Spatial Responsiveness of Monkey Hippocampal Neurons to Various Visual and Auditory Stimuli

Transcription

1 HPPOCAMPUS, VO. 2, NO. 3, PAGES , JUY 1992 Spatial Responsiveness of Monkey Hippoampal Neurons to Various Visual and Auditory Stimuli Ryoi Tamura, Taketoshi Ono, Masaji Fukuda, and Kiyomi Nakamura Department of Physiology, Faulty of Mediine, Toyama Medial and Pharmaeutial University, Toyama, Japan ABSTRACT To investigate involvement of the hippoampal formation in spatial information proessing, ativity of neurons in the hippoampal formation of the onsious monkey was reorded during presentation of various visual and auditory stimuli from several diretions around the monkey. Of 1,47 neurons reorded, 16 (1.1%) responded to some stimuli from one or more diretions. Of these 16 neurons with diretionally differentiating responsiveness, 49 responded to visual stimulation, 35 to auditory stimulation, and 22 to both. Among 81 neurons, eah tested with more than 1 different stimuli, one type responded independent of the nature of the stimulus (nonseletive, n = 39), and responses of the other type depended on the nature of the stimulus (seletive, n=42). To investigate effets of hange in spatial relations between test stimuli and bakground stimuli fixed on the monkey or fixed in the environment, 59 of 16 neurons were tested while the experimental apparatus holding the stimulus was moved relative to the monkey. Of these 59 neurons, 36 hanged their responsiveness; 7 maintained the magnitude of their responses but hanged the response diretion with the movement of the apparatus, 5 hanged diretion regardless of the movement, and 24 did not hange diretion, but dereased or extinguished responses from the preferred diretion. Thirty-two of 16 neurons were also tested by rotating the monkey. The diretionally differentiating responsiveness of 1 1 neurons followed the monkey (egoentri neurons), that of 9 remained in plae in the environment (alloentri neurons), and responses of 12 were reversibly extinguished when the monkey was rotated. The results suggest that these hippoampal neurons may be involved in identifiation of relations among various kinds of stimuli in different spatial frameworks (egoentri or alloentri) and this identifiation may be developed from multiple sensory modalities. Key words: hippoampal formation, single unit reording, spatial memory, egoentri neurons, alloentri neurons t is now widely aepted that the hippoampal formation (HF) is involved in spatial memory as well as nonspatial memory. Humans with damage in the medial temporal lobe, inluding the HF, exhibit spatial memory defiits (Smith and Milner, 1981; 1984; 1989; Cave and Squire, 1991). n the monkey, bilateral lesions in the H F system impair performane in several kinds of spatial tasks (Zola-Morgan and Squire, 1986; Rupniak and Gaffan, 1987; Parkinson et al., 1988). n the rat, bilateral lesions in the H F system also impair performane in spatial tasks (Olton et al., 1978; Moms et al., 1982). Unit-reording studies of the rat H F suggest the existene of plae units in the H F that respond preferentially to partiular loations of the animal in the field (O Keefe and Correspondene and reprint requests to Professor Taketoshi Ono, Department of Physiology, Faulty of Mediine, Toyama Medial and Pharmaeutial University, Sugitani, Toyama 93-1, Japan. Dostrovsky, 1971; O Keefe and Nadel, 1978; Muller et al., 1987; Breese et al., 1989; Foster et al., 1989). Unit-reording reports indiate that some primate H F neurons respond in spatial tasks (Watanabe and Niki, 1985; Cahusa et al., 1989; Miyashita et al., 1989; Rolls et al., 1989; Feigenbaum and Rolls, 1991). These unit-reording studies in the monkey H F suggest that some neurons in the H F may be involved in spatial learning or memory developed from vision. As suggested in a previous study of the rat hippoampus (O Keefe and Conway, 1978), plae-speifi ativity hange in plae units an depend not only on vision but also on other sensory modalities, suh as audition. However, relations between neuronal responses of the monkey H F to spatial fators and properties of stimuli (suh as sensory modalities or a kind of stimulation in one sensory have not been asertained- The H F has reiproal onnetions with both the posterior region of the inferior parietal ortex and the prefrontal as- 37

2 38 HZPPOCAMPUS VO. 2, NO. 3, JUY 1992 soiation ortex, whih are involved in spatial information proessing (Teuber, 1964; Pohl, 1973; Petrides and versen, 1979; Andersen and Mountastle, 1983; Passingham, 1985; Andersen, 1987; Goldman-Raki, 19871, mainly via the parahippoampal orties (PH) (Jones and Powell, 197; Van Hoesen, 1982; Amaral, 1987; Tranel et al., 1988). Reent results indiate that the HF, and the system to whih it belongs, are essential for aquisition, relation, ombination, and onjuntion among stimuli (Eihenbaum et al., 1988; Squire et al., 1989; Sutherland and Rudy, 1989; Wiener et al., 1989) and suh funtions are important for alloentri spatial mapping. A omputational theory suggests the possibility that egoentri representation of information is onverted to alloentri form in the hippoampus (O Keefe, 199). Assuming that there are neurons in the monkey H F that respond to stimuli presented in a partiular loal ontext (or onfiguration of bakground stimuli), responsiveness of suh neurons may also depend on some referene stimuli as well. Furthermore, if referene stimuli are fixed to the animal or things that move with the animal, responses of neurons may have egoentri properties, and if referene stimuli are fixed in the environment, responses of neurons may have alloentri properties. t was thus thought to be useful and important to investigate whether there are neurons in the monkey H F that respond to some spatial aspets of stimuli presented, and if there are suh neurons, whih relations between stimuli are represented in their responses. n this study, we analyzed the responses of single neurons in the monkey H F to the presentation of various test stimuli, visual or auditory or both, from various diretions relative to the monkey. To investigate relations between test stimuli and referene stimuli further, the same tests were performed either while keeping the monkey fixed and hanging part of the environment, or while moving the monkey relative to a fixed environment. A preliminary report of this work has appeared elsewhere (Tamura et al., 199), and some supporting evidene has reently been reported (Feigenbaum and Rolls. 1991). MATERAS AND METHODS Exept for differenes in some parts of the experimental design and the stimulus presentation paradigm, the experimental methods were essentially the same as those used in the aompanying paper (Tamura et al.). Therefore, the Materials and Methods setion has been minimized: the reader is referred to the aompanying paper (pp ) and our previous papers (Ono et al., 198; 1981; 1989; Fukuda et al., 1986; Nishijo et al., 1988a; 1988b; Yamatani et al., 199). Experimental design and stimulus presentation The monkeys normally sat in a hair faing an apparatus (Fig. 1). The apparatus had a front panel and two wings of aluminum plate that ould be easily detahed. The panel had a window (Wm, about 1 m x 2 m) overed by a oneway-mirror shutter. Eah wing also had a window (W1 or Wr, about 15 m x 15 m). Behind eah window was a stage for setting objets. The normal arrangement of the experimental room is shown in Figure 1A. The apparatus was in front of the monkey (M). The experimenter(s) (H) usually sat in front of the monkey to its right and was hidden by a wing of the apparatus. n this situation, various visual and auditory stimuli were presented to the monkey from several diretions. Many different objets hosen from a pool of about 1,, as well as some parts of the human body, were used as visual stimuli. Sometimes food (raisin, a piee of apple, ookie, et.) was given to the monkey to retain its attention to the presented objets. Sine, in the experimental situation reported here, the monkey s limb movements (e.g., reahing his hands to foods) were restrited by attahing an aryli plate in front, the foods were plaed diretly in the monkey s mouth by the experimenter. The range that the monkey ould see was restrited to about 28 from the enter by attahing opaque aryli plates at the sides of its fae. Usually eah objet was presented by the experimenter putting it on the stage behind the window for about 2. seonds. The distane between the monkey s fae and the objet was usually about 4 m. Sometimes the same objet was also presented by the experimenter s hand, whih approahed loser to the monkey. When large objets, suh as the human body, were presented, the wings were detahed or the experimental apparatus in front of the monkey was ompletely removed. When the apparatus was removed and the room light was on, the monkey ould see the white walls of the room in front and to the left anterior, and the reording setup to the right anterior (Fig. 1A). n this situation, the experimenter ould walk around the monkey, stop at various plaes, and display various ations, suh as sitting, standing, or presenting an objet by hand that had been shown in the experimental apparatus. Many different kinds of sounds were used for auditory stimulation, inluding meaningful or omplex sounds (step sound, lap, human voie, rash, et.) and omputer-synthesized sounds (harmoni rih sounds or pure tones) using a sound board (PC K, NEC). Meaningful or omplex sounds were usually presented by some at of the experimenter (stepping, lapping, voalizing, rashing metal materials, et). Computer-synthesized sounds were presented through speakers arranged around the monkey. Eah sound soure (experimenter s at or speaker) was usually loated about 8 m from the monkey s head exept for sound from fixed soures, suh as the sound of opening or losing the entrane door. The intensity of omputer-synthesized sounds was usually ontrolled to about 8 db. ntensities of omplex sounds made by the experimentr were estimated to range between 7 and 9 db. Most auditory stimuli were presented from various diretions around the monkey by moving the soure of the sound. f a stimulus was onsidered to ontain both visual and auditory omponents, we attempted to separate the modalities by attenuating the intensity of the sound or by masking it with white noise, or by reduing illumination to a minimum to restrit the monkey s vision. When ativity of a single neuron in the H F or PH was deteted, a few kinds of objets or human ations were presented as visual stimuli at the left anterior (about 45 from the anterior-posterior plane), anterior, and right anterior positions (Fig. 1B). Eah visual stimulus was presented five or more times at eah position before the next objet was presented. A few kinds of sounds were similarly presented five or more times eah from the left anterior, anterior, right anterior, right posterior, posterior, and left posterior diretions.

3 A AC 1 - MONKEY HPPOCAMPA RESPONSES TO SPACE / Tarnura et al. 39 PC 981 Task Control / - ' 1, H ) -/ Am piif ier et. ATAC-45 (On lna A M Y S ~ ) Wm w r \ H : Human M : Monkey : Objet S : Speaker B Door Anterior J eft Anterior eft Posterior t Posterior Right Posterior - VSUA _-. AUDTORY Fig. 1. Shema of experimental situation. (A) Arrangement of 4 m x 6 m experimental room. Monkey sat in a hair faing a front panel. Monkey's limb movements were restrited by plate attahed in front of its body. Two wings (left wing and right wing) were at left and right sides of the front panel. There was a window in the front panel (Wm) and one in eah wing (Wl, Wr). Experimenter(s) were usually at right anterior of the monkey to observe behavior of the monkey and diret on line analysis of unit ativity. Various kinds of objets () were presented to monkey from left anterior (W), anterior (Wm), or right anterior (Wr) diretions as visual stimuli. Various kinds of sounds were presented to monkey from one of eight speakers (S) set around monkey as auditory stimuli. Entrane door to room was behind monkey. (B) Diretions of visual and auditory stimulation soures. Various kinds of visual and/or auditory stimuli were presented from several diretions around the monkey. Solid lines, visual stimulation; broken lines, auditory stimulation. M, monkey; H, experimenter(s);, objets presented; S, speakers used for omputer-generated auditory stimuli; AC, air onditioner. f neuronal ativity hanged when the relative diretion of the origin of a stimulus was hanged, then more detailed relations between the diretionally differentiating responsiveness and the kind of stimulation were examined using various kinds of visual and auditory stimuli. We use the term diretionally differentiating sine the neurons responded differently when stimuli were presented from different positions (or diretions) around the monkey. This phrase is then used to desribe responses of these neurons as they are subjeted to orresponding proedures. Stimulus presentations were separated by intervals of at least 3 seonds. After these tests were onluded, and if the neuron being reorded was thought to have diretionally differentiating responsiveness, we tested its responses in the following two proedures. Proedure 1 The experimental apparatus in front of the monkey was moved to several plaes in the experimental room. Sine the experimental apparatus was usually loated in front of the monkey and overed a large part of its visual field, this apparatus was onsidered to have many visual stimuli that ould be dominant landmarks (referenes) for loational information of test stimuli. f the neuron did not hange responses

4 31 HPPOCAMP~S VO. 2, NO. 3, JUY 1992 in this proedure, it was deemed to be responding to relations between a test stimulus and bakground stimuli other than those on the experimental apparatus. f the responses of the neuron followed movement of the apparatus, the neuron ould have been responding to relations between test stimuli and stimuli on the experimental apparatus. Bakground stimuli is a general term meaning stimuli that ould provide spatial referene for the test stimulus. Therefore, bakground stimuli inlude all stimuli fixed in the environment (experimental room), on the monkey, and all visible parts of the restraining devie. Proedure 2 The monkey hair was rotated 45 to the left and then 45 to the right of the experimental apparatus, and the test stimuli remained in their initial positions. f the responses follow this rotation, as in Figure 9A, the neuron was onsidered to respond to some spatial relation between the test stimulus and some stimuli that rotated with the monkey. f the responses remained onstant relative to the environment, as in Figure 9B, the neuron was onsidered to respond to some spatial relation between the test stimulus and some stimuli that were fixed in the environment. Reording and data analysis Most of the reording tehnique and data analysis were the same as those desribed in the aompanying paper (pp ). Briefly, single-unit ativity was reorded from the HF and PH. Neuronal ativity was proessed in a window disriminator and displayed as peristimulus time histograms using miniomputers on-line and off-line. During auditory stimulation, the stimuli themselves triggered data olletion by onverting sounds to eletrial start signals through a mirophone near the monkey s head. For visual stimulation, the experimenter pressed a foot swith to trigger olletion of data when stimuli were presented. Neuronal responses for the first 1. seonds after presentation were averaged for eah of several stimulus presentations. Responses to the presented stimuli were ompared by one-way analysis of variane (ANOVA) with signifiane at P.5. ateny of eah response to the stimulus was measured mainly for auditory stimulation off-line sine timing of auditory stimulation ould be preisely determined from reorded sound data, whereas timing of visual stimulation, espeially of human ations, ould not be preisely determined. During reording sessions, eye movement of the monkey was monitored through a monohrome television amera or eletrooulograms (EOG) or both, as well as diretly by the experimenter. During reording sessions, behavior of the monkey was also monitored through a olor television amera or eletromyograms (EMG) or both, as well as diretly by the experimenter. Animal behavior RESUTS All monkeys usually turned their attention to a visual stimulus when it was presented. Sine the head was restrained, only the eyes of the monkeys ould move in the diretion of a visual stimulus. When food was presented, the monkeys would usually fix their eyes on it until they got it or it was withdrawn. No other onsistent overt behavior was notied when visual stimuli were presented. When auditory stimuli were presented, they did not usually exhibit any onsistently overt behavior or eye movement, exept when a loud or frightening sound was made. General properties of HF neurons with diretionally differentiating responsiveness Of 1,47 neurons tested in the HF and PH, 16 (1.1%) responded (all by exitation) preferentially to the presentation of visual andlor auditory stimuli from a partiular diretion(s) around the monkey. Many of the neurons responded phasially, but a few responded tonially during stimulus presentation. Judging from EOGs, television monitoring, and inspetion of experimenter, there was no lear orrelation between visual responses of most neurons and initial eye position or eye movement. Responses of these neurons did not orrelate to partiular movements of the monkey, suh as arm or leg movements. The spike ativity of some of these neurons was omplex, that is, a derementing series of spikes at short intervals (25. ms) with.4-.6 ms spike durations and low spontaneous firing rates (<2. spikesls). However, many of the neurons with diretionally differentiating responsiveness did not show typial omplex spike ativity, but showed irregular bursting with shorter spike duration of spikes (.2-.5 ms) and slightly higher spontaneous firing rates ( spikeds). Neurons that showed regular firing ativity with short spike duration (.2-.3 ms) and high frequeny of spontaneous firing (> 1. spikesh) usually did not have diretionally differentiating responsiveness. There was no overlap between the neurons with diretionally differentiating responsiveness desribed in this paper and neurons that responded to the sight of objets during performane of the objet disrimination task desribed in the aompanying paper (Tamura et al., 1992). Relations among sensory modalities, kinds of stimuli, responsiveness, and diretion Of these 16 neurons with diretionally differentiating responsiveness, 49 responded to visual stimulation, 35 responded to auditory stimulation, and 22 responded to both visual and auditory stimulation. The relations among sensory modalities, neuronal responses, and orientation are shown in Figure 2. The number of neurons in eah ategory that also responded in the other modality are shown in parentheses. The vision-responsive neurons tended to respond to stimulation presented from the right anterior diretion, and the sound-responsiveness neurons tended to respond more to stimulation presented from the posterior. Of the 22 visual plus auditory neurons with diretionally differentiating responsiveness, the visual and auditory responses of 11 overlapped (4, left anterior; 2, anterior; 5, right anterior) and responses of 11 did not overlap diretion. Of the 16 neurons, 81 ould be tested with more than 1 kinds of stimuli, and these 81 neurons were divided into two groups. n one group, responsiveness was independent of the nature (or kind) of the stimulus in one sensory modality (non-

5 MONKEY HPPOCAMPA RESPONSES TO SPACE / Tarnura et al A (2) A (1 1) Fig. 2. Relations among sensory modalities, neuronal responses, and stimulus orientation. Ordinates, numbers of neurons with diretionally differentiating responsiveness that respond in eah position. Hathed areas (V), neurons with visual responses; solid areas (A), neurons with auditory responses. Number in parentheses below eah olumn indiate number of neurons in that olumn that are bimodal. Totals may exeed those in text beause of overlap. seletive group, n = 39). n the other group, responsiveness depended on the nature of the stimulus, and these neurons tended to respond more to some stimuli than to others (seletive group, n = 42). Of 49 visual neurons, 34 were tested with more than 1 kinds of stimuli and 21 of 34 were lassified into a nonseletive group. Responses of a visual nonseletive neuron are shown in Figure 3. This neuron, loated in the CA area of the HF, was tested with 1 1 kinds of visual stimuli (not all shown) and 4 kinds of auditory stimuli. t responded phasially to visual stimuli presented from the right anterior diretion (A, B), and did not respond to the same visual stimuli presented from any other diretion (Aa, Ab). This neuron did not respond to any of four kinds of auditory stimuli presented from any diretion (Aa-f, pure tone; B, human voie, step, lap, pure tone). Thus the responses of this neuron were nonseletive to the kind of visual stimuli tested, but they were seletive to the diretion of stimulation (right anterior) and to sensory modality (vision). As shown in Figure 4, there seemed to be no lear relation between the response of this neuron and eye movement. For example, there was no signifiant eye movement in the seond presentation of apple (A2) and seond presentation of stik (C2) sine the monkey had already direted its eyes toward right anterior, and there was signifiant movement to the right in the other presentations. However, the responses of the neuron were almost the same in all of these situations. The other 13 of 34 visual neurons were lassified into a seletive group (1 1 responded strongly to partiular human movements and 2 responded to the sight of rewarding objets). Responses of a typial, visual seletive neuron are shown in Figure 5. This neuron, loated in the CAl area of the HF, was tested with 15 kinds of visual stimuli and 8 kinds of auditory stimuli (not all shown). t responded seletively to human ation (standing up or sitting down) in the right anterior position (A,B), but not to any of 14 other kinds of visual stimuli or 8 kinds of auditory stimuli (B). This neuron did not respond to any other test stimuli presented from any diretion (A). When room illumination was redued to a minimum or the monkey was prevented from seeing human ation by an opaque plate, responses to human ation at right anterior were ompletely extinguished (B, human ation in dark). The response magnitude of this neuron to the presentation of visual stimuli did not depend on the length of the period that the monkey fixed his eyes on the visual stimulus, sine the monkey almost always fixed his eyes longer on rewarding objets, suh as a piee of apple, raisin, or ookie, than on human ation. However, responses to human ation were muh stronger than the responses to these rewarding objets. Of 35 auditory neurons, 32 were tested with more than 1 kinds of stimuli, and 14 of 32 ould be lassified as nonseletive. Responses of an auditory nonseletive neuron are shown in Figure 6. This neuron, loated in the CA3 area of the HF, was tested with six kinds of visual stimuli and seven kinds of auditory stimuli. t responded strongly to auditory stimuli presented from the left posterior diretion (A). t also responded weakly to auditory stimuli from the posterior diretion, but not if they were presented from any other diretion. This neuron did not respond to any of six kinds of visual stimuli presented from anterior diretions. Magnitudes of responses were almost the same for all kinds of different auditory stimuli presented from the left posterior diretion (B) (ANOVA, P >. 1). Other auditory neurons (18 of 32) were lassified as seletive. All of these neurons exhibited strongly seletive responses to omplex sounds made by the experimenter: seven responded to a step, four to the sound of a hair moving, two to the sound of a door opening, two to a data reorder being swithed, two to board tapping, one to the sound of paper rumpling. Responses of an auditory seletive neuron are shown in Figure 7. This neuron, loated in the dentate gyrus, was tested with 8 kinds of visual stimuli (not shown) and 1 kinds of auditory stimuli. The neuron responded strongly to sounds related to human movement presented from posterior diretions (A, B); the strongest response was to the door being opened in the posterior diretion (B). When the diretionally differentiating responsiveness of this neuron was further tested by using a human voie, a lap, or a pure tone, responses to the human voie and lap diminished as the soure of the stimulus approahed the anterior diretion. There were no responses to a pure tone from any diretion nor to any auditory stimulation from anterior diretions within the monkey's vision (k 8" from the enter) (A). Judging from EOGs and television monitoring, there was no orrelation between eye movement and responses of these auditory neurons. Of 22 neurons that responded to both visual and auditory stimuli, 15 were tested with more than 1 kinds of stimuli. Of these 15 neurons, 4 were lassified as nonseletive and 11 as seletive (all 1 1 responded strongly to human ation suh as walking and to omplex sounds made by the experimenter).

6 31 2 HZPPOCAMPUS VO. 2, NO. 3, JUY 1992 U ",,, / Syringe,, i 111 Stik m. f 1' > n 1 1 ll ae se 3 f eft Posterior e Posterior t ae 1 1 r s e \ Y U a a 4 - E if 2 - E lu a - a E > 5 E O E a T - al - P) ' P v) T Y v) T h N r v P) C - Stimuli Presented Fig. 3. Responses of visual nonseletive neuron. (A) Diretionally differentiating responsiveness to visual and auditory stimuli shown as raster displays. Visual stimuli (e.g., apple, syringe, and stik) were presented from left anterior (a), anterior (b), and right anterior () diretions. Auditory stimuli (e.g., pure tone) were presented from six diretions (a-. Note phasi responses to visual stimuli presented from right anterior diretion (). Absissa, time (se); vertial line in eah plot indiates time when eah stimulus was presented; horizontal line above absissa, period of stimulus presentation (2 seonds). P, front panel. Other desriptions as for Figure 1. (B) Comparison of visual nonseletive responses to various kinds of visual or auditory stimuli presented from right anterior diretion. This neuron responded to all visual stimuli tested with almost the same response magnitude when presented from right anterior, but not from other diretions, and did not respond to auditory stimuli. Eah histogram, mean and SEM of response size of five presentations of eah stimulus. (Response size defined as differene between 1.O seond measurement of firing number before stimulus presentation and 1.O seond measurement of firing number during stimulus presentation.)

7 A Apple Bill 1 MONKEY HPPOCAMPA RESPONSES TO SPACE / Tamura et al \ H J u b - - -, B Syringe a M Stik the human ation was not hanged. *m ll 111 ll 4 3 J se Fig. 4. Relation between neuronal responses to visual stimulation (upper rastergrams) and eletrooulograms (EOGs, lower traes). These neuronal responses are those to the visual stimulation from the right anterior diretion in Figure 3 but shown relative to EOGs. There were no lear orrelations between eye movement and diretionally differentiating neuronal responses even if the visual stimulus was the sight of a rewarding (A, apple), an aversive (B, syringe), or a neutral (C, stik) objet. EOG, upward defletion, movement to left (); downward defletion, movement to right (R). Absissa, time (se); time, eah stimulus was presented; horizontal line above absissa, period of stimulus presentation (2 seonds). Spontaneous firing rates and response latenies of eah type of neuron are summarized in Table. Dependene of responsiveness on relations between test stimulus and bakground stimuli Fifty-nine of 16 neurons were tested by hanging the loation of experimental apparatus in front of the monkey while the monkey s position was fixed in the environment. Thirtytwo of 16 neurons were also tested by rotating the hair in whih the monkey was seated while the loation of the ex- R perimental apparatus was fixed. Changing the loation of the experimental apparatus modified the responses of 36 of 59 neurons. For 12 of these neurons, the diretion of maximum response hanged (7 neurons hanged the diretion aording to the movement of the apparatus and 5 hanged the diretion independent of the movement). For the other 24, the preferred diretion did not hange, but responses in the preferred diretion were very weakened, or the diretionally differentiating responses were reversibly extinguished. The responsiveness of a neuron whose responsiveness was ompletely extinguished by moving the apparatus is shown in Figure 8. This is the same neuron shown in Figure 5. The strong responses to human ation presented from the right anterior diretion (Aa) were ompletely extinguished when the experimental apparatus, whih was onsidered to be a part of the environment, was removed to a plae where the monkey ould not see it (Ab). The response magnitude also depended on the preise loation of the apparatus when the monkey ould see it. When the apparatus was moved about 5 m to either the left or the right, responses in the preferred diretion were steeply diminished irrespetive of the relations between the test stimulus and the experimental apparatus (Fig. 8A, B), and when the apparatus was moved more than 1 m to the left or 15 m to the right, responses were ompletely extinguished (Fig. SAb, B) although the visibility of When the monkey was rotated, the diretionally differentiating responsiveness of 11 neurons followed the monkey. We designate these neurons as egoentri, beause the diretions of their responses followed the rotation of the mon- key, and these responses were onsidered to represent relations between the test stimulus and bakground stimuli on the monkey or on the restraining devie that rotated with the monkey. An example of responsiveness of an egoentri neuron is shown in Figure 9A. This neuron responded strongly to both visual and auditory stimuli related to human movement, suh as walking, presented from the right anterior position. The responses to walking at the right anterior were weakened but not extinguished when all of the lights in the experimental room were turned off or when the monkey was prevented by a sreen from seeing the human walking (not shown). Thus, the diretionally differentiating responsiveness of this neuron may inlude both vision and audition. The responsiveness to a human walking at the right anterior (Aa) followed the monkey when it was rotated 45 to the left (Ab) or to the right (A). Eletroulograms and television monitoring of eye movement did not show orrelation between responses of these egoentri neurons and eye movement. The diretionally differentiating responsiveness of nine neurons remained in plae in the environment and did not follow the monkey. We designated these neurons as alloentri, beause the diretions of their responses did not follow the rotation of the monkey but remained onstant relative to the environment, and these responses were onsidered to represent relations between test stimulus and bakground stimuli fixed in the environment. An example of the responses of an alloentri neuron is shown in Figure 9B. This neuron also responded to visual or auditory stimulation aused by human ations (walking) presented from right anterior (Ba). Responses to human ation presented from the right anterior

8 314 HZPPOCAMPUS VO. 2, NO. 3, JUY 1992 A r m 1 - Y n 1 m T 5 - O UJ - O C Y O C Y) o E n m O P % a T v) v) C : -. T Y V v) Stimuli P r e s e n t e d - N -5- Fig. 5. Responses of visual seletive neuron. (A) Comparison of responses to human movement (from left anterior and right anterior diretions) and pure tone from six diretions. This neuron responded strongly when human sat down or stood up at right anterior. There were no responses to human movements from left side diretions or to auditory stimulation from any diretion. (B) Comparison of responses to various visual or auditory stimuli presented from right anterior diretion. This neuron responded seletively to human movement (sitting down or standing up). lhere were no responses to presentation of other stimuli, nor were there responses when all room lights were turned off and illumination was minimized (human movement in dark). Eah histogram, mean and SEM of response size of six presentations of eah stimulus. Hathed area in A and B, responses to human movement: filled area in A and B, responses to pure tone. Other desriptions as for Figures 1 and 3. A --+- P * 2 B A i a m m 8 Y u) a n 4 E r 2 (D a2 z * v) - 6 u) 6 Stimuli Presented Fig. 6. Responses of auditory nonseletive neuron. (A) Comparison of responses to stimuli presented around monkey. Visual stimuli (e.g., apple) were presented from left anterior, anterior, and right anterior diretions. Auditory stimuli (e.g., pure tone) were presented from six diretions. This neuron responded to auditory stimuli presented from left posterior diretion. (B) Comparison of responses to several auditory stimuli presented from left posterior diretion. This neuron responded to all auditory stimuli tested from the left posterior diretion with almost the same response magnitude. Eah histogram, mean and SEM of response size of six presentations of eah stimulus. Dotted area in A, responses to sight of apple; filled area in A and B, responses to pure tone. Calibration at right side of bottom histogram (responses to auditory stimuli from posterior diretion) in A, firing rate. Other desriptions as for Figures 1 and 3.

9 A l- T B 15 v) \ UJ P, 1 x n v) v 2 5 a En C i i o C (d 2-5 a, n T > C) a E. 3 (d r - T C - P v) P n Stimuli P r e s e n t e d Fig. 7. Responses of auditory seletive neuron. (A) Comparison of responses to human voie (hathed area), lap (dotted area), and 1, Hz pure tone (filled area) from various diretions. This neuron responded to several kinds of auditory stimuli when presented from posterior diretions, but not to visual stimuli (not shown) from anterior diretions. (B) Comparison of responses to various auditory stimuli presented from posterior diretion. This neuron responded strongly to sound of door being opened behind monkey, moderately to human voie, tapping, and lap, and weakly to step and rash. There were no responses to omputer-synthesized omplex sounds nor to pure tone. Other desriptions as for Figures 1 and 3. U N rn - N l- Y P, r + 7 n

10 31 6 HPPOCAMPUS VO. 2, NO. 3, JUY 1992 Table 1. Spontaneous Firing Rates and Response atenies of Eah Type of Neuron Visual Auditory Visual and Auditory NS S NS S NS S No. of neurons Spontaneous firing rate Range (spikesis) Mean f SEM (spikesis) 3.6 k f k f f.4 Response lateny Range (ms) k28 Mean i SEM (ms) 173 f k i i 24 No. of neurons NS, nonseletive neurons; S, seletive neurons. were weakened but not extinguished when all of the lights in the experimental room were turned off or when the monkey was prevented by a sreen from seeing the human movement (not shown). The responsiveness to walking at the right anterior diretion (Ba) remained fixed in the environment and did not follow the monkey when the monkey was rotated 45" to the left (Bb) or to the right (B). Response magnitude was dereased to about half of the original when the monkey was rotated 45" to the left and ould not see the human beause of the restrition of monkey's vision to?so". These responses to walking at the right are onsidered to be responses to the auditory omponent of the stimulus (step). The diretionally differentiating responsiveness of 12 other neurons was reversibly extinguished when the monkey was rotated. Reording sites Reording sites of all neurons sampled are depited in Figure 1. None of the three monkeys was signifiantly different from the other two, nor was any of the six hemispheres signifiantly different from any of the other five (hi-square test, P >. l ). Therefore, reording sites from both hemispheres of three monkeys are plotted on representative setions of the left hemisphere. The reording sites of eah type of neuron are depited in Figure 1 1 and summarized in Table 2. These neurons were loated throughout the rostroaudal extent of the HF, but tended to be loated audally (hi-square test, fy.1). There were no lear relations between the subfields in the H F and sensory modalities nor between the subfields C 5 m * -- 5 C : p : m eft Right U Fig. 8. Effets of hange of environment on responsiveness of the neuron shown in Figure 5. (A) Effets of hange of front panel loation. Strong responses to human movement (hathed area) at right anterior position (a) extinguished when front panel was moved from sight (b) or attenuated when it was moved about 5 m to left (). There were no responses to pure tone (filled area). (B) Analysis of effet of small loation hange of front panel on diretionally differentiating responsiveness. Responsiveness of this neuron was strongest when panel was in its usual position ( m), attenuated as panel was moved from usual position, and extinguished when panel was moved more than 1 m left or more than 15 m right. Eah point, mean and SEM of response size of six presentations of eah stimulus. Absissa, distane (m) from usual position.

11 MONKEY HPPOCAMPA RESPONSES TO SPACE / Tarnura et al A a Usual Situation - b 45"Rotation t o eft T 45"Rotation to Right - 1 \ B a Usual Situation - b 45"Rotation to eft 45"Rotation to Righ! Fig. 9. Effets of monkey rotation on responsiveness of neurons. (A) Egoentri neuron. (Aa) Strong responses to human walking at monkey's right anterior while the monkey was in its usual position. (Ab) Diretionally differentiating responsiveness to human movement after monkey was rotated 45" left from its initial position. (A) Diretionally differentiating responsiveness to human movement after monkey was rotated 45" right from its initial position. Note that responsiveness to human movement at right anterior (a) followed monkey (b and ). (B) Alloentri neuron. (Ba) Diretionally differentiating responsiveness to human walking; there were strong responses at right anterior. (Bb) Diretionally differentiating responsiveness to human walking (step sound) remained fixed in environment after monkey was rotated 45" left from initial position. (B) Diretionally differentiating responsiveness remained fixed in environment after monkey was rotated 45" right from initial position. Eah histogram in A and 3, mean and SEM of response size of five and three (A and B, respetively) presentations of eah stimulus. Hathed areas, responses to human movement. Filled irles in B, no responses to stimuli presented (spontaneous firing rate of this neuron was almost ). Other desriptions as for previous figures. and nonseletive or seletive nature of the neurons. n the subiular omplex, the number of neurons that hanged their diretionally differentiating responsiveness when the loation of the experimental apparatus was hanged exeeded the number of stable neurons that did not hange (8 of 1 hanged, 2 of 1 stable), but there were no lear differenes between the numbers of these neurons in other areas (dentate gyrus: 7 of 13 hanged, 6 of 13 stable; CA3: 8 of 16 hanged, 8 of 16 stable; CA: 1 of 17 hanged, 7 of 17 stable; EC: 2 of 2 hanged; PH: 1 of 1 hanged). The 11 egoentri neurons were loated in the dentate gyrus (n = 4) and in the CA3 (n = 2) and CA (n = 5) subfields. The nine alloentri neurons were loated in the dentate gyrus (n = 3), the CA subfield (n = 4), and the subiular omplex (n=2). The 12 neurons that reversibly extinguished their diretionally differentiating responsiveness with rotation of the monkey were loated in the dentate g y m (n = 2), CA3 (n = 3), CA (n = 6), and subiular omplex (n = 1). DSCUSSON Relations between diretionally differentiating responsiveness and sensory modalities The neurons with diretionally differentiating responsiveness tended to respond to visual stimuli presented from the right anterior diretion and to auditory stimuli presented from the posterior. This tendeny was observed in all three monkeys. The reasons for this disrimination are not entirely lear, but it seems reasonable that vision should depend on stimuli presented within the visual field, and audition might be more valuable to the monkey when stimuli annot be seen. Also, sine the experimenters were usually positioned at the right front of the monkey during most experiments (training sessions and reording sessions for operant tasks), this loation may have ahieved speial importane to the monkey. t is also probable that the monkey has spent more time with its attention direted to its right anterior than in any other

12 318 HPPOCAMPUS VO. 2, NO. 3, JUY 1992 Fig. 1. Reording sites of sampled neurons. Solid irles, neurons with diretionally differentiating responsiveness; dots, neurons without diretionally differentiating responsiveness. DG, dentate gyms; CA1 and CA3, hippoampal subfields; SUB, subiular omplex; EC, entorhinal ortex; PH, parahippoampal orties. Number below eah setion: distane (mm) anterior from interaural line. diretion in the experimental room. f learning or memory should be related to neuronal ativity, it might be possible that the inreased time of attention to that area aused the differene seen in the number of neurons, that is, more neurons represent relations to various ombinations of stimuli that were seen in the right anterior. The tendeny of neurons to respond to visual stimuli presented from the anterior diretion annot be explained by leftright differene of hemispheres sine there was no signifiant differene in the inidene of reording rate in the left and right hemispheres. t is also diffiult to believe that H F neurons have this tendeny by nature or that this tendeny was observed in all three monkeys by hane. Rather it seems reasonable that this tendeny may be formed in the experimental ondition of training and reording sessions. Existene of neurons that respond to visual or auditory stimuli or both is not surprising, sine the HF reeives input mainly from multimodal or so-alled supramodal ortial areas (Swanson, 1983; Amaral, 1987). This means that, onsistent with a previous study of the rat hippoampus (O'Keefe and Conway, 1978), some neurons in the HF of the monkey may be involved in the representation of relations between stimuli through both vision and audition. Nonseletive and seletive neurons Nonseletive neurons responded to different kinds of stimulus presentation in one or both sensory modalities with almost the same response magnitude. The responses of these stimulus nonseletive neurons might partly reflet attention. However, even if a visual stimulus was not presented, the monkey often glaned toward the spae where we believed it expeted the next stimulus to be put, thus direting its attention to that spae. Visual nonseletive neurons never responded onsistently in this situation. Therefore, these neurons did not respond even if the monkey did diret its attention to that spae and pereive only bakground stimuli.

13 MONKEY HPPOCAMPA RESPONSES TO SPACE / Tamura et al A 19 A 17 P H ' A 15 A 1 3 * Visual Nonseletive A Visual Seletive Auditory Nonseletive = Auditory Seletive Visual and Auditory Nonseletive Visual and Auditory Seletive A 1 1 Fig Reording sites of different types of neurons with diretionally differentiating responsiveness. Sensory modality of neurons: visual, triangles; auditory, squares; visual and auditory, irles. Seletivity of neurons: nonseletive, open; seletive, solid. Other desription as for Figure 1. A 9 Table 2. Reording Sites of Eah Neuron Type Reording Visual Auditory Visual and Auditory Total Diretional Neurons Total Neurons Sites (NS, S) (NS, S) (NS, S) (NS, S) Sampled 2 (1, 1) 22 (1, 12) 25 (12, 13) 11 (6, 5 ) 2 (1, 1) 1 (, 1) 81 (39, 42) NS, nonseletive neurons; S, seletive neurons: DG, dentate gyrus; CA3 and CA, subfields of hippoampus proper; SUB, subiular omplex; EC, entorhinal ortex; PH, parahippoampal orties

14 32 HPPOCAMPUS VO. 2, NO. 3, JUY 1992 However, these neurons required that visual stimulation, suh as an objet, be presented from a partiular loation. Seletive neurons responded more to some stimuli than to others in at least one sensory modality. Hippoampal neurons with seletive (or differential) responses have been previously reported (Wible et al., 1986; Rolls et al., 1989; Wiener et al., 1989; Creutzfeldt, 199; Vidyasagar et al., 1991; Tamura et al., 1992). Furthermore, there were some reports (Creutzfeldt, 199; Vidyasagar et al., 1991) of hippoampal neurons that were ativated by presentation of partiular human ations during what they alled the raisin show. This report is very similar to the present results in that seletive neurons tended to respond to human ation. Two visual information-proessing pathways have been reported (Mishkin, 1972). One is the dorsal pathway, whih is involved in analysis of the loation of objets, and the other is the ventral pathway, whih is involved in analysis of the physial properties of objets. For deteting a partiular loation in whih a partiular objet is loated, interation between these two pathways is neessary. The H F and PH would be positions in whih suh interation ould our, sine anatomial studies (Jones and Powell, 197; Van Hoesen, 1982; Amaral, 1987; Tranel et al., 1988) show inputs to the H F via the PH from the visual and auditory assoiation orties after proessing pereption of physial properties (Rolls et al., 1977; Gross et al., 1979; Sato et al., 198; Fuster and Jervey, 1982; Ungerleider and Mishkin, 1982; Desimone et al., 1984; Miyashita, 1988; Miyashita and Chang, 1988), and from the posterior parietal and prefrontal orties, after analysis of spatial information (Teuber, 1964; Pohl, 1973; Petrides and versen, 1979; Andersen and Mountastle, 1983; Passingham, 1985; Andersen, 1987; Goldman-Raki, 1987). Responsiveness of the visual seletive neurons might reflet results of suh interation. t has been reported that bilateral lesions of the HF impaired performane in a spatial memory task (Parkinson et al., 1988) that required objet-plae assoiation, so we an imagine that assoiation between a stimulus and its loation in spae, as refleted in responses of the seletive group, would be formed in the HF. Dependene of responsiveness on relations between test stimulus and bakground ues Responsiveness of neurons that was not modulated by hanging the loation of experimental apparatus may represent relations between the test stimulus and bakground stimuli, other than those fixed on the apparatus. Responsiveness of the neurons that was hanged by movement of the apparatus may represent relations between the test stimulus and bakground stimuli fixed on the apparatus, and these might orrespond to neurons with responses in the frame-of-referene of alloentri oordinates reported by Feigenbaum and Rolls (1991). t is diffiult to interpret responsiveness of neurons that hange irrespetive of movement of the apparatus. t is possible that newly appearing diretionally differentiating responsiveness reflets representation of other relations between the test stimulus and bakground stimuli, and this phenomenon might orrespond to the firing of HF plae ells in the rat that an be hanged by adding, or removing, intramaze or extramaze ues (Muller et al., 1987). Responsiveness of neurons that was extinguished by move- ment of the apparatus might represent relations between test stimuli, bakground stimuli fixed on the apparatus, and other bakground stimuli. However, other fators might have influened the responsiveness, suh as suppression by partiular stimuli that ould not be seen in the usual position. Responsiveness of egoentri neurons is onsidered to represent relations between the test stimulus and bakground stimuli fixed to the monkey or the restraining devie, whereas responsiveness of alloentri neurons is onsidered to involve representation of relations between the test stimulus and bakground stimuli fixed in the room. Responsiveness of neurons that is extinguished by rotation of the monkey may be involved in representation of relations between the test stimulus, bakground stimuli apparent after rotation of the monkey, and stimuli fixed in the room, though it is also possible that some other fators might influene the responsiveness. To represent a visible objet in an egoentri (head-entered) framework, it is neessary to ombine and ompute the information about loation of the image of the objet on a visual field, eye position, nek position, and so on. Some neurons in the neoortial areas have been reported to respond to retinotopi spae, eye position, or ombinations of these two fators (for review, see Andersen, 1987). However, there is still no evidene to suggest neurons in the neoortial areas that represent aspets of spatial fators in the egoentri framework or alloentri framework. Present results suggest that neurons in the monkey H F or PH may relate stimuli in an egoentri framework or an ailoentri framework. Therefore, some omputation or onversion of information from the primary step of spatial representation (suh as retinotopi spae) to a more ompliated spatial representation (suh as egoentri or alloentri spae) may be made between the sensory assoiation orties and the H F and PH, or within the H F and PH. Further studies are neessary to investigate suh omputation or onversion. ACKNOWEDGMENTS The authors wish to thank Dr. A. Simpson, Showa University, for help in preparing the manusript, and Mrs. M. Yamazaki and Mrs. A. Tabuhi for typing. Supported in part by the Japanese Ministry of Eduation, Siene and Culture, Grants-in-Aid for Sientifi Researh and , by a Biosiene Grant for nternational Joint Researh Projet from the NEDO, Japan, and by Brain Siene Foundation, Japan. Referenes Amaral, D. G. (1987) Memory: Anatomial organization of andidate brain regions. n Handbook of Physiology, The Nervous System, V.B. Mountastle, ed., vol. V, pp , Amerian Physiology Soiety, Bethesda, MD. Andersen, R. A. (1987) lnferior parietal lobule funtion in spatial pereption and visuomotor integration. n Handbook of Physiology, The Nervous System, V.B. Mountastle, ed., vol. V, pp , Amerian Physiology Soiety, Bethesda, MD. Andersen, R. A,, and V. B. Mountastle (1983) The influene of the angle of gaze upon the exitability of the light-sensitive neurons of the posterior parietal ortex. J. Neurosi Breese, C. R., R. E. Hampson, and S. A. Deadwyler (1989) Hip-

15 MONKEY HPPOCAMPA RESPONSES TO SPACE / Tamura et al. 321 poampal plae ells: Stereotypy and plastiity. J. Neurosi. 9: Cahusa, P. M. B., Y. Miyashita, and E. T. Rolls (1989) Responses of hippoampal formation neurons in the monkey related to delayed spatial response and objet-plae memory tasks. Behav. Brain Res. 33: Cave, C. B., and. R. Squire (1991) Equivalent impairment of spatial and nonspatial memory following damage to the human hippoampus. Hippoampus 1: Creutzfeldt,. D. (199) Neuronal responses in human and monkey temporal lobe during memory tasks: Compartmentalization of memory funtions. n Vision, Memory and the Temporal obe, E. wai and M. Mishkin, eds., pp , Elsevier, New York. Desimone, R., T. D. Albright, C. G. Gross, and C. Brue (1984) Stimulus-seletive properties of inferior temporal neurons in the maaque. J. Neurosi. 4: Eihenbaum. H., A. Fagan, P. Mathews, and N. Cohen (1988) Hippoampal system dysfuntion and odor disrimination learning in rats: mpairment or failitation depending on representational demands. Behav. Neurosi. 12: Feigenbaum, J. D., and E. T. Rolls (1991) Alloentri and egoentri spatial information proessing in the hippoampal formation of the behaving primate. Psyhobiology 19:21-4. Foster, T. C., C. A. Castro, and B.. MNaughton (1989) Spatial seletivity of rat hippoampal neurons: Dependene on preparedness for movement. Siene 244: Fukuda, M., T. Ono, H. Nishino, and K. Sasaki (1986) Visual responses related to food disrimination in monkey lateral hypothalamus during operant feeding behavior. Brain Res. 374: Fuster, J. M., and J. P. Jervey (1982) Neuronal firing in the inferotemporal ortex of the monkey in a visual memory task. J. Neurosi. 2: Goldman-Raki, P. S. (1987) Ciruitry of primate prefrontal ortex and regulation of behavior by representational memory. n Hundbook of Physiology, The Nervous System, vol. V, V. B. Mountastle, ed., pp , Amerian Physiology Soiety, Bethesda, MD. Gross, C. G., D. B. Bender, and G.. Gerstein (1979) Ativity of inferior temporal neurons in behaving monkeys. Neuropsyhologia 17: Jones, E. G., and T. P. S. Powell (197) An anatomial study of onverging sensory pathways within the erebral ortex of the monkey. Brain 93: Mishkin, M. (1972) Cortial visual areas and their interations. n Brain and Human Behavior, A. G. Karzmar and J. C. Eles, eds., pp , Springer-Verlag, Berlin. Miyashita, Y. (1988) Neuronal orrelate of visual assoiative longterm memory in the primate temporal ortex. Nature Miyashita, Y., and H. S. Chang (1988) Neuronal orrelate of pitorial short-term memory in the primate temporal ortex. Nature 331:68-7. Miyashita, Y., E. T. Rolls, P. M. B. Cahusa, H. Niki, and J. D. Feigenbaum (1989) Ativity of hippoampal formation neurons in the monkey related to a onditional spatial response task. J. Neurophysiol. 61 : Morris, R. G. M., P. Garrud, J. N. P. Rawlins, and J. O Keefe (1982) Plae navigation impaired in rats with hippoampal lesions. Nature Muller, R. U., J.. Kubie, and J. B. Rank Jr. (1987) Spatial firing patterns of hippoampal omplex-spike ells in a fixed environment. J. Neurosi. 7: Nishijo, H., T. Ono, and H. Nishino (1988a) Topographi distribution of modality-speifi amygdalar neurons in alert monkey. J. Neurosi. 8: Nishijo, H., T. Ono, and H. Nishino (1988b) Single neuron responses in amygdala of alert monkey during omplex sensory stimulation with affetive signifiane. J. Neurosi. 8: O Keefe, J. (199) A omputational theory of the hippoampal ognitive map. Prog. Brain Res. 83: O Keefe, J., and J. Dostrovsky (1971) The hippoampus as a spatial map: Preliminary evidene from unit ativity in freely-moving rat. Brain Res. 34: O Keefe, J., and Conway, D. H. (1978) Hippoampal plae units in the freely moving rat: Why they fire where they fire. Exp. Brain Res. 3 1 : O Keefe, J., and. Nadel (1978) The Hippoampus as a Cognitive Map, Oxford, Clarendon, UK. Olton, D. S., J. A. Walker, and F. H. Gage (1978) Hippoampal onnetions and spatial disrimination. Brain Res. 139: Ono, T., H. Nishino, K. Sasaki, M. Fukuda, and M. Muramoto (198) Role of the lateral hypothalamus and the amygdala in feeding behavior. Brain Res. Bull. 5 (Suppl. 4): Ono, T., H. Nishino, K. Sasaki, M. Fukuda, and M. Muramoto(l981) Monkey lateral hypothalami neuron responses to sight of food, and during bar press and ingestion. Neurosi. ett Ono, T., R. Tamura, H. Nishijo, K. Nakamura, and E. Tabuhi(l989) Contribution of amygdalar and lateral hypothalami neurons to visual information proessing of food and nonfood in monkey. Physiol. Behav. 45: Ono, T., K. Nakamura, M. Fukuda, and R. Tamura (1991) Plae reognition responses of neurons in monkey hippoampus. Neurosi. ett. 121: Parkinson, J. K., E. A. Murray, and M. Mishkin (1988) A seletive mnemoni role for the hippoampus in monkeys: Memory for the loation of objets. J. Neurosi. 8: Passingham, R. E. (1985) Memory of monkeys (Maaa mulatta) with lesions in prefrontal ortex. Behav. Neurosi. 99:3-21. Petrides, M., and S. D. versen (1979) Restrited posterior parietal lesions in the rhesus monkey and performane on visuospatial tasks. Brain Res. 161: Pohl, W. (1973) Dissoiation of spatial disrimination defiits following frontal and parietal lesions in monkeys. J. Comp. Physiol. Psyhol. 82: Rolls, E. T., S. J. Judge, and M. K. Sanghera (1977) Ativity of neurons in the inferotemporal ortex of the alert monkey. Brain Res. 13: Rolls, E. T., Y. Miyashita, P. M. B. Cahusa, R. P. Kesner, H. Niki, J. D. Feigenbaum, and. Bah (1989) Hippoampal neurons in the monkey with ativity related to the plae in whih a stimulus is shown. J. Neurosi. 9: Rupniak, N. M. J., and D. Gaffan (1987) Monkey hippoampus and learning about spatially direted movements. J. Neurosi. 7: Sato, T., T. Kawamura, and E. wai (198) Responsiveness of inferotemporal single units to visual pattern stimuli in monkeys performing disrimination. Exp. Brain Res. 38: Smith, M.., and B. Milner (1981) The role of the right hippoampus in the reall of spatial loation. Neuropsyhologia 19: Smith, M.., and B. Milner (1984) Differential effets of frontal-lobe lesions on ognitive estimation and spatial memory. Neuropsyhologia 22: Smith, M.., and B. Milner (1989) Right hippoampal impairment in the reall of spatial loation: Enoding defiit or rapid forgetting? Neuropsyhologia 27: Squire,. R., A. P. Shimamura, and D. G. Amaral(l989) Memory and hippoampus. n Neural Models of Plastiiry, H. Byrne and W. Berry, eds., pp , Aademi Press, New York. Sutherland, R. J., and J. W. Rudy (1989) Configural assoiation theory: The role of the hippoampal formation in learning, memory and amnesia. Psyhobiology 17: Swanson,. W. (1983) The hippoampus and the onept of the

16 322 HPPOCAMPUS VO. 2, NO. 3, JUY 1992 limbi system. n Neurobiology of ihe Hippoampus, W. Seifert, ed., pp. 3-19, Aademi Press, New York. Tamura, R., T. Ono, M. Fukuda, and K. Nakamura (199) Reognition of egoentri and alloentri visual and auditory spae by neurons in the hippoampus of monkeys. Neurosi. ett. 19: Tamura, R., T. Ono, M. Fukuda, and H. Nishijo (1992) Monkey hippoampal neuron responses to omplex sensory stimulation during objet disrimination. Hippoampus 2: Teuber, H. (1964) The riddle of frontal lobe funtion in man. n The Frontal Granular Cortex and Behavior, J. M. Warren and K. Akert, eds., pp , MGraw-Hill, New York. Tranel, D., D. R. Brady, G. W. Van Hoesen, and A. R. Damasio (1988) Parahippoampal projetions to posterior auditory assoiation ortex (area Tpt) in Old-World monkeys. Exp. Brain Res. 7~ Ungerleider, W., and M. Mishkin (1982) Two ortial visual systems. n The Analysis of Visual Behavior, D. J. ngle, M. A. Goodale, and R. J. W. Mansfield, eds., pp , MT Press, Cambridge, MA. Van Hoesen, G. W. (1982) The parahippoampal gyms: New obser- vations regarding its ortial onnetions in the monkey. Trends Neurosi. 5: Vidyasagar, T. R., E. Salzmann, and. D. Creutzfeldt (1991) Unit ativity in the hippoampus and the parahippoampal temporobasal assoiation ortex related to memory and omplex behaviour in the awake monkey. Brain Res. 544: Watanabe, T., and H. Niki (1985) Hippoampal unit ativity and delayed response in the monkey. Brain Res. 325: Wible, C. G., R.. Findling, M. Shapiro, E. J. ang, S. Crane, and D. S. Olton (1986) Mnemoni orrelates of unit ativity in the hippoampus. Brain Res Wiener, S.., C. A. Paul, and H. Eihenbaum (1989) Spatial and behavioral orrelates of hippoampal neuronal ativity. J. Neurosi. 9: Yamatani, K., T. Ono, H. Nishijo, and A. Takaku (199) Ativity and distribution of learning-related neurons in monkey (Maaa fusaia) prefrontal ortex. Behav. Neurosi Zola-Morgan, S., and. R. Squire (1986) Memory impairment in monkeys following lesions limited to the hippoampus. Behav. Neurosi. 1: