The Journal of Neurophysiology Vol. 77 No. 2 February 1997, pp. 587-598
Copyright ©1997 by the American Physiological Society

Functional Double Dissociation Between Two Inferior Temporal Cortical Areas: Perirhinal Cortex Versus Middle Temporal Gyrus

M. J. Buckley¹, D. Gaffan¹, and E. A. Murray²

¹ Department of Experimental Psychology, Oxford University, Oxford OX1 3UD, United Kingdom; and ² Laboratory of Neuropsychology, National Institute of Mental Health, Bethesda, Maryland 20892

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Buckley, M. J., D. Gaffan, and E. A. Murray. Functional double dissociation between two inferior temporal cortical areas: perirhinal cortex versus middle temporal gyrus. J. Neurophysiol. 77: 587-598, 1997. There is both anatomic and cytoarchitectural evidence for dorsal-ventral subdivisions of the inferior temporal cortex. Despite this, there has been only limited evidence of corresponding functional subdivisions and no evidence that two adjacent cortical areas within the inferior temporal cortex, namely area TE and the perirhinal cortex, have distinctly different roles in vision and memory. We assessed the color discrimination abilities of cynomolgus monkeys with either bilateral ablation of the perirhinal cortex or bilateral ablation of the middle temporal gyrus. The stimuli were isoluminant colored squares presented on a touch screen. In each trial the subject had to learn to discriminate and select the correct choice (green) from among a maximum of eight other foils, each varying in either hue or saturation. Relative to unoperated controls, monkeys with middle temporal gyrus lesions were severely impaired in the color discrimination task, whereas monkeys with perirhinal lesions were unimpaired on this task. We also assessed the visual recognition abilities, as measured by a basic delayed nonmatching-to-sample task with trial-unique objects presented in a Wisconsin General Test Apparatus, of rhesus monkeys with bilateral middle temporal gyrus lesions. We then tested the monkeys' postoperative performance on a delayed nonmatching-to-sample task with delays and extended list lengths. The results from this experiment were compared with those from two other groups of rhesus monkeys, an unoperated control group and a group with bilateral perirhinal cortex lesions, both of which had performed the identical tasks in a previous experiment. Relative to unoperated controls, monkeys with perirhinal cortex lesions were severely impaired both in relearning the basic delayed nonmatching-to-sample task and on the postoperative performance test. In contrast, monkeys with middle temporal gyrus lesions were only mildly affected in relearning the basic nonmatching task and were unimpaired on the postoperative performance test. Thus our data demonstrate a clear functional double dissociation between the perirhinal cortex and the middle temporal gyrus. This result gives strong support to the hypothesis that the perirhinal cortex and the adjacent area TE have distinctly different roles in visual learning and memory.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Anatomic, ablation, and physiological evidence all suggest that the neuronal mechanisms that connect vision and memory in primates are located within the inferior temporal cortex (IT) of the temporal lobe. Anatomically, IT consists of the middle temporal gyrus (MTG) dorsally and the inferior temporal gyrus (ITG) ventrally, which corresponds largely to Brodmann's areas 21 and 20, respectively (Brodmann 1905). The MTG and ITG have been established to be the visual part of the temporal lobe (see Gross 1994 for a review). Recently it has been shown that within the medial temporal lobe it is the perirhinal cortex, part of the ITG, that is critical for stimulus recognition memory rather than the hippocampus and amygdala as had previously been believed (Meunier et al. 1993). Therefore much interest has focused on the perirhinal cortex to try to elucidate its function and to discover how this may differ from the rest of the cortex in IT.

von Bonin and Bailey (1947) considered the anterior part of the MTG and ITG to be part of one cytoarchitectural division in Macaca mulatta, and they labeled this area TE. They also identified a more caudal area, TEO. On the basis of observations of cell morphology, myelination patterns, and connectivity, Seltzer and Pandya (1978) discerned five different subdivisions of von Bonin and Bailey's area TE of M. mulatta; these five areas were named TE1, TE2, TE3, TEm, and TEa. These areas are roughly parallel to the gyri but they do not respect the middle temporal sulcus, although TE1 is primarily in ITG and TE3 and TEm are primarily in MTG. However, on the basis of the distributions and densities of amygdalar connections with MTG and ITG, Iwai and Yukie (Iwai and Yukie 1987; Iwai et al. 1987) divided areas TE and TEO of the Japanese monkey into dorsal and ventral subdivisions bordered by the anterior middle temporal sulcus. There is further evidence for dorsal-ventral subdivision of areas TE and TEO. Barbas (1985) injected the orbitofrontal cortex with horseradish peroxidase and later found labelled cells in the superior temporal sulcus, the ventral surface of the temporal pole, and ITG, but not in MTG. Van Essen et al. (1990) used retrograde tracer injections to determine patterns of connectivity of IT with V4. They suggested that IT of Macaca fascicularis could be divided into six regions: dorsal and ventral subdivisions of the posterior inferotemporal area, namely PITd and PITv; dorsal and ventral subdivisions of a more anterior central inferotemporal area, namely CITd and CITv; and dorsal and ventral subdivisions of an anterior temporal region, namely AITd and AITv. Finally, although ITG was previously thought to be composed of cortex designated as TE (von Bonin and Bailey 1947) or area 20 (Brodmann 1909), it now appears on the basis of connectional grounds that the lateral boundary of perirhinal cortex (Brodmann's areas 35 and 36) may be located more laterally than previously believed (Amaral et al. 1987; Insausti et al. 1987; Suzuki et al. 1993), perhaps near the ITG-MTG boundary (i.e., the fundus of the anterior middle temporal sulcus).

There is some functional evidence of subdivisions within IT that correspond to these dorsal-ventral cytoarchitectonic and connectional subdivisions. Horel et al. (1987) examined the effects of suppressing different segments of the inferotemporal cortex of M. fascicularis by cold. They found that cooling ITG produced a deficit in delayed matching to sample, whereas cooling MTG did not, and they reproduced the same effects after MTG or ITG ablation. Horel (1994b) also showed that suppressing the dorsal aspect of the inferotemporal cortex by cooling disrupted the retrieval of color discriminations but not the retrieval of form discriminations.

The perirhinal cortex lies in the anterior medial part of the ITG. The perirhinal cortex is made up of areas 35 and 36 and is situated in the lateral bank of the rhinal sulcus and in the cortex laterally adjacent to it; however, the recognized extent of the perirhinal cortex differs slightly between species and across investigators (Amaral et al. 1987; Brodmann 1909; Insausti et al. 1987). The entorhinal cortex is situated medial to the rhinal sulcus, including the medial bank of the sulcus. It is uniquely defined by its robust layer II projection to the molecular layer of the dentate gyrus (Witter et al. 1989) and includes the areas designated as prorhinal and entorhinal cortices by Van Hoesen and Pandya (1975). Together the perirhinal and entorhinal cortex can be called, for brevity, the rhinal cortex.

Murray et al. (1989) showed that a severe impairment in delayed nonmatching to sample (DNMS) was produced by ablation of the rhinal cortex. However, as well as being involved in recognition memory, the rhinal cortex has also been implicated in associative memory as evidenced by the effect of rhinal cortex lesions on the formation of stimulus-stimulus associative memories (Murray et al. 1993). Meunier et al. (1993) showed that damage to the perirhinal cortex alone produced a deficit in recognition memory nearly as severe as that found after rhinal cortex lesions, whereas damage to the entorhinal cortex alone produced only a mild deficit. It was demonstrated not only that damage limited to the perirhinal cortex was sufficient to produce a severe loss in visual recognition, but also that such damage leads to a far greater loss than damage to any other single structure within the medial part of the temporal lobe. Gaffan (1994) provided further evidence that a severe impairment in visual recognition memory follows ablations restricted to the perirhinal cortex, and also showed that the effects of these ablations can be doubly dissociated from the effects of fornix transection. After a series of studies, Eacott et al. (1994) concluded that ablation of the rhinal cortex did not produce an impairment in all forms of visual recognition memory, and only in visual recognition memory, but rather produced a general impairment in the capacity for knowledge about objects. Damage limited to the perirhinal cortex alone has now also been shown to produce impairments in visual object discrimination learning with 24-h intertrial intervals (Buckley and Gaffan, unpublished data). All these results taken together suggest that the rhinal cortex, and the perirhinal cortex in particular, forms the kernel of a system specialized for processing and storing knowledge about objects.

In contrast to the evidence that the perirhinal cortex is involved in knowledge about objects, there is evidence that the anterior cortex area of TE is more involved in color vision. As already mentioned, Horel (1994b) showed that cooling of MTG disrupts retrieval of color discriminations. Further, cerebral achromatopsia is a human clinical condition in which, after brain damage, there is severe impairment of color vision with relative sparing of nonchromatic vision. Heywood et al. (1995) showed that lesions made in the temporal lobe anterior to area V4, unlike lesions within V4, produced a similar achromatopsia in monkeys. In addition, there is evidence that this region is important for the processing of high-spatial-frequency information, as opposed to global form information (Horel 1994a).

Thus both TE and the perirhinal cortex have clearly been shown to be involved in aspects of visual learning and memory. However, as yet there has been no double dissociation of function between the perirhinal cortex and TE to demonstrate distinct functional differences between these two adjacent cortical areas. One hypothesis is that there are no distinct functional differences between TE and the perirhinal cortex and instead there is a continuity or gradation of function across the IT areas, perhaps with some areas having a greater functional specialization than others. With this hypothesis in mind, more extensive deficits following TE or MTG lesions than following perirhinal lesions could be explained by the fact that these lesions are much larger in area than perirhinal lesions. The alternative hypothesis is that the perirhinal lesion actually produces a distinctly different set of deficits than lesions to other anterior cortical areas within IT.

The experiments we report here were designed to investigate this issue further by looking at pre- and postoperative performance in two different visual tasks following bilateral lesions to one or other of two possible functional subdivisions of IT based on the anatomic and functional evidence we have reviewed, namely the MTG and the perirhinal cortex. The first task was color discrimination (experiment 1). The second task was DNMS with trial-unique objects, a task that tests visual object recognition memory (experiment 2).

	METHODS

Abstract Introduction Methods Results Discussion References

Experiment 1: color discrimination

SUBJECTS. Eleven male cynomolgus monkeys (M. fascicularis) served as subjects. They were housed either individually or in pairs in rooms with automatically regulated lighting and with water always freely available. The subjects were experimentally naive juveniles, apart from two subjects with previous experience as control subjects in object recognition tasks (Eacott et al. 1994). After preoperative training the subjects were assigned into three groups matched on the basis of preoperative learning scores. Three subjects received bilateral perirhinal cortex ablations (group PRh-A), three subjects received MTG ablations (group MTG-A), and five subjects remained unoperated controls (group CON-A). One control animal underwent the initial stages of surgery but the surgery was aborted before the brain itself was operated on because of irregular breathing. Of the two subjects that had experience in object recognition tasks, one was assigned to the control group and the other to the perirhinal group.

APPARATUS AND MATERIALS. The color discrimination pretraining and tasks were performed in an automated test apparatus similar to that used in a previous study on color vision in macaques recently carried out in the same laboratory (Heywood et al. 1995). The subject sat in a wheeled transport cage fixed in position in front of a touch sensitive screen (380 × 280 mm) on which the color stimulus patterns could be displayed. The subject could reach out between the horizontal or vertical bars (150 mm apart) at the front of the transport cage to touch the screen. An automated pellet delivery system controlled by the computer delivered reward pellets into a food well 80 mm diam positioned in front of and to the right of the subject. Reward pellets (190 mg) were only delivered in response to a correct choice made by the subject to the touch screen. Pellet delivery was accompanied by an audible click. An automated lunch box (length 200 mm, width 100 mm, height 100 mm) was positioned in front of and to the left of the subject. The lunch box was spring loaded and opened immediately with a loud crack only on completion of the whole session. The lunch box contained cereals, seeds, proprietary primate pellets, nuts, raisins, and half an apple or banana. An infrared camera was positioned looking down into the transport cage from above to allow the subject to be observed while engaged in the task. The whole apparatus was housed in an experimental cubicle that was dark apart from a 25-W incandescent lamp positioned on the floor below the level of the touch screen to avoid any reflection onto the screen yet still allow the subject to see into the cup and lunch box when the screen was dark. The presentation of the visual stimuli on the touch screen was controlled by a computer. The computer also recorded the responses the subjects made to the touch screen and controlled the delivery of reward pellets and the opening of the lunch box.

PREOPERATIVE TRAINING. Preliminary training. The animals were first accustomed to the apparatus and taught to touch patterns appearing on the screen for food reward as described in Gaffan et al. (1984).

Color discrimination. After pretraining the subjects commenced several stages of training with colored stimuli. The stimuli for this task were colored squares (80 × 80 mm) presented on a plain gray background on the touch screen. The stimuli varied in hue and saturation but were isoluminant as measured by a photometer. The positions in which the stimuli could appear took the form of a three-by-three array so that a maximum of nine stimuli could be presented at any one time. Adjacent stimuli were separated by 45 mm both horizontally and vertically. No two stimuli presented within any single trial were identical in both hue and saturation. In each trial, no matter how many stimuli were displayed, the positions of the stimuli were always randomized between the nine possible stimulus positions.

Green, referred to as square color 1, was the correct choice (S+) in every trial. The eight possible foils (S) were referred to as square colors 2-9. Square colors 1-6 gradated in hue from green through yellow and orange to red, and square colors 1, 7, 8, and 9 gradated in saturation from green to gray. The colors were judged to be roughly equally perceptually spaced.

Before each trial began there was an intertrial interval of 10 s. Any touch to the screen during the intertrial interval restarted the 10 s interval. After the intertrial interval the stimuli were presented. If the subject touched anywhere within the area of the green square, then a reward pellet was delivered with a loud click from the automatic pellet dispenser. The screen then blanked, apart from the green square itself, which remained on the screen for a further 1 s before the intertrial interval for the next trial commenced. Alternatively, if the subject touched within the area of any of the foils, then the whole screen immediately blanked, no reward pellet was dispensed, and a new intertrial interval was commenced before the next trial. If the subject touched elsewhere on the screen than on any of the stimuli, then all the stimuli remained until a correct or incorrect choice was made. After a specified number of correct choices had been made, the lunch box automatically opened and the food within became available. The subjects were left for 10-15 min to eat out of the lunch box. Each subject performed the task once per day.

On the 1st day of training, 25 correct responses were required to open the lunch box. On the 2nd day, 50 correct responses were required, and from the 3rd day onward, 100 correct responses were required. Thus in the full task in which 100 correct responses were required to open the lunch box, there were 100 correct responses in every session, with in addition a variable number of errors that were committed in the course of accumulating 100 correct responses. For example, if a subject made 11 errors, this was a 90% correct performance (100 correct in 111 trials).

In the first stage of training, the three most distinctly different stimuli were presented in each trial, one green S+ (square color 1), one red foil (square color 6), and one gray foil (square color 9). When the subject attained criterion of 90% correct responses within a session on a particular stage, then on the following day the subject progressed to the next stage of training. The second stage of training had three foils: red, gray, and an intermediate color between green and red that was closest to red (square colors 6, 9, and 5). The third stage of training had five foils (square colors 6, 9, 5, 4, and 8), so that the additional foil was a color intermediate in saturation between green and gray that was closest to gray. The fourth stage had six foils (square colors 6, 9, 5, 4, 8, and 3), and the final stage, which is the full task, had all eight foils present. Thus, as the difficulty level increased, the differences in hue and saturation between the S+ and those foils with the most similar hue and saturation decreased.

The subjects were required to attain the criterion of 90% of all choices made within a session being correct for three consecutive sessions on the hardest level of difficulty to complete the task. After attaining this criterion the subjects performed three sessions on a task that was identical in all respects except that the red, green, and blue gun values were now set at values derived from a subjective method of determining isoluminance. We used a modified chromatic flicker procedure to determine luminance equivalence (Kaiser 1991). There is physiological (Lee 1991) and behavioral evidence (DeValois et al. 1974) that the visual systems of macaques and humans are very similar with respect to spectral sensitivity and flicker perception. Both species have behavioral chromatic flicker thresholds of ~12-15 Hz. Our procedure was carried out on human subjects. The aim was to make each of the foils isoluminant with the green S+. The method entailed a series of experiments in which square stimuli of different chromacities were square-wave modulated with the green S+ with the use of a staircase procedure with suprathreshold and subthreshold flicker. In each of several stages a nine-choice procedure was used in which subjects were asked to select the stimulus of minimum perceived flicker and thus the stimulus that was most similar in luminance to the green S+. Minimum perceived flicker was generally selected at a flicker rate of 15 Hz, which is close to the upper threshold of human chromatic flicker perception. Human subjects found it subjectively harder to discriminate the S+ from the closest color (square color 2) with the use of the values derived from the flicker fusion method. Thus this stage of the color discrimination learning task was at least as hard if not harder than the previous stages. All subjects also learned preoperatively a concurrent visual learning task with pairs of complex colored shapes that forms part of a different study; those results are not reported here. The subjects completed this other task in an average of 19 days.

SURGERY. The operations were performed in sterile conditions with the aid of an operating microscope and the monkeys were anesthetized throughout surgery with barbiturate (5% sodium thiopentone solution) administered through an intravenous cannula.

For the perirhinal cortex ablation, the arch of the zygoma was removed and the temporal muscle was detached from the cranium and retracted. Surgery was carried out on one hemisphere at a time. A bone flap was raised over the frontal and temporal lobe. The medial and posterior limits of the flap were in a crescent shape extending from within 5 mm of the midline at the brow to the posterior insertion of the zygoma. The anterior limit of the flap was the brow and the orbit. Ventrally the flap extended from the posterior insertion of the zygoma to the level of the superior temporal sulcus in the lateral wall of the temporal fossa anteriorly. The ventral anterior part of this bone flap was extended with a rongeur to the base of the temporal fossa. The dura mater was cut to expose the dorsolateral frontal and lateral temporal lobes. The frontal lobe was retracted from the orbit with a brain spoon to enable access to the anterior medial temporal lobe. Pia mater was cauterized and the underlying cortex was removed by aspiration in the lateral bank of the anterior part of the rhinal sulcus and in the adjacent 2 mm of cortex on the third temporal convolution. The monkey's head was then tilted to an angle of 120° from the vertical, and the base of the temporal lobe was retracted from the floor of the temporal fossa with a brain spoon. The posterior tip of the first part of the ablation was identified visually and the removal was then extended in the lateral bank of the rhinal sulcus to the posterior tip of the sulcus, again removing 2 mm of laterally adjacent tissue. The dura mater was then sewn, the bone flap was replaced, and the wound was closed in layers. The ablation was made in both hemispheres in a single operation.

For the MTG ablation, the zygomatic arch was likewise removed and the temporal muscle detached from the cranium and retracted. Surgery was carried out on one hemisphere at a time. A bone flap was raised over an area larger than the intended lesion and then extended with a rongeur down to the temporal fossa. The dura mater was then cut to reveal the MTG. The anterior boundary of the lesion was an imaginary line drawn between the rostral tips of the superior temporal sulcus and the anterior middle temporal sulcus. The caudal boundary of the lesion was an imaginary line drawn 7 mm anterior to the inferior occipital sulcus, perpendicular to the superior temporal sulcus; however, in some hemispheres, part of the posterior middle temporal sulcus could be seen running parallel to the inferior occipital sulcus in this position, and in those hemispheres the posterior middle temporal sulcus was the posterior limit of the ablation. The dorsal boundary of the lesion was the fundus of the superior temporal sulcus. The ventral boundary of the lesion was the fundus of the anterior middle temporal sulcus, anteriorly, and a line extrapolated along the line of the sulcus, continued posteriorly, parallel to the superior temporal sulcus. In some monkeys, the posterior middle temporal sulcus, which is variable in position, lay caudal to the anterior middle temporal sulcus along this imaginary line and formed the ventral limit of the lesion. The gray matter within the area of the lesion was removed by aspiration. Pia mater was cauterized along the banks of the sulci. The pia mater in the sulci was left to enable the cortex on the remaining bank of the sulcus to survive. The dura was then sewn, the bone flap was replaced, and the wound was closed in layers. This ablation was also made in both hemispheres in a single operation.

HISTOLOGY. After the conclusion of all behavioral experiments the monkeys with ablations were sedated, deeply anaesthetized, and then perfused through the heart with saline solution (0.9%), which was followed by formol saline solution (10% Formalin in 0.9% saline solution). The brains were blocked in the coronal stereotaxic plane posterior to the lunate sulcus, removed from the skull, allowed to sink in sucrose Formalin solution (30% sucrose, 10% Formalin), and sectioned coronally at 50 µm on a freezing microtome. Every 10th section through the temporal lobe was stained with cresyl violet and mounted.

Lesion extent. Removals in all three monkeys with ablations of the perirhinal cortex were essentially as intended. The intended and actual extents of the lesions in two monkeys in group PRh-A are depicted in Fig. 1; the third animal in each group was similar to the two illustrated. In all three cases there was some inadvertent damage due to slight involvement of the laterally adjacent area TE, on the left in monkeys PRh-A1 and PRh-A3 and bilaterally in monkey PRh-A2 for an anteroposterior extent of 2 and 4 mm on the left and right sides, respectively.

View larger version (75K): [in this window] [in a new window]   FIG. 1. Middle column: shaded regions show intended location and extent of the ablation of the perirhinal cortex on the ventral view of the brain (top) and coronal sections (bottom) from a standard macaque monkey brain. Ablation of the perirhinal cortex in 2 monkeys (PRh-A1, left column, and PRh-A3, right column) are shown by the black area on the ventral views (top) and actual drawings of coronal sections through the lesion at levels matching those in the "Intended Lesion" column. Heavy black lines: region in which cortex was removed. Different sections were required to match the levels for the left and right hemispheres for monkey PRh-A3. To aid in visual matching of coronal sections to ventral views, the ventral view is reversed (i.e., left hemisphere is on the left). Numerals: distance in mm from the interaural plane.

Removals in all three monkeys with ablations of the MTG were also essentially as intended, or nearly so. The intended and actual extents of the lesions in two monkeys in group MTG-A are depicted in Fig. 2; the third animal in each group was again similar to the two illustrated. There were small differences in the posterior extent of the lesions both within (left-right difference) and across animals, presumably as a result of the posterior boundary being set by either a measurement from the inferior occipital sulcus or from the posterior middle temporal sulcus, if available. For example, the ablations in monkeys MTG-A1 and MTG-A3 did not extend quite as far posteriorly on the right side as on the left. In addition, there was inadvertent damage to a small portion of the medial bank of the anterior middle temporal sulcus on the left in monkey MTG-A3, because the lesion extended somewhat more medially than intended, and to a small portion of the temporal polar region on the right in monkey MTG-A2, because the ablation extended slightly further rostrally than intended.

View larger version (80K): [in this window] [in a new window]   FIG. 2. Middle column: shaded regions show the intended location and extent of the ablation of the middle temporal gyrus on the ventral view of the brain (top) and coronal sections (bottom) from a standard macaque monkey brain. Ablation of the middle temporal gyrus in 2 monkeys (MTG-A1, left column, and MTG-A3, right column) are shown by the black area on the ventral views (top) and actual drawings of coronal sections through the lesion at levels matching those in the intended. Heavy black lines: region in which cortex was removed. In both cases, different sections were required to match the levels for the left and right hemispheres. To aid in visual matching of coronal sections to ventral views, the ventral view is reversed (i.e., left hemisphere is on the left). Numerals: distance in mm from the interaural plane.

POSTOPERATIVE TESTING. A minimum of 14 days was allowed for recovery before testing of the operated animals resumed; the controls rested for a similar period. Postoperatively the subjects relearned the color discrimination task with all eight foils present in every trial. The red, blue, and green gun values for each of the colored stimuli on the computer screen were derived from the flicker fusion method of determining isoluminance as in the final stage of preoperative training. The color discrimination task was the first postoperative task.

Experiment 2: DNMS

SUBJECTS. The subjects were four naive rhesus monkeys (M. mulatta), three males and one female, weighing from 3.9 to 4.2 kg at the time of surgery. They were housed individually. Monkeys were fed a diet of Purina primate chow supplemented with fruit; water was always available.

Behavioral scores of four unoperated rhesus monkeys and four rhesus monkeys that had received ablations of the perirhinal cortex, reported in Meunier et al. (1993), served as the basis for comparison. As in the present study, all the monkeys in the earlier study were naive before the initiation of training. To enable a direct comparison with the two groups from the previous study, the present DNMS task was also administered in the same laboratory and in the same way for all groups under consideration; specifically, the rate and sequence of training, delay intervals, and intertrial intervals were in all cases the same as those used here.

APPARATUS AND MATERIALS. Training was conducted in a modified Wisconsin General Testing Apparatus inside a darkened room. Sound masking was provided by a white-noise generator. The test tray measured 720 × 190 mm and contained a row of three food wells spaced 180 mm apart, center to center, that were located 90 mm from the front edge of the tray. Rewards consisted of a single banana-flavored pellet (300 mg, P. J. Noyes). Preliminary training employed several square gray plaques (76 mm square) and three objects dedicated to this stage. Test material consisted of >1,120 different objects that varied widely in size, shape, texture, and color.

PREOPERATIVE TRAINING. Preliminary training. The monkeys were first trained by successive approximation to displace cardboard plaques that covered the food wells to obtain a food reward hidden underneath. Then one of three objects used only in preliminary training was presented over a baited well. When the monkeys would displace the baited objects without hesitation, formal training began.

DNMS. The monkeys were trained in DNMS with trial-unique objects. Each trial was composed of two parts: sample presentation followed by choice test. On each trial, the monkey was presented with the sample object overlying the baited central well of the test tray; the monkey displaced the sample to obtain the food reward hidden underneath. After a 10-s delay, the sample object, now unbaited, and the baited novel object were presented for choice over the lateral wells of the test tray, and the monkey could obtain an additional reward by displacing the novel object. A 30-s intertrial interval ensued, after which the procedure was repeated with a novel pair of objects, and so on, until the 20 trials making up a test session were completed. During intertrial and delay intervals, an opaque screen separated the monkey from the test tray. When the opaque screen was raised to permit responses, a one-way vision screen blocked the monkey's view of the experimenter. The left-right position of the novel object on the choice test followed a balanced pseudorandom order, and there was no correction for errors. Monkeys were trained at a rate of 20 trials per day, 5 or 6 days/wk, to a criterion of 90 correct responses in 100 consecutive trials. After learning this basic recognition task with 10-s delays, the monkeys received bilateral ablations of the cortex of the MTG.

SURGERY. Monkeys were anesthetized with ketamine hydrochloride (10 mg/kg im) followed by isoflurane (1-2% to effect). After induction of anesthesia, the animal was treated with atropine sulfate (0.04 mg/kg im) to reduce secretions. Surgery was carried out with the use of aseptic techniques and heart rate, respiration rate, body temperature, expired CO₂ levels, and blood pressure were monitored throughout the procedure. The ablation was made by subpial aspiration of tissue under visual control with the aid of an operating microscope. When the ablation was completed, the wound was closed in anatomic layers. All monkeys received dexamethasone phosphate (0.4 mg/kg im) and Di-Trim (0.1 ml/kg im, 24% solution) for 1 day before surgery, and daily for 1 wk after surgery to reduce swelling and prevent infection, respectively. Monkeys also received acetominophen (40 mg) for 3 days after surgery as an analgesic.

In each monkey, removal of the zygoma preceded the taking of a large bone flap that extended over the lateral surface of the frontal and temporal lobes. One monkey had an irregularity of the cranium, and received a craniotomy (bilaterally) instead of a bone flap. The area of the intended lesion was the same as in experiment 1. Three monkeys received the ablation bilaterally in one stage and the fourth received it in two unilateral stages; this is the same number of one- and two-stage surgeries that had been carried out in the four monkeys with perirhinal cortex ablations (Meunier et al. 1993). Quarantine of an animal housing room disrupted the usual timing of the training and surgery sequence for two monkeys. As a result, 8 wk intervened between surgery and the initiation of postoperative testing for one monkey (MTG-B1), whereas 10 wk intervened between the completion of preoperative testing and surgery for another monkey (MTG-B2).

HISTOLOGY. At the conclusion of behavioral testing, the operated animals were given a lethal dose of pentobarbital sodium and were perfused intracardially with normal saline followed by aldehyde fixatives. The brains were removed, allowed to sink in a glycerol-Formalin solution, and cut at 50 µm in the coronal plane on a freezing microtome. Every fifth section was mounted, stained with thionin, and coverslipped.

Lesion extent. The removals in three of the four monkeys that received ablations of the MTG (MTG-B2, MTG-B3, and MTG-B4) were complete, or nearly so. This is illustrated in Fig. 3, which depicts the intended and actual extent of the MTG lesions in two monkeys best representative of the MTG group. Of the three, there was slight, partial sparing of the lateral bank of the anterior middle temporal sulcus bilaterally in monkey MTG-B4, near the fundus. As for the fourth monkey (MTG-B1), there was again slight sparing of the lateral bank of the anterior middle temporal sulcus bilaterally, near the fundus, and sparing of the most caudal 3 mm of the MTG bilaterally as well. Inadvertent damage was limited to a caudal portion of the posterior ITG on the left in monkey MTG-B3.

View larger version (80K): [in this window] [in a new window]   FIG. 3. Middle column: shaded regions show the intended location and extent of the ablation of the middle temporal gyrus on the ventral view of the brain (top) and coronal sections (bottom) from a standard macaque monkey brain. Ablation of the middle temporal gyrus in 2 monkeys (MTG-B2, left column, and MTG-B4, right column) are shown by the black area on the ventral views (top) and actual drawings of coronal sections through the lesion at levels matching those in the intended. Heavy black lines: region in which cortex was removed. To aid in visual matching of coronal sections to ventral views, the ventral view is reversed (i.e., left hemisphere is on the left). Numerals: distance in mm from the interaural plane.

The ablations of the MTG in monkeys of experiment 2 were compared with those in monkeys in experiment 1 by plotting, for each monkey, the extent of the lesion onto the same set of standard sections and reconstructing the lesion onto the standard ventral view of the brain. Examination of the lesion plots and reconstructions indicates that the extents of the lesions sustained by the monkeys comprising the two groups is comparable.

The removals in the monkeys with perirhinal cortex lesions, already reported in Meunier et al. (1993), were also essentially as intended. The ablations were compared with those sustained by monkeys in experiment 1 by plotting, for each monkey, the extent of the lesion onto the same set of standard sections and preparing reconstructions of the lesion onto ventral views of the brain. In both groups, the lesions usually extended slightly more laterally than planned, involving more of the cortex between the rhinal sulcus and anterior middle temporal sulcus than depicted in the illustration of the intended lesion (see Fig. 2). The only systematic difference between the two groups appeared to be sparing of the most rostrolateral portion of the perirhinal cortex in the monkeys in experiment 2 but not in those in experiment 1. Thus the monkeys in experiment 2 had slightly smaller perirhinal lesions than those in experiment 1.

POSTOPERATIVE TESTING. Approximately 2 wk after surgery, with the exception noted above, the monkeys were retrained on the basic DNMS task (with 10-s delays) to the same criterion as before. Each monkey was then given a performance test adapted from Gaffan (1974) in which, first, the delay between sample presentation and choice test was lengthened in stages from the initial delay of 10 s to 30, 60, and finally 120 s, and then the list of sample objects to be remembered was increased in steps from the original single object to 3, 5, and finally 10 objects. In the list-length tests, the sample objects were presented one at a time at 20-s intervals, and then each sample was paired successively with a different novel object, also at 20-s intervals. Consequently, the minimum retention interval for each trial was 20 s multiplied by the length of the list. For each delay condition, the monkeys received five consecutive daily sessions of 20 trials each, and, for each list-length condition, the monkeys received five consecutive daily sessions of 30 trials each.

	RESULTS

Abstract Introduction Methods Results Discussion References

Experiment 1: color discrimination

For the color discrimination task the number of pre- and postoperative errors accumulated and the number of trials performed respectively by each animal are shown in Table 1. Preoperatively the 11 monkeys attained criterion on the color discrimination task in an average of 1,675 trials (range 1,292-2,017) with an average of 200 errors (range 105-320). A parametric one-way analysis of variance (ANOVA) confirmed that the three groups that were formed (PRh-A, MTG-A, and CON-A) did not differ from each other in either of these preoperative learning scores (errors: F = 1.196,df = 2, 8; trials: F < 1, df = 2, 8).

Postoperatively the animals in the control group (CON-A) attained criterion in an average of 541 trials and made on average 81 errors. Group PRh-A attained criterion in an average of 599 trials with an average of 66 errors. The MTG-A animals, in contrast, performed an average of 704 trials while making on average 504 errors, an error rate far in excess of that of the other two groups. The postoperative rate of producing a correct response was so low for the MTG-A animals that they were only required to produce 25 correct responses per session rather than the 100 correct responses per session required from the other groups. When each MTG-A animal had eventually accumulated 200 correct responses, further testing was ceased. At this point the MTG-A animals were all deemed to have "failed" to relearn the task. Because of the failure of the MTG-A animals to attain criterion on the color discrimination task, nonparametric tests were used to compare the error scores from the different groups. The errors to criterion relearning score differed significantly between groups [Kruskal-Wallis 1-way ANOVA, H(2) = 6.01, P < 0.05]. The effect is shown in Fig. 4, left.

View larger version (18K): [in this window] [in a new window]   FIG. 4. Mean preoperative learning and postoperative relearning scores (errors to criterion) in the color discrimination task (experiment 1) and in the basic delayed nonmatching-to-sample (DNMS) task (experiment 2). Experiment 1: group CON-A (normal controls; n = 5), group PRh-A (bilateral ablation of the perirhinal cortex; n = 3), and group MTG-A (bilateral ablation of the middle temporal gyrus; n = 3). Experiment 2: group MTG-B (bilateral ablation of the middle temporal gyrus; n = 4). Group CON-B (normal controls; n = 4), and group PRh-B (bilateral ablation of the perirhinal cortex; n = 4) are from Meunier et al. (1993).

To compare the postoperative error scores from the color discrimination task (experiment 1) with the postoperative performance scores from the DNMS task (experiment 2), a measure of the accuracy of the discrimination of the S+ from each of the eight foils was calculated for each monkey; these values are shown in Table 1. To calculate these values, for each monkey and each of these foils the number of correct responses made to the S+ was expressed as a percentage of the total number of responses made to the S+ and to that particular foil. Thus an accuracy value of 100% against a foil would reflect that the monkey never chose that foil in preference to the S+, whereas an accuracy value of 50% against a foil would reflect that the monkey's preference between the S+ and that foil was at chance. These accuracy scores were analyzed by a 3 × 8 ANOVA with one repeated measure. Both main effects were significant (lesions: F = 40.59, df = 2, 8, P < 0.001; foil color: F = 20.88, df = 7, 56, P < 0.001); however, the interaction term was also significant (F = 5.52, df = 14, 56, P < 0.001). Thus the type of lesion and the difficulty of the discrimination had a significant interactive effect on hue discrimination accuracy. This is illustrated in Fig. 5, left, which shows that the poor hue discrimination accuracy of group MTG-A relative to both group PRh-A and group CON-A becomes even more marked when the green S+ and foil are closer in hue or saturation and therefore harder to discriminate from each other.

View larger version (15K): [in this window] [in a new window]   FIG. 5. Postoperative performance on the color discrimination task (experiment 1) and on the DNMS task with extended delays and list lengths (experiment 2). Scores in the color discrimination task are mean discrimination accuracy; % accuracy score for each foil is a measure of the number of times that the foil was chosen relative to the number of times that the green S+ was chosen in preference to it (all stimuli were isoluminant). Curves: mean accuracy for each group toward foils 6-2, which vary in hue from red to green, and toward foils 9-7, which vary in saturation from gray to green. Scores in the DNMS task are mean % correct responses. Curves: mean score for each group to conditions 1-4 when increasingly longer delays were imposed between sample and choice test, and to conditions 5-7 when increasingly longer lists of items were presented.

Pairwise comparisons of the accuracy scores showed that group MTG-A differed significantly from both the control group and group PRh-A, whereas group PRh-A did not differ significantly from the control group (Tukey's HSD tests; MTG-A vs. CON-A: P < 0.05; MTG-A vs. PRh-A: P < 0.05; PRh-A vs. CON-B: P > 0.05).

Experiment 2: DNMS

For the basic DNMS task the number of pre- and postoperative errors accumulated and the number of trials performed respectively by each animal are shown in Table 2; this table also shows the percentage of correct responses on the extended delays and list lengths presented postoperatively.

Preoperatively the 12 monkeys attained criterion on the basic DNMS task in an average of 120 trials (range 0-260) with an average of 34 errors (range 0-56). A parametric one-way ANOVA confirmed that the three groups that were formed (PRh-B, MTG-B, and CON-B) did not differ from each other in either of these preoperative learning scores (errors: F = 1.198, df = 2, 9; trials: F = 0.445, df = 2, 9).

Postoperatively the animals in the control group (CON-B) relearned the basic DNMS task immediately; all took zero trials and made zero errors preceding criterion. The MTG-B animals took 105 trials on average and made an average of 24 errors preceding criterion. Group PRh-B took 380 trials on average and made an average of 84 errors preceding criterion, which was >3 times as many trials and >3 times as many errors as group MTG-B. Because of the control animals' immediate relearning of the basic DNMS task, nonparametric tests were used to compare the error scores from the different groups. Both of the relearning error scores, trials and errors preceding criterion, differed significantly between groups; see Fig. 4 [Kruskal-Wallis 1-way ANOVA, trials: H(2) = 9.12, P < 0.02; errors: H(2) = 9.087, P < 0.02].

The groups differed markedly on the postoperative performance test (Table 2, Fig. 5). Monkeys in group PRh-B obtained an average of only 78.2% correct responses over the six conditions and made more errors than groups CON-B and MTG-B, which showed an average of 92.6 and 90.2% correct responses, respectively. The mean scores on the performance test were analyzed by a 3 × 7 ANOVA with one repeated measure (scores at the 10-s delay were included). Both main effects were significant (lesions: F = 58.83, df = 2, 9, P < 0.001; conditions: F = 9.06, df = 6, 54, P < 0.001) whereas the interaction term was not significant (F = 1.71, df = 12, 54, P > 0.05). Thus both the type of lesion and the difficulty of the task each had a significant additive effect on the DNMS performance scores. This is illustrated in Fig. 5, which shows that the performance level on DNMS, although distinctly worse in group PRh-B than in the other two groups, falls off to some extent in all three groups with the increasing difficulty between conditions. Pairwise comparisons of the scores on the delay conditions alone (excluding the 10-s delay, which was not an equivalent performance measure because this score was based on relearning to criterion), on the list-length conditions alone, and on all the conditions together all showed that group PRh-B differed significantly from both the control group and group MTG-B, whereas group MTG-B did not differ significantly from the control group (Tukey's HSD tests; PRh-B vs. CON-B: P < 0.05; PRh-B vs. MTG-B: P < 0.05; MTG-B vs. CON-B: P > 0.05).

Comparison of experiment 1 and experiment 2 results

Figure 4 shows the pre- and postoperative errors to criterion both in the color discrimination task (experiment 1) and in the DNMS task (experiment 2). In the color discrimination task the MTG group shows a large impairment in postoperative relearning, whereas the perirhinal group is not impaired. In the basic DNMS task the converse is true: the MTG group is mildly affected, whereas the perirhinal group shows a large impairment in postoperative relearning.

Figure 5 shows the accuracy scores for the discrimination of the S+ from each of the foils in the color discrimination task and also shows the postoperative performance scores on the DNMS task with extended delays and list lengths. In the color discrimination task the perirhinal group and the control group are able to discriminate the S+ from the foils with mean accuracy levels of 98.6 and 98.4%, respectively. The MTG group, however, performed the color discriminations with a consistently lower level of accuracy. The mean accuracy with which the MTG group discriminated the green S+ from each of the five foils that varied in hue ranged from 86% toward the red foil to 61.3% toward the foil that was closest in hue to the green S+. The mean accuracy with which the MTG group discriminated the green S+ from each of the three foils that varied in saturation ranged from 80.4% accuracy toward the gray foil to 73.5% accuracy toward the foil that was closest in saturation to the green S+. The pattern of results of the two lesioned groups in the DNMS task differed markedly from those in the hue discrimination task. Whereas the MTG group and the control group achieved a mean level of performance of 90 and 93% correct responses, respectively, across all the conditions of extended delays and list lengths in the DNMS task, the perirhinal group's performance fell off with the extended delays to an average of 79% correct responses at the 120-s delay, and ranged from 76 to 68% correct responses with list lengths of 3 and 10, respectively.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Effects of perirhinal and MTG lesions

Meunier et al. (1993) reported that bilateral ablation of the perirhinal cortex produced a striking deficit in visual recognition memory as measured by DNMS with trial unique objects. Monkeys with this lesion were severely impaired both in postoperative relearning of the basic nonmatching principle and in the subsequent performance test with extended delays and lists. We have shown that bilateral ablation of the perirhinal cortex does not, however, produce an impairment in color discrimination. The performance of monkeys with this lesion did not differ from that of the control group in color discrimination (Table 1; Figs. 4 and 5).

Our results have shown that monkeys with bilateral ablation of the MTG were only mildly affected in their postoperative relearning of the basic nonmatching rule and were unimpaired in the subsequent performance test with extended delays and lists (Table 2; Figs. 4 and 5). In contrast, bilateral ablation of the MTG produced a striking deficit in color discrimination (Table 1; Figs. 4 and 5). Monkeys with this lesion were severely impaired in their postoperative relearning of this task.

Thus, whereas bilateral perirhinal cortex lesions produced large impairments on the DNMS task but not on the color discrimination task, bilateral MTG lesions produced large impairments on the color discrimination task but not on the DNMS task. Therefore we have shown a double dissociation between the impairments produced by bilateral lesions of the perirhinal cortex from the impairments produced by bilateral lesions of the MTG.

The subjects performing the color discrimination task in this study were cynomolgus monkeys (M. fascicularis), whereas the subjects that performed DNMS in this study and the subjects that performed DNMS in the previous study that was used as the basis for comparison (Meunier et al. 1993) were rhesus monkeys (M. mulatta). There is experimental evidence that the effects of lesions are comparable between these two groups. First, the effects of perirhinal or rhinal lesions on delayed matching to sample and DNMS have been found in both rhesus monkeys (Meunier et al. 1993) and cynomolgus monkeys (Eacott et al. 1994; Horel et al. 1987). Further, the dissociation of the effects of MTG and ITG lesions on delayed matching to sample and DNMS have now been found in both cynomolgus monkeys (Horel et al. 1987) and in rhesus monkeys (present study). Second, there were no species differences in preoperative performance of color discrimination in a previous study (Heywood et al. 1995), nor were there any species differences in unoperated monkeys in DNMS (Meunier et al. 1993; Murray and Mishkin 1986). The evidence supports the argument that the lesions are comparable and the double dissociation is stable across animal groups.

How does the role of the perirhinal cortex differ from that of area TE?

Although there is both anatomic and cytoarchitectural evidence for a dorsal-ventral subdivision of IT, there has been only limited evidence to suggest corresponding functional subdivisions. Recently much interest has been directed to one particular area of IT, the perirhinal cortex. Despite this, the perirhinal cortex had not previously been shown to be functionally distinct from the adjacent cortical area TE.

In this paper the functional double dissociation we have reported between two anatomically and cytoarchitecturally defined subdivisions of IT, namely the perirhinal cortex and the MTG, demonstrates that these areas are indeed functionally distinct cortical areas within IT and gives strong support to the hypothesis that the perirhinal cortex and its neighboring cortical area TE have different roles in visual learning and memory.

In the introduction we reviewed recent perirhinal lesion studies that suggest that the role of the perirhinal cortex is concerned with knowledge about objects. On the basis of the functional double dissociation we have shown, and on the evidence from the perirhinal lesion studies we reviewed earlier, we suggest that the perirhinal cortex is a functionally distinct area of IT cortex primarily involved in processing knowledge about whole objects or about the constitutive parts of whole objects.

TE has also long been associated with object identification. Horel (1994a) found that reversible suppression of the dorsal half of TE, the area of the MTG, disrupted the retrieval of some well-learned object discriminations but not others, and different animals had difficulty with different objects. They interpreted this as indicating that dorsal TE processes elements or features of the visual image, not the entire image. Evidence was found that stimulus size may be one such feature, because suppressing dorsal TE, rather than disrupting the retrieval of all forms, disrupted the retrieval of small details but not the retrieval of global forms. Another such feature that dorsal TE may be involved in processing is color; Horel (1994b) found that cooling dorsal TE suppressed new and recent learning of color discrimination, Heywood et al. (1995) showed that lesions in the temporal lobe anterior to area V4 resulted in cerebral achromatopsia, Komatsu et al. (1992) classified 71% of the neurons they recorded from in anterior IT as being color selective, and the MTG lesions we made resulted in striking deficits in color discrimination. In contrast to the perirhinal cortex, which we infer to be involved in processing whole objects or constituent parts of whole objects, dorsal TE may be involved in processing the more general attributes of objects, local details and color being two examples of these attributes.

Although the exact nature of the difference between the role of the perirhinal cortex and that of the more anterior cortex within IT has not been fully ascertained by these experiments, a difference in role has now been firmly established by the demonstration of a functional double dissociation. Further research will be needed to provide a fuller description of the nature of the functional differences between TE, or subdivisions of TE, and the perirhinal cortex.

	FOOTNOTES

  Address for reprint requests: M. J. Buckley, Dept. of Experimental Psychology, Oxford University, South Parks Rd., Oxford OX1 3UD, UK.

  Received 9 July 1996; accepted in final form 3 October 1996.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• AMARAL, D. G., INSAUSTI, R., COWAN, W. M. The entorhinal cortex of the monkey. I. Cytoarchitectonic organization. J. Comp. Neurol. 264: 326-355, 1987.[Medline]
• BARBAS, H. Topography and laminar organization of cortical visual input to the prefrontal cortex of the rhesus monkey. Soc. Neurosci. Abstr. 11: 324, 1985.
• BRODMANN, K. Beitrage zur histologischen Lokalisation der Grosshirnrinde. Dritte Mitteilung: die Rindenfelder der niederen Affen. J. Psychol. Neurol. 4: 177-226, 1905.
• BRODMANN, K. In: Vergleichende Lokalisationslehre der Grosshirnrinde in Ihren Prinzipen Dargestellt auf Grund des Zellenbaus., . Liepzig, Germany: Barth, 1909
• DEVALOIS, R. L., MORGAN, H. C., POLSON, M. C., MEAD, W. R., HULL, E. M. Psychophysical studies of monkey vision. I. Macaque luminosity and color vision tests. Vision Res. 14: 53-67, 1974.[Medline]
• EACOTT, M. J., GAFFAN, D., MURRAY, E. A. Preserved recognition memory for small sets, and impaired stimulus identification for large sets, following rhinal cortex ablations in monkeys. Eur. J. Neurosci. 6: 1466-1478, 1994.[Medline]
• GAFFAN, D. Recognition impaired and association intact in the memory of monkeys after transection of the fornix. J. Comp. Physiol. Psychol. 86: 1100-1109, 1974.[Medline]
• GAFFAN, D. Dissociated effects of perirhinal cortex ablation, fornix transection and amygdalectomy: evidence for multiple memory systems in the primate tempoal lobe. Exp. Brain Res. 99: 411-422, 1994.[Medline]
• GAFFAN, D., SAUNDERS, R. C., GAFFAN, E. A., HARRISON, S., SHIELDS, C., OWEN, M. J. Effects of fornix transection upon associative memory in monkeys: role of the hippocampus in learned action. Q. J. Exp. Psychol. B Comp. Physiol. Psychol. 36: 173-221, 1984.[Medline]
• GROSS, C. G. How inferior temporal cortex became a visual area. Cereb. Cortex 5: 455-469, 1994.
• HEYWOOD, C. A., GAFFAN, D., COWEY, A. Cerebral achromatopsia in monkeys. Eur. J. Neurosci. 7: 1064-1073, 1995.[Medline]
• HOREL, J. A. Local and global perception examined by reversible suppression of temporal cortex with cold. Behav. Brain Res. 65: 157-164, 1994a.[Medline]
• HOREL, J. A. Retrieval of color and form during suppression of temporal cortex with cold. Behav. Brain Res. 65: 165-172, 1994b.[Medline]
• HOREL, J. A., PYTKO-JOINER, D. E., VOYTKO, M. L., SALSBURY, K. The performance of visual tasks while segments of the inferotemporal cortex are suppressed by cold. Behav. Brain Res. 23: 29-42, 1987.[Medline]
• INSAUSTI, R., AMARAL, D. G., COWAN, W. M. The entorhinal cortex of the monkey. II. Cortical afferents. J. Comp. Neurol. 264: 356-395, 1987.[Medline]
• IWAI, E., YUKIE, M. Amygdalofugal and amygdalopetal connections with modality-specific visual cortical areas in macaques (Macaca fuscata, M mulatta, and M fascicularis). J. Comp. Neurol. 261: 362-367, 1987.[Medline]
• IWAI, E., YUKIE, M., SUYAMA, H., SHIRAKAWA, S. Amygdalar connections with the middle and inferior temporal gyri of the monkey. Neurosci. Lett. 83: 25-29, 1987.[Medline]
• KAISER, P. K. Flicker as a function of wavelength and heterochromatic flicker photometry. In: Vision and Visual Dysfunction: Limits of Vision, edited by J. J. Kulikowski, V. Walsh, and I. J. Murray . London: Macmillan, 1991, vol. 5, p. 171-179
• KONATSU, H., IDEURA, Y., KALI, S., YAMANE, S. Color selectivity of neurons in the inferior temporal cortex of the awake macaque monkey. J. Neurosci. 12: 408-424, 1992.[Abstract]
• LEE, B. B. Spectral sensitivity in primate vision. In: Vision and Visual Dysfunction: Limits of Vision, edited by J. J. Kulikowski, V. Walsh, and I. J. Murray . London: Macmillan, 1991, vol. 5, p. 191-199
• MEUNIER, M., BACHEVALIER, J., MISHKIN, M., MURRAY, E. A. Effects on visual recognition of combined and separate ablations of the entorhinal and perirhinal cortex in rhesus monkeys. J. Neurosci. 13: 5418-5432, 1993.[Abstract]
• MURRAY, E. A., BACHEVALIER, J., MISHKIN, M. Effects of rhinal cortical lesions on visual recognition memory in rhesus monkeys. Soc. Neurosci. Abstr. 15: 342, 1989.
• MURRAY, E. A., GAFFAN, D., MISHKIN, M. Neural substrates of visual-stimulus association in rhesus monkeys. J. Neurosci. 13: 4549-4561, 1993.[Abstract]
• MURRAY, E. A., MISHKIN, M. Visual recognition in monkeys following rhinal cortical ablations combined with either amygdalectomy or hippocampectomy. J. Neurosci. 6: 1991-2003, 1986.[Abstract]
• SELTZER, B., AND PANDYA, D. N. Afferent cortical connections and architectonics of the superior temporal sulcus and surrounding cortex in the rhesus monkey. Brain Res. 149: 1-24, 1978.[Medline]
• SUZUKI, W. A., ZOLA-MORGAN, S., SQUIRE, L. R., AMARAL, D. G. Lesions of the perirhinal and parahippocampal cortices in the monkey produce long-lasting memory impairment in the visual and tactual modalities. J. Neurosci. 13: 2430-2451, 1993.[Abstract]
• VAN ESSEN, D. C., FELLEMAN, D. J., DEYOE, E. A., OLAVARRIA, J., KNIERIM, J. Modular and hierarchical organization of extrastriate visual cortex in the macaque monkey. Cold Spring Harbor Symp. Quant. Biol. 55: 679-696, 1990.[Medline]
• VAN HOESEN, G. W., PANDYA, D. N. Some connections of the entorhinal (area 28) and the perirhinal (area 35) cortices of the rhesus monkey. I. Temporal lobe afferents. Brain Res. 95: 1-24, 1975.[Medline]
• VON BONIN, G., BAILEY, P. In: The neocortex of Macaca mulatta., . Urbana, IL: Univ. of Illinois, 1947
• WITTER, M. P., VAN HOESEN, G. W., AMARAL, D. G. Topograpical organization of the entorhinal projection to the dentate gyrus of the monkey. J. Neurosci. 9: 216-228, 1989.[Abstract]

0022-3077/97 $5.00 Copyright ©1997 The American Physiological Society