access to cortex via the pulvinar nucleus of the thalamus (e.g.,
Benevento and Standage, 1983). Portions of the inferior and
lateral pulvinar project to MT (Maunsell and Van Essen, 1983~;
Standage and Benevento, 1983), and some cells in the inferior
and lateral pulvinar remain responsive in the absence of striate
cortex input (Burman et al., 1982; Bender, 1983). In addition,
the sensitivity of MT neurons to stimulus motion and their
relative insensitivity to stimulus form suggest a potentially im-
portant role for pathways involving the SC in the genesis of
response properties in MT.
In the present study, we continued our investigation of the
afferent basis of MT response properties by recording in MT
after lesions of the SC alone and after combined lesions of striate
cortex and the SC. We found that lesions of the SC alone pro-
duced only minor alterations in the characteristic properties of
MT neurons. On the other hand, the residual responses seen in
MT after the striate cortex lesion were abolished by the addition
of an SC lesion. Some of these results have appeared briefly in
abstract form (Rodman et al., 1986).
Materials and Methods
Subjects
Subjects were 4 male
Macaca fascicularis
weighing between 3.1 and 5.4
kg.
SC lesions. Unilateral (left) SC lesions were made in 2 monkeys (case
nos. 552 and 579). After a l- to 2-week recoverv ueriod. each animal
was recorded from 5 to 6 times in the space of j-5 week;.
Combined striate cortex-SC lesions.
SC lesions were made in 2 mon-
keys that had previously been studied following bilateral ablation of
striate cortex. The SC lesions (bilateral
in case no. 555, right hemisphere
only in case no. 56 1) were followed by a 2-week recovery period.
MT
was subsequently recorded from 3 to 6 times over a 2- to 4-week period
in each animal.
In addition, single units in normal MT were studied with the same
methods prior to the striate cortex lesion.
Surgical preparation
Access both to MT and to the SC was obtained via a 5-cm-diameter,
stainless-steel chamber implanted over the midline at least 1 week prior
to the SC lesion (or the initial striate cortex lesions), as described pre-
viously (Rodman et al., 1989).
SC lesions were made using methods described previously (Bender,
1983; Bruce et al., 1986). Briefly, 4 radiofrequency lesions were made
in each SC, using the locations of visual receptive fields recorded on
several electrode penetrations into the SC and Cynader and Berman’s
(1972) retinotopic map of the SC as a guide. On the day of the SC lesion,
the animal was anesthetized, placed in a stereotaxic apparatus, and
prepared for recording visual responses (see below). The cap of the
recording chamber was removed and any tissue growth over the midline
The Journal of Neuroscience, April 1990, IO(4) 1155
of the skull cleared away. A small craniotomy, centered approximately
on AP-0, was made over the midline. The dura was then cut and re-
tracted. Electrode penetrations were made vertically into the SC until
its most dorsal point, where cells have receptive fields with centers at
about 5” along the horizontal meridian, was located. The placement of
the lesions was calculated relative to the stereotaxic coordinates of this
point. The lesions were produced by a Teflon-insulated stainless-steel
probe with an exposed tip of 0.5 mm using a radiofrequency current
generator (Radionics, model RFG-4). The current was passed through
the probe at an intensity sufficient to heat the tip to a temperature of
74-80°C for 45 sec. After the lesions were made, the dura was sewn up,
the cap replaced on the recording well, and the monkey allowed to
recover.
The surgical methods for the striate cortex lesions have been described
in Rodman et al. (1989).
Recording procedure and visual stimulation
Recording sessions were carried out under conditions of N,O/O, an-
esthesia and immobilization identical to those used in our previous
studies (Albright, 1984; Rodman et al., 1989). The procedures for re-
cording extracellular potentials from single isolated neurons or small
groups of neurons were conventional and are also described fully in the
above reports. The electrodes were angled 40” from vertical in the par-
asagittal plane, passing dorsoanteriorly to ventroposteriorly before
reaching MT. Single and multiunits in MT in the combined-lesion cases
were tested for responsiveness with moving and flashed spots and slits
of light and dark edges from a hand-held projector. Large stimuli, such
as squares of cardboard, a yardstick, large brushes, etc., were also used.
Responses to these stimuli were characterized as strong, weak, or absent.
For the cases with SC lesions alone, we used both hand-held stimuli
and computer-assisted techniques to examine the response properties
of single isolated neurons in MT. A PDP 11/34A computer automati-
cally controlled the presentation of stimuli by an optical bench equipped
with X- and Y-axis mirror galvanometers, an electronic shutter, and a
stepping-motor-controlled rotation mechanism. The computer also col-
lected spikes for peristimulus time histograms, which were displayed
on-line and updated during each new trial.
Three types of stimuli were used for computer-controlled testing of
MT neurons in the SC lesion cases. Moving slits of light and
moving
spots (single small squares) were presented at the speed estimated to be
optimal on the basis of hand testing. Dimensions of the moving slit
used for quantitative testing ranged from 0.5” to 1.0” in width and 3
to 20” in length; for the spot, the dimensions used ranged from 0.25” to
0.5” on a side. Speeds used ranged from 2Vsec to 64Vsec. The slits and
spots were swept approximately 30” across a tangent screen along a path
centered on the neuron’s receptive field (RF). Each type of moving
stimulus was presented in 16 directions of motion, with a constant
angular deviation (22.59 separating neighboring orientations.
Station-
aryflashed slits, positioned in the RF so that they evoked the maximal
response from a given orientation, were presented for 2.5 set in each
of 8 orientations, with a constant angular deviation of 22.5” separating
neighboring orientations. For each of the 3 types of stimuli, a test con-
sisted of a series of 5 pseudorandomly interleaved presentations of each
direction or orientation, Intertrial intervals for all stimuli were a min-
imum of 2 sec. The above details of presentation of both moving and
stationary stimuli were identical to those used for the cells studied by
Albright (1984).
The sequence of testing performed on each penetration was as follows.
Upon entering presumptive MT cortex, we would first assess the overall
sound of the cortex (spontaneous activity, burstiness) and the respon-
siveness and direction selectivity of the MT activity. We then proceeded
to study individual neurons and, in the case of the combined lesions,
small groups ofcells (multiunits). First, the responsiveness ofeach single
or multiunit was assessed as described above. Next, for responsive
recording sites, the RF was plotted using the smallest visual stimulus
that would evoke a reliable response (usually a short bar of light). For
the SC lesion cases, we next undertook, as described above, the moving-
slit series, followed by the moving-spot series, and then the stationary-
slit series. Finally, we compared responsiveness through the contralat-
era1 and ipsilateral eyes using the scale of Hubel and Wiesel(1962).
Data analysis
As in our previous studies (Albright, 1984; Rodman and Albright, 1989),
responses were considered statistically significant if they exceeded the
baseline rate plus (or minus) twice its standard error. To account for
asymmetries in RF dimensions, responses to moving stimuli were com-
pensated by considering only the spike rate within a time window set
to correspond to the width of the RF along its narrowest axis. Responses
to moving stimuli were measured as the average rate of firing during
this time window minus the spontaneous rate. Responses to stationary
slits were calculated as the average rate of firing during the entire stim-
ulus exposure minus the spontaneous rate.
Histology
The animals were killed with an overdose of intravenous sodium pen-
tobarbital and perfused through the heart with saline followed by 10%
buffered formalin. The brains were removed, blocked stereotaxically,
and sectioned at 50 Km in the parasagittal plane. Recording sites were
reconstructed from serial sections stained with cresyl violet. The bound-
aries of area MT were identified in neighboring sections stained with
the Gallyas (1969) silver myelin stain. The cresyl- and fiber-stained
sections were also used to examine the extent of damage to the SC and
surrounding structures. The methods for reconstructing the striate cortex
lesions are described in Rodman et al. (1989).
Results
Effects of SC lesions alone
Overview
We recorded from 103 isolated single neurons in MT on a total
of 23 penetrations in the hemispheres ipsilateral to the SC le-
sions in the 2 animals. All of these units were histologically
determined to be within the myeloarchitectonic borders of MT;
cells recorded outside of MT are not included in the present
sample. Normal MT comparison data were drawn both from a
sample of single units studied in MT prior to the striate lesions
(for estimates of incidence of responsiveness) and from Albright
(1984) (for quantitative measurements of response magnitude,
selectivity, and tuning). We will first present histological veri-
fication of the SC lesions; second, we will present data on re-
sponsiveness and response strength in MT after the SC lesion;
third, we will discuss direction and orientation selectivity in the
sample; and fourth, we will address RF size. Finally, we will
touch on a few additional characteristics of MT after SC re-
moval, including the presence of “Type II” (“pattern-motion”)
neurons (Rodman and Albright, 1989).
Histologicaljindings
Parasagittal sections through the brain stem for all the cases in
the present report are shown in Figure 1. Sections from the SC
lesion cases are shown on the far left (case nos. 552 and 579).
On the far right (case no. 56 1, left hemisphere) are shown sec-
tions through the SC in an intact hemisphere for purposes of
comparison; the SC and neighboring structures are labeled.
Missing tissue is indicated by solid black; areas of prominent
gliosis or necrotic tissue are indicated by stippling. In both cases,
all layers of the SC were destroyed throughout virtually the
entire extent of the structure. Some involvement of the medial
pulvinar and pretectum was also present in each case.
Responsiveness
Figure 2 shows the incidence of strong, weak, and absent re-
sponses to the hand-held stimuli in MT after an ipsilateral SC
lesion and, for comparison, in normal MT. The distributions
were virtually identical (x2 = 2.20,
df
= 2, p > 0.10); nearly all
MT cells in both samples gave strong responses.
For each single unit responding to at least one direction of
1156 Rodman et al. - MT After SC Removal
COMBINED STRIATE - SC CASES
no. 552R no. 579R no. 555R no. 555L no. 561R no. 561L
6.0 La
5.0 La
4.0 La
3.0 La
2.0 La
1.0 La
Figure 1.
Parasagittal sections through the brain stem showing superior colliculus lesions. For purposes of comparison, the SC and neighboring
structures are labeled in sections taken from an intact hemisphere (no. 56 1 L). Missing tissue is indicated by solid
black,
necrotic tissue or gliosis
is shown by
stippling.
cg, Central gray;
ic,
inferior colliculus;
mp,
medial pulvinar;
pt,
pretectal area; SC, superior colliculus; th, other thalamic nuclei.
the automatically presented moving slit, we calculated a differ-
ential response magnitude measure (see Albright, 1984) as the
algebraic difference between the largest and smallest deviation
from baseline (spontaneous activity level) in spikes/set.
A sim-
ilar procedure was used to quantify responses to the moving
spot and stationary slit.
Moving stimuli. Figure 3 shows distributions of differential
response magnitudes to the moving slit and moving spot for
cells in MT after an SC lesion and for cells in normal MT
(replotted from Albright, 1984). For both types of stimuli, there
was a tendency for response magnitudes to be higher after the
SC lesion than in the normal animal (moving slit: K = 25.6 s/set
after SC lesion vs 19.8 s/set in normal MT; moving spot: K =
27.9 s/set after SC lesion vs 19.0 in normal MT). The difference
between the samples was significant for the moving slit (median
test, x2 = 5.11, df = 1, p < 0.05) but not for the moving spot
(median test, x2 = 3.29, df = 1, p > 0.05).
Stationary stimuli. Figure 4 shows distributions of differential
response magnitude values for the stationary slit for cells in MT
following an SC lesion and for cells in normal MT. The mean
value after the SC lesion (10.0 s/set) was somewhat higher than
that for normal MT (5.8 s/set), and there were significantly more
high values in the post-SC distribution (median test, x2 = 7.33,
df = 1, p < 0.01).
Direction selectivity and tuning
Directional properties were characterized using 2 standard mea-
sures, the direction index (DI) and direction tuning bandwidth.
The DI, defined as 1 - (response in antipreferred direction/
response in preferred direction), is a measure of the relative
response to motion in the 2 directions along the axis of motion
eliciting the best response. Distributions of this measure for the
moving slit and moving spot are shown in Figure 5 for cells in
MT after an SC lesion and for cells in normal MT. For both
The Journal of Neuroscience, April 1990, 70(4) 1157
samples, the distributions of direction index values were clus-
tered around 1 .O for both types of stimuli, and the samples did
not differ significantly (median test) for either stimulus type.
The average DI value was close to 1 .O for both samples and for
both stimulus types (moving slit: K = 0.90 after SC lesion vs
1 .O in normal MT; moving spot: K = 1.07 after SC lesion vs
0.98 in normal MT), indicating an average tendency for uni-
directional tuning with no response in the antipreferred direc-
tion. On the other hand, the finding of many MT neurons with
DI values well over 1 .O indicates that strong inhibition to mo-
tion in the antipreferred direction was still common in MT after
the SC lesion.
Direction tuning bandwidth, a measure of the sharpness of
direction tuning, is defined as the width in degrees of the di-
rection tuning curve (a plot of response as a function of direction
of motion) at half of its maximum height. Figure 6 shows dis-
tributions of the direction bandwidth measure for the moving
slit and moving spot for cells in MT after an SC lesion and for
cells in normal MT. There was no statistical difference between
the samples (median test) for either stimulus type.
Orientation tuning
Orientation tuning bandwidth (width of the orientation tuning
curve at half its maximum height) was calculated for each unit
that responded selectively to stationary oriented slits. The dis-
tributions of orientation tuning bandwidths found after the SC
lesion and in normal MT are shown in Figure 7. There was no
significant difference between the distributions.
RF size
In normal MT, RF size increases with eccentricity in a manner
that can be described with a linear function (Albright and De-
simone, 1987). We plotted RF size as a function of eccentricity
for MT units studied after the SC lesion and compared it to the
function for units in normal MT (replotted from Albright and
Desimone, 1987; Fig. 8). Regression lines fitted to each set of
data are shown. There was no significant difference between the
slopes of the 2 regression equations (t-test for slopes). Likewise,
the y-intercepts for the two regression lines were not significantly
different (t-test for intercepts).
Other properties
Binocularity.
In normal MT, most neurons can be activated
about equally well through either eye, and monocular units are
rare (Maunsell and Van Essen, 1983b; Albright et al., 1984).
The responses of 3 1 single units in the SC-lesioned animals were
tested for binocularity using the estimated optimal stimulus and
rated according to the scale introduced by Hubel and Wiesel
(1962). In this scheme, a 1 corresponds to a unit driven only
through the contralateral eye, and a 7, a unit that can be driven
only through the ipsilateral eye. A 4, at the midpoint of the
scale, indicates a cell that can be driven equally well through
either eye. In the sample of MT neurons studied following an
ipsilateral SC lesion, 30/3 1 had a binocularity rating of 3, 4, or
5, indicating little or no eye dominance.
Pattern-motion cells.
Movshon et al. (1985) have shown that
about a quarter of MT cells are “pattern-motion selective”; that
is, they respond to the motions of whole patterns independent
of the motions of the constituent contours. We recently reported
(Rodman and Albright, 1989) that these “pattern-motion se-
lective” cells make up the same population as our “Type II”
SC LESION
F 100
f
1
k
50
5
i?
% 0 I
NORMAL
CONTROL
N=103
-
s
N=40
S = STRONG RESPONSE
W = WEAK RESPONSE
d = NO RESPONSE
Fig.
2. Incidence of responses judged as strong, weak, or absent in MT
after an ipsilateral SC lesion, and in normal MT.
Hatching
in diagram
above graph represents lesion zone resulting from SC destruction rel-
ative to receptive fields in MT.
MT cells, defined as cells whose orientation preference is parallel
to their preferred direction of motion (Albright, 1984). In the
present study, of 24 MT neurons in the SC lesion animals, 6
were classified as Type II. This proportion is similar to that
found in normal MT (Rodman and Albright, 1989). Thus, SC
lesions do not appear to alter the incidence of MT cells sensitive
to the motion of whole patterns as opposed to only individual
contours.
Effects of combined striate cortex-SC lesions
A total of 16 1 recording sites (89 single units and 72 multiunits)
were studied on 22 penetrations into MT in the right hemi-
spheres of the 2 animals following the combined striate cortex-
SC lesions. All of the recording sites
were
determined to lie
within the myleoarchitectonic borders of area MT upon recon-
struction of the electrode tracks.
Histological findings
Striate cortex lesions.
Reconstructions of the striate cortex le-
sions and ensuing lateral geniculate degeneration are shown in
Rodman et al. (1989). In case no. 555, striate cortex on the
dorsolateral surface of the hemisphere was removed bilaterally,
resulting in a roughly circular estimated field defect (see Fig. 9,
left), which extended about 4” into the upper field and 6”-8” into
the lower field and along the horizontal meridian. In case no.
56 1, virtually all of striate cortex was removed bilaterally from
the dorsolateral and medial surfaces and from within the cal-
1158 Rodman
et al. *
MT After SC Removal
MOVING SLIT
SC LESION NORMAL
20
N = 53
II z = 25.6
i SD = 22.3
g 10
5
8
ii
0 I
0 10 20 30 40 50 60 >60
RESPONSE MAGNITUDE
SC LESION NORMAL
y = 99
x = 19.9
SD = 21.0
0 10 20 30 40 50 60 >60
RESPONSE MAGNITUDE
MOVING SPOT
Figure
3. Distributions of differential
response magnitude (algebraic differ-
ence between the largest and smallest
deviation from baseline) for moving slits
(top) and moving spots (bottom) for cells
in MT after an SC lesion. In Figures 3-
7, normal MT data replotted from Al-
bright (1984) are shown for compari-
son. Dimensions and speed of motion
of the stimuli used for each cell were
those estimated optimal with hand test-
ing.
0 10 20 30 40 50 60 >60 0 10 20 30 40 50 60 >60
RESPONSE MAGNITUDE RESPONSE MAGNITUDE
carine sulcus. The estimated field defect was thus total for at
least the central 60” (Fig. 9, right). Dorsally, the lesion extended
slightly onto the posterior bank of the lunate sulcus and thus
into V2 in each hemisphere. Ventrally, extrastriate damage was
more extensive in the right hemisphere, involving portions of
V2, V3, and cortex anteroventral to V3. Less marked involve-
ment of the same areas was present in the left hemisphere.
SC lesions. Parasagittal sections through the SC lesion for
case nos. 555 and 561 are shown in Figure 1. Sections from the
left hemisphere of no. 56 1, in which no SC lesion was made,
show the location of the SC and surrounding structures in the
intact midbrain. Removal of the SC of the right hemisphere of
each animal was nearly complete, sparing only a small region
(sections C and D) at the posterior pole [where the far periphery
is represented (Cynader and Berman, 1972)] in case no. 56 1. In
the left hemisphere of no. 555, a larger portion of the posterior
pole was spared. (However, no MT recordings were made in
this hemisphere subsequent to the SC lesions.) In all 3 hemi-
spheres with SC lesions in these 2 cases, there was some damage
to the medial pulvinar and pretectum, as well as slight damage
to the inferior colliculus.
Responsiveness
Single and multiunits in the combined lesion cases were divided
into 3 categories on the basis of their RF locations relative to
the estimated defect from the striate cortex lesion (see Fig. 9).
SC LESION NORMAL
N = 29
R=10.0
SD = 6.3
Figure
4. Distributions of differential
response magnitude measure for sta-
tionary slit stimuli. See also legend to
Figure 3.
B ‘”
P
o-
0 10 20 30 40 50 60 >60
RESPONSE MAGNITUDE
60
I:
d 60
0
k
40
5
N = 59
x = 5.8
SD = 7.8
0 1,. -
0 10 20 30 40 50 60 > 60
RESPONSE MAGNITUDE
The Journal of Neuroscience, April 1990, IO(4) 1159
MOVING SLIT
SC LESION NORMAL
0 .4 .a 1.2 1.6 2.0 >2.0 0 .4 .s 1.2 1.6 2.0 >2.0
DIRECTION INDEX DIRECTION INDEX
MOVING SPOT
SC LESION NORMAL
jJi;i; ; LjJ,:’
0 .4 .a 1.2 1.6 2.0 > 2.0 0 .4 .s 1.2 1.6 2.0 > 2.0
DIRECTION INDEX DIRECTION INDEX
Since the SC lesions were nearly total in the hemispheres re-
corded from, sparing only the far periphery (which part of the
representation we did not study in MT), all single and multiunits
in the sample were presumably devoid of their ipsilateral SC
input. Accordingly, units with RFs entirely within the overlap
of the lesion zones reflect the effects of removal of all of both
striate cortex and SC input; units with RFs partly within the
overlap of the lesion zones reflect the effects of removal of all
of their SC input and part of their striate cortex input; finally,
units with RFs categorized as outside the overlap reflect only
the effects of SC removal.
Since the striate cortex lesion for case no. 561 was virtually
total, all single and multiunits recorded in MT in this animal
fell, by definition, into the first category (entirely within the
overlap of lesion zones). Thus, only units from case no. 555 fell
into the second and third categories. For purposes of categori-
zation, RFs of unresponsive single and multiunits were esti-
mated on the basis of the fields plotted for the nearest responsive
units on the same penetration, and, on penetrations with no
responsive recording sites, on the basis of fields plotted on the
adjacent penetrations made prior to the lesions.
Figure 10 illustrates the incidence of strong, weak, and absent
responses in each of the RF categories following the combined
striate cortex-SC lesions. Within the portion of MT visuotop-
ically correspondent with the overlap of the striate cortex and
SC lesions, responses were virtually absent: only 3 recording
sites out of 99 tested gave even a weak response. Moreover, the
background (spontaneous) activity appeared abnormal: fre-
Figure
5. Distributions of direction
index values for stimulation with mov-
ing slits (top) and moving spots (bot-
tom), where DI = 1 - (response to mo-
tion in antipreferred direction/response
to motion in preferred direction). See
also legend to Figure 3.
quently there was no clear cellular activity, or only a faint,
rhythmic “swishing” that could be heard from the audio am-
plifier. Well outside the portion of the representation covered
by both lesions (i.e., from which only the SC input had been
removed) both the overall sound of the cortex and responses
were quite normal: 17 recording sites studied in this group all
were responsive, and the distribution of strong, weak, and absent
responses did not differ significantly from either the normal
sample or the sample of single MT units studied in the cases
with SC lesions alone. The distribution of strong, weak, and
absent responses for single and multiunits with RFs only partly
within the overlap of the lesion zones was intermediate between
the distributions for the other postlesion categories.
We did not quantitatively assess direction selectivity and bin-
ocularity after the combined lesion. However, in the course of
testing responsiveness and plotting RFs in the portions of MT
from which only SC input had been completely removed (i.e.,
where responsive units were still found), we noticed no marked
change in the strength or sharpness of direction selectivity. Sev-
en single units with RFs outside the overlap of the lesion zone
tested for responsiveness through both eyes showed no clear eye
dominance. These observations are consistent with the results
reported in the section on the SC lesion cases.
Discussion
The results presented here show that destruction of the SC alone
produces only minor alterations in the characteristic properties
of neurons in visual area MT. Although there is a tendency for
1160 Rodman et al. - MT After SC Removal
MOVING SLIT
SC LESION
NORMAL
30
1
N
= 44 30
1
N
=
90
07 !I =
96
03
x=91
5, ,h=I ; ;:kD;
0 40 60 120 160 >I60 0 40 60 120 160 >160
TUNING BANDWIDTH TUNING BANDWIDTH
MOVING SPOT
SC LESION NORMAL
Figure 6.
Distributions of direction
tuning bandwidth values for stimula-
tion with moving slits (top) and moving
spots (bottom), where bandwidth is de-
fined as the width of the direction tun-
inn curve at half maximum height. See
also legend to Figure 3. -
CJ~i; cJ,,Ilu;:
0 40 60 120 160 >160 0 40 60 120 160 >160
TUNING BANDWIDTH TUNING BANDWIDTH
increased responsiveness to slit stimuli, there are no effects on
direction selectivity or tuning, orientation tuning, RF size and
other properties. As in the normal monkey, virtually all MT
neurons in monkeys with SC lesions alone respond to visual
stimulation. In contrast, adding an SC lesion to striate removal
obliterates the considerable responsiveness remaining in MT
after a striate cortex lesion.
to the effects of SC damage alone and in combination with a
striate lesion. Finally, we will discuss implications of the findings
in regard to the nature of visual area MT and known visual
system organization.
Effects of SC lesions alone on properties of cells in visual
regions
In this section, we will first discuss the effects of SC lesions In the present study, SC lesions alone caused only slight changes
alone on response properties in the monkey visual system. Sec- in response magnitude to slit stimuli while leaving directional
ond, we will contrast the effects on neural responses and on properties, RF size, incidence of responsiveness, and other char-
behavior that result from damage to the SC with the effects of acteristic MT properties intact. In the inferior pulvinar of the
combining damage to the striate cortex and damage to the SC. monkey, a region that is both a recipient of SC input and a
Third, we will consider in some detail the various pathways source of projections to MT (Harting et al., 1980; Benevento
potentially interrupted by the SC lesion which may contribute and Standage, 1983; Maunsell and Van Essen, 1983~; Standage
SC LESION NORMAL
Figure 7.
Distributions of orientation
tuning bandwidth values for stationary
slit stimuli. See also legend to Figure 3.
0
40 60 120 160 >160 0 40 80 120 160 >I60
TUNING BANDWIDTH TUNING BANDWIDTH
SC LESION
The Journal of Neuroscience, April 1990, fO(4) 1161
NORMAL MT
25
1
N=
514
0-
25
ECCENTRICITY (deg.) ECCENTRICITY (deg.)
Figure
8. Plots of RF size (square root
of RF area) as a function of eccentricity
for cells in MT after an SC lesion and
in normal MT (replotted from Albright
and Desimone, 1987). Straight lines are
regressions fitted to the data by the
method of least squares. Slopes of
regression lines: SC lesion, 0.5 1; nor-
mal MT, 0.6 1. Intercepts of regression
lines: SC lesion, 1.77; normal MT, 1.04.
and Benevento,
1983),
Bender
(1983) found similarly weak ef-
fects of SC lesions alone. Direction and orientation selectivity,
RF size, and binocularity of inferior pulvinar cells were all un-
affected by SC lesions. A barely significant increase in the percent
of unresponsive cells was found (13% vs 7% in normal animals),
along with an increase in the proportion of cells giving sustained
responses to stationary flashed stimuli.
Bruce et al. (1986) have examined the effects of SC removal
on the superior temporal polysensory area (STP) of the monkey,
a primarily visual extrastriate cortical region which may receive
input multisynaptically from MT via area MST (Boussaoud et
al., 1987; C. Bruce, C. G. Gross, and R. Desimone, unpublished
observations). The effects of SC lesions alone on STP resemble
those found for MT in several respects: directional properties
of STP neurons were not affected by the SC lesions, and all
neurons continued to respond to stimuli in the hemifield con-
tralateral to the lesion. However, there were some minor dif-
ferences: STP neurons became slightly more difficult to drive
regardless of stimulus location or modality, and a small increase
in the percent of completely unresponsive neurons was found
(17% vs 2% in intact animals). There were also slight decrements
in RF size, notably a shrinking of the contralateral borders be-
yond about 60” in the periphery. Even these weak effects seem
more pronounced than those seen in MT and the pulvinar; STP
may be relatively more dependent on the midbrain (including
the deep, multimodal layers of the SC) for the excitability of its
neurons. Moreover, STP emphasizes the peripheral visual field
much more than does area MT, and the portions of STP RFs
that were affected by SC lesions were in the same part of the
visual field in which detection of visual stimuli is disrupted by
SC damage, namely, beyond about 45” (Butter et al., 1978).
Since all the cells in the present study had RFs within the central
30” of the visual field in MT, the possibility remains open that
SC lesions might produce more severe effects in more peripheral
MT or in parts of the caudal superior temporal sulcus repre-
senting more peripheral parts of the visual field, such as MT,
(Desimone and Ungerleider, 1986).
Efects of adding SC lesions to striate cortex damage
In the present study, we found that adding an SC lesion to striate
cortex damage abolished the residual responsiveness found in
MT after damage to the striate cortex alone. However, SC lesions
alone had only minor effects on MT responses. A similar pattern
of results was found for area STP by Bruce et al. (1986): the
addition of an SC lesion to existing striate cortex damage com-
pletely obliterated neuronal responses in this area, although SC
lesions alone produced relatively minor changes, as described
above. These sets of findings indicate that pathways damaged
by the SC lesion are sufficient, but not necessary, to sustain
visual responsiveness in at least 2 areas in the primate extrastri-
ate cortex.
Lesion-behavior studies have demonstrated relatively mild
effects of SC lesions in primates without concomitant striate
damage. SC lesions alone produce a number of small changes
in eye movements, including increased latency of saccade ini-
tiation, increased number of corrective saccades needed to fixate
a target, and decrease in spontaneous saccade frequency and
velocity (Mohler and Wurtz, 1977; Schiller et al., 1980, 1987).
SC lesions in primates also produce decreases in responsiveness
to peripheral stimuli (e.g., Butter et al., 1978; Albano et al.,
1982) and impairments in visual search (Bender and Butter,
no. 555 no. 561
“M “M
Figure 9. Estimated field defects resulting from striate cortex lesions
in case no. 555 and no. 561. In no. 555, in which only the dorsolateral
surface of the occipital lobe was targeted, the field defect was small,
central, and roughly circular. In no. 56 1, in which we aimed to remove
all of striate on the dorsolateral and medial surfaces and in the calcarine
fissure, the field defect was total for at least the central 60”. Sections
through the striate lesions and resultant LGN degeneration are shown
in Rodman et al. (1989).
1162 Rcdman et al.
l
MT After SC Removal
Figure 10. Incidence of responses
judged as strong, weak, or absent in MT
following combined striate cortex-SC
lesions and in normal MT. In the sche-
matic above each receptive field cate-
gory, the shaded urea represents the
striate zone and the hatching, the lesion
zone resulting from SC destruction. Lo-
cations of RFs of unresponsive units
were estimated on the basis of those
RF IN
OVERLAP OF
LESION ZONES
00
50
0
i
1
SW
0
N=99
RF PARTLY IN RF OUTSIDE
OVERLAP OF OVERLAP OF
LESION ZONES LESION ZONES
/
.:.:.:.:.:.:.:..
. . . . . . . . . . . . . . .
~ ., ..:
s wa
N=45
plotted on adjacent prelesion penetra-
tions. Compare with Figure 2.
S = STRONG RESPONSE W = WEAK RESPONSE 0= NO RESPONSE
N’17
NORMAL
CONTROL
ii S
-l
w0
N=40
1987). Although significant, these effects are relatively small
compared to the effects of SC removal subsequent to striate
ablation; the SC lesion then abolishes the recovery of visually
guided behavior seen following the striate lesion in both central
and peripheral parts ofthe visual field (Mohler and Wurtz, 1977;
Solomon et al., 198 1). Thus, there is a parallel between the
devastating effects of combined striate-SC damage on visual
cortical responses and on visual behavior.
Pathways potentially damaged by SC lesions
It is well established that in species with geniculocortical systems
less well developed than those of the higher primates, the tec-
tofugal pathways play a dominant role in visual processing (e.g.,
Trevarthen, 1968; Karten, 1979; Diamond et al., 1985). Because
of this fact, and because of the strong, reciprocal connections
of area MT with the inferior and lateral pulvinar, we have sug-
gested (Rodman et al., 1986, 1989) that the residual responses
found in MT after striate removal derive from SC inputs via
the pulvinar. In support of this notion, it is well known that the
superficial layers of the SC receive direct retinal input (Hubel
et al., 1975; Pollack and Hickey, 1979) and these regions project
to the inferior and lateral subdivisions of the pulvinar (Beneven-
to and Fallon, 1975; Harting et al., 1980; Benevento and Stan-
dage, 1983). A small proportion of neurons in both the inferior
and lateral pulvinar remain responsive after striate cortex re-
moval (Burman et al., 1982; Bender, 1983). A pathway from
the retina to MT passing through only the SC and its pulvinar
targets via a minimal number of synapses is thus quite plausible.
However, the portions of the inferior and lateral pulvinar that
project to MT may not be the same subdivisions that receive
SC input (see Kaas and Huerta, 1988). Moreover, there are
several other potential routes through the midbrain and dien-
cephalon which may be affected by SC damage and which may
contribute to the effects of such damage on MT. These routes
are summarized in the Appendix. In particular, the pretectum
and the dorsal lateral geniculate nucleus (LGN) may play a role
in the effects of SC lesions on MT responses and on visual
behavior.
Potential role of the pretectum.
Lesions of the SC often pro-
duce direct damage to the pretectal region, which lies just an-
terior to the fovea1 representation in the SC. Some pretectal
damage was apparent in each of the cases in the present study
(Fig. 1). Efferents of the pretectum go to a number of subcortical
visual structures, including the SC, inferior and lateral pulvinar,
and LGN (Carpenter and Pierson, 1973; Benevento et al., 1977;
Benevento and Standage, 1983). Moreover, since some of these
efferents pass through the brachium of the SC (Carpenter and
Pierson, 1973), even lesions limited to the SC might interrupt
axons en route to the pulvinar and LGN from the pretectum
(Benevento and Fallon, 1975). A role for the pretectum in the
visual behavior that survives striate cortex damage is suggested
by the work of Pasik and Pasik (1973), who found that monkeys
failed to relearn a light-no light discrimination after striate dam-
age only when additional midbrain lesions were large and in-
cluded extensive pretectal damage. Thus, it is not possible to
rule out a role for the pretectum in the effects of SC damage on
MT.
Potential role
of
the LGN.
Although MT itself does not receive
projections from any portion of the LGN (Standage and Be-
nevento, 198 1; Maunsell and Van Essen, 1983c), there are weak
projections from primarily the interlaminar and magnocellular
zones to extrastriate areas V4 and possibly V3 (Benevento and
Yoshida, 198 1; Yukie and Iwai, 1981); both these extrastriate
areas project to MT (Maunsell and Van Essen, 1983~; Unger-
leider and Desimone, 1986). This LGN-extrastriate pathway
remains intact years after striate cortex removal (Cooper and
Cowey, 1988). Two additional pieces of evidence suggest that
the LGN may play a role in MT responses and in visual behavior
above and beyond that serving as a relay from the retina to
striate cortex. First, reversible inactivation of the magnocellular
layers of the LGN produces a nearly total cessation of activity
in retinotopically corresponding parts of MT (Maunsell et al.,
The Journal of Neuroscience, April 1990, 70(4) 1163
1989) whereas reversible inactivation of striate cortex does not
(Rodman et al., 1985, 1989; Bullier and Girard, 1988). Second,
lesions of the LGN abolish the visually guided eye movements
that persist after striate cortex lesions (Schiller et al., 1985).
Whatever role the LGN may have in the MT responses that
survive striate cortex damage is obviously undermined by the
SC lesion, since this lesion completely abolishes the residual
responsiveness. In this regard, it is interesting that intraocular
injections of amino acids label all projection structures of the
LGN except extrastriate cortex (Benevento and Yoshida, 198 l),
suggesting that the LGN cells which project to extrastriate do
not receive direct retinal input; rather, Benevento and Yoshida
hypothesize that the LGN-extrastriate projection might be anal-
ogous to pulvinar-extrastriate systems which receive visual in-
puts from the midbrain and cortex. However, our 2 cases with
minimal (no. 555) and extensive (no. 561) extrastriate damage
attending striate cortex removal did not differ in sparing of MT
responsiveness, arguing against an important role for the LGN-
extrastriate route in the remaining responses (see Rodman et
al., 1989). Clearly, more work is needed to specify the role of
the LGN in the visual properties of cells in MT and other
extrastriate areas both in the normal monkey and in animals
with visual system damage.
Parallel pathways in cortex
The residual responsiveness in MT after a striate cortex lesion
is rather striking in view of the weakness of the pathways likely
to be involved in relaying visual information from the midbrain
to MT: as described in the previous section, only a small pro-
portion of the cells in the lateral and inferior pulvinar remain
visually responsive after the striate lesion, and none are direc-
tionally selective, and the LGN-extrastriate pathway is rela-
tively minor. Thus, MT appears able to generate a surprising
amount of neural function on the basis of rather weak signals
from pathways originating in the midbrain.
This relationship between MT and midbrain, moreover, may
reflect the role of MT in spatial functions. The notion that
anatomically separable pathways subserve spatial and pattern
vision in primates was introduced in the 1960s (Trevarthen,
1968). In the original formulation, spatial or “ambient” func-
tions, such as the detection and localization of brief visual stim-
uli, were attributed to a tectopulvinar system, whereas pattern
or “focal” vision, such as the identification of objects, was at-
tributed to geniculocortical pathways. We now know that there
is a degree of separation of pathways subserving spatial and
pattern vision at the cortical level (e.g., Ungerleider and Mish-
kin, 1982). In the current view, pattern vision is represented
cortically by a pathway that originates within certain subdivi-
sions of striate cortex, goes next to V4 via parts of V2, and then
courses ventrally to areas in the inferior temporal lobe, whereas
spatial vision is subserved by a pathway that originates in other
compartments of striate cortex and V2 and then courses dorsally
to parietal cortex via area MT and other zones within the su-
perior temporal sulcus. Bruce et al. (1986) have suggested that
while cortical areas involved in pattern vision (the ventral path-
way) may be dependent solely on information deriving from
geniculostriate projections, the SC may influence visual pro-
cessing within the dorsal or “spatial” pathway. Major compo-
nents of the dorsal, spatial vision system [namely, MT and
surrounding zones in the superior temporal sulcus (Rodman et
al., 1985, 1989; Bullier and Girard, 1988)] retain visual re-
sponsiveness in the absence of striate cortex input. Similarly,
area STP, which receives converging input from both the dorsal
and ventral cortical systems (Jones and Powell, 1970; Seltzer
and Pandya, 1978, Baizer et al., 1988) contains visually re-
sponsive cells after striate lesions (Bruce et al., 1986). On the
other hand, insofar as it has been examined, it appears that
components of the ventral, pattern vision system (namely, V2
and inferior temporal cortex) are completely dependent on striate
cortex for visual responsiveness (Rocha-Miranda et al., 1975;
Schiller and Malpeli, 1977; Bullier and Girard, 1988).
Appendix
Pathways that might provide visual input to MT and might be
disrupted by SC lesions:
1. Retina - SC + P,, P, + MT
2. Retina + pretectum - P,, P, --) MT
3. Retina 4 pretectum + SC + P,, P, - MT
4. Retina - SC + LGN + extrastriate - MT
5. Retina + pretectum - LGN + extrastriate - MT
See text for references. Abbreviations: P,, inferior pulvinar; P,, lateral
pulvinar; LGN, dorsal lateral geniculate nucleus.
References
Albano, J. E., M. Mishkin, L. E. Westbrook, and R. H. Wurtz (1982)
Visuomotor deficits following ablation of monkey superior colliculus.
J. Neurophysiol. 48: 338-35 1.
Albright, T. D. (1984) Direction and orientation selectivity of neurons
in visual area MT of the macaque. J. Neurophysiol. 52: 1106-l 130.
Albright, T. D., and R. Desimone (1987) Local precision of visuotopic
organization in the middle temporal area (MT) of the macaque. Exp.
Brain Res. 65: 582-592.
Albright, T. D., R. Desimone, and C. G. Gross (1984) Columnar
organization of directionally selective cells in visual area MT of the
macaque. J. Neurophysiol.
51:
16-3 1.
Allman, J. M., and J. H. Kaas (197 1) A representation of the visual
field in the caudal third of the middle temporal gyrus of the owl
monkey @lotus
trivirgutus).
Brain Res. 31: 85-105.
Allman, J. M., F. Miezen, and E. McGuinness (1985) Stimulus-specific
responses from beyond the classical receptive field: Neurophysiolog-
ical mechanisms for local-global comparisons in visual neurons. Annu.
Rev. Neurosci. 8: 407-430.
Baizer, J. S., L. G. Ungerleider, and R. Desimone (1988) Organization
of visual projections to parietal and temporal cortex in the macaque.
Neurosci. Abstr. 14: 11.
Bender, D. B. (1983) Visual activation of neurons in the primate
pulvinar depends on cortex but not colliculus. Brain Res. 279: 258-
261.
Bender, D. B., and C. M. Butter (1987) Comparison of the effects of
superior colliculus and pulvinar lesions on visual search and tachis-
toscopic pattern discrimination in monkeys. Exp. Brain Res. 69: 140-
154.
Benevento, L. A., and J. H. Fallon (1975) The ascending projections
of the superior colliculus in the rhesus monkey (Macaca
muluttu).
J.
Comp. Neurol. 160: 339-362.
Benevento, L. A., and G. P. Standage (1983) The organization of
projections of the retinorecipient and nonretinorecipient nuclei of the
pretectal complex and layers of the superior colliculus to the lateral
pulvinar and medial pulvinar in the macaque monkey. J. Comp.
Neurol. 271: 307-336.
Benevento, L. A., and K. Yoshida (1981) The afferent and efferent
organization of the lateral geniculo-prestriate pathways in the ma-
caque monkey. J. Comp. Neurol. 203: 455474.
Benevento, L. A., M. Rezak, and F. Santos-Anderson (1977) An au-
toradiographic study of the projections of the pretectum in the rhesus
monkey
(Mucucu muluttu):
Evidence for sensorimotor links to the
thalamus and oculomotor nuclei. Brain Res. 127: 197-2 18.
Boussaoud, D., L. G. Ungerleider, and R. Desimone (1987) Cortical
pathways for motion analysis: Connections of areas MST and FST
in macaques. Neurosci. Abstr. 13: 1625.
1164 Rodman et al. - MT After SC Removal
Bruce, C., R. Desimone, and C. G. Gross (1986) Both striate cortex
and the superior colliculus contribute to the visual properties of neu-
rons in the superior temporal polysensory area in the macaque. J.
Neurophysiol. 5: 1057-1075.
Bullier, J., and P. Girard (1988) Visual responses of neurons in area
V2 and in the superior temporal sulcus of the macaque monkey during
reversible inactivation of area V 1. Neurosci. Abstr. 14: 602.
Burman, D., G. Felsten, and L. Benevento (1982) Visual properties
of neurons in the lateral pulvinar of normal and occipital lobecto-
mized macaques. Assoc. Res. Vis. Ophthal. Suppl. 22: 237.
Butter, C. M., C. Weinstein, D. B. Bender, and C. G. Gross (1978)
Localization and detection of visual stimuli following superior col-
liculus lesions in rhesus monkeys. Brain Res. 156: 3349.
Carpenter, M. B. and R. J. Pierson (1973) Pretectal region and the
pupillary light reflex: An anatomical analysis in the monkey. J. Comp.
Neurol. 149: 27 l-300.
Cooper, A. M., and A. Cowey (1988) A geniculo-prestriate projection
in long-term destriate macaque monkeys. Neurosci. Abstr. 14: 1123.
Cynader, M. and N. Berman (1972) Receptive-field organization of
monkey superior colliculus. J. Neurophysiol. 35: 187-20 1.
Desimone, R., and L. G. Ungerleider (1986) Multiple visual areas in
the caudal superior temporal sulcus of the macaque. J. Comp. Neurol.
248. 164-189.
Diamond, I. T., D. Fitzpatrick, and J. M. Sprague (1985) The extrastri-
ate visual cortex: a historical approach to the relation between “visuo-
sensory” and “visuo-psychic” areas. In Cerebral Cortex, Vol. 4, A.
Peters and E. G. Jones, eds., New York.
Dubner, R., and S. M. Z&i (197 1) Response properties and receptive
fields of cells in an anatomically defined region of the superior tem-
poral sulcus in the monkey. Brain Res. 35: 528-532.
Gallyas, F. (1969) Silver staining of myelin by means of physical
development. Orvostucomany 20: 433-489.
Gattass, R. G., and C. G. Gross (198 1) Visual topography of the striate
projection zone in the posterior superior temporal sulcus (MT) of the
macaque. J. Neurophysiol. 46: 62 I-638.
Harting, J. K., M. F. Huerta, A. J. Frankfurter, N. L. Strominger, and
G. J. Royce (1980) Ascending pathways from the monkey superior
colliculus: An autoradiographic analysis. J. Comp. Neurol. 192: 853-
882.
Hubel, D. H., and T. N. Wiesel (1962) Receptive fields, binocular
interaction and functional architecture in the cat’s visual cortex. J.
Physiol. 160: 106-l 54.
Hubel, D. H., S. LeVay, and T. N. Wiesel (1975) Mode oftermination
of retinotectal fibers in macaque monkey: An autoradiographic study.
Brain Res. 96: 25-40.
Jones, E. G., and T. P. S. Powell (1970) An anatomical study of
converging sensory pathways within the cerebral cortex of the ma-
- -
caque. Brain 93: 7-98-820. -
Kaas. J. H.. and J. F. Huerta (1988) The subcortical visual svstem of
primates: In Comparative P&ate Biology Vol. 4, H. Steklis and J.
Erwin, eds., Liss, New York.
Karten, H. J. (1979) Visual lemniscal pathways in birds. In Neural
Mechanisms
of
Behavior in the Pigeon, A. M. Granda and J. H.
Maxwell, eds.,.Plenum, New York. -
Maunsell. J. H. R.. and D. C. Van Essen (1983a) Functional uronerties
of neurons in middle temporal visual irea oi the macaque mbnkey.
I. Selectivity for stimulus direction, speed and orientation. J. Neu-
rophysiol. 49: 1127-I 147.
Maunsell, J. H. R., and D. C. Van Essen (1983b) Functional properties
of neurons in middle temporal visual area of the macaque monkey.
II. Binocular interactions and sensitivity to binocular disparity. J.
Neurophysiol. 49: 1148-l 167.
Maunsell, J. H. R., and D. C. Van Essen (1983~) The connections of
the middle temporal visual area (MT) and their relationship to a
cortical hierarchy in the macaque monkey. J. Neurosci. 3: 2563-2586.
Maunsell, J. H. R., D. D. DePriest, and T. A. Nealy (1989) The middle
temporal visual area receives excitatory drive primarily via the mag-
nocellular layers of the LGN. Assoc. Res. Vision Ophthal. Suppl. 30:
427.
Mohler, C. W., and R. H. Wurtz (1977) Role of striate cortex and
superior colliculus in visual guidance of saccadic eye movements in
monkeys. J. Neurophysiol. 40: 74-94.
Movshon, J. A., E. H. Adelson, M. S. Gizzi, and W. T. Newsome (1985)
The analysis of moving visual patterns. In Study Group on Pattern
Recognition Mechanisms, C. Chagas, R. G. Gattass, and C. G. Gross,
eds., Pontitica Academia Scientiarum, Vatican City.
Pasik, T., and P. Pasik (1973) Extrageniculostriate vision in the mon-
key. IV. Critical structures for light vs. no-light discrimination. Brain
Res. 56: 165-182.
Pollack, J. G., and T. L. Hickey (1979) The distribution of retino-
collicular axon terminals in rhesus monkey. J. Comp. Neurol. 185:
587-602.
Rocha-Miranda, C., D. Bender, C. G. Gross, and M. Mishkin (1975)
Visual activation of neurons in inferotemporal cortex depends on
striate cortex and forebrain commissures. J. Neurophysiol. 38: 475-
491.
Rodman, H. R., and T. D. Albright (1987) Coding of visual stimulus
velocity in area MT of the macaque. Vis. Res. 27: 2035-2048.
Rodman, H. R., and T. D. Albright (1989) Single-unit analysis of
pattern-motion selective properties in the middle temporal visual area
(MT). Exp. Brain Res. 75: 53-64.
Rodman, H. R., C. G. Gross, and T. D. Albright (1985) Removal of
striate cortex does not abolish responsiveness of neurons in visual
area MT of the macaque. Neurosci. Abstr. II: 1246.
Rodman, H. R., C. G. Gross, and T. D. Albright (1986) Responses
of neurons in visual area MT after removal of the superior colliculus.
Neurosci. Abstr. 12: 1369.
Rodman, H. R., C. G. Gross, and T. D. Albright (1989) Afferent basis
of visual response properties in area MT of the macaque: I. Effects of
striate cortex removal. J. Neurosci. 9: 2033-2050.
Schiller, P. H., and J. G. Malpeli (1977) The effect of striate cortex
cooling on area 18 cells in the monkey. Brain Res. 126: 366-369.
Schiller, P. H., S. True, and J. Conway (1980) Deficits in eye move-
ments following frontal eye field and superior colliculus ablations. J.
Neurophysiol. 44: 1175-l 189.
Schiller, P. H., J. H. R. Maunsell, and J. G. Malpeli (1985) The effect
of lateral geniculate nucleus lesions on the detection of visual stimuli.
Invest. Ophthalmol. Vis. Sci. Suppl. 26: 195.
Schiller, P. H., J. H. Sandell, and J. H. R. Maunsell (1987) The effect
of frontal eye field and superior colliculus lesions on saccadic latencies
in the rhesus monkey. J. Neurophysiol. 57: 1033-1049.
Seltzer, B. and D. N. Pandya (1978) Afferent cortical connections and
architectonics of the superior temporal sulcus and surrounding cortex
in the rhesus monkey. Brain Res. 149: l-24.
Solomon, S. J., T. Pasik, and P. Pasik (198 1) Extrageniculostriate
vision in the monkey. VIII. Critical structures for spatial localization.
Exp. Brain Res. 44: 259-270.
Standage, G. P., and L. A. Benevento (1981) Lack of dorsal lateral
geniculate input to extrastriate cortical area MT and dorsal visual 2
in the macaque monkey. Neurosci. Abstr. 8: 680.
Standage, G. P., and L. A. Benevento (1983) The organization of
connections between the pulvinar and visual area MT in the macaque
monkey. Brain Res. 262: 288-294.
Trevarthen, C. (1968) Two mechanisms of vision in primates. Psychol.
Forsch. 31: 299-337.
Ungerleider, L. G., and R. Desimone (1986) Cortical connections of
visual area MT in the macaque. J. Comp. Neurol. 248: 190-222.
Ungerleider, L. G., and M. Mishkin (1982) Two cortical visual sys-
terns. In Analysis of Visual Behavior,.D. J.‘Ingle, M. A. Goodale, aid
R. J. W. Mansfeld. eds.. MIT Press. Cambridge. MA.
Ungerleider, L. G., RI Desimone, T. W: Galkin, aid M. Mishkin (1984)
Subcortical projections of area MT in the macaque. J. Comp. Neural.
223: 368-386.
Van Essen, D. C., J. H. R. Maunsell, and J. L. Bixby (198 1) The
middle temporal visual area in the macaque: Myeloarchitecture, con-
nections, functional properties and topographic organization. J. Comp.
Neural. 199: 293-326.
Yukie, M. and E. Iwai (198 1) Direct projection from dorsal lateral
geniculate nucleus to the prestriate cortex in macaque monkey. J.
Comp. Neurol. 201: 81-98.
Zeki, S. M. (1974) Functional organization of a visual area in the
posterior bank of the superior temporal sulcus of the rhesus monkey.
J. Physiol. 236: 549-573.
