An Afferent Hippocampal Fiber System in the Fornix of the Monkey

CHARLES L. VOTAW AND EDWARD W. LAUER Department of Anatmy, The University of Michigan, Ann ATboT, Michigan

In a previous investigation (Votaw, ’60b) and weighed from 2.2 to 5.3 kgs. Seven it was noted that, in spite of massive bilat- were used only for this investigation and eral removal of the hippocampal formation seven others were used as well in another (Ammon’s horn and dentate gyms), there investigation on the function of the hippo- was a considerable percentage of the fi- campus and are included in this study bers in the fornix, anterior to the hippo- because the lesions placed were ideally carnpal commissure, that did not degener- suited for comparison with the former ate. This would imply, since the lesions seven cases. to all intents and purposes eliminated all All animals were subjected to aseptic possible fibers going into the fornix from surgical procedures. The only surgical pro- the region of the temporal lobe, the pres- cedure on seven was the production of the ence of fibers in the fornix which are a€- lesions to be reported here; lesions were ferent with respect to the hippocampus, placed in the remaining seven after other and further, that these fibers are not com- studies had been carried out. General missural in nature. In another study anesthesia was used and consisted of (Votaw, ’60a), it was shown that lesions ether, sodium pentobarbital (20-25 mg/kg in the septa1 area would produce a de- of body weight given intravenously) or a generation which could be followed from combination of the two anesthetics, ether the area of the lesion in a caudal direction, being used to supplement a somewhat coursing through the body of the fornix, lower dose of sodium pentobarbital. Fol- from which it would enter the fimbria and lowing closure of the surgical wounds, the alveus. The final termination could not be animal was given a dose of penicillin- determined with certainty. chloramphenicol intramuscularly and re- There is accumulating in the current turned to an isolated cage. The monkeys literature a considerable body of evidence were observed daily, and additional anti- concerning a hippocampal afferent sys- biotics given if it was thought necessary. tem in the fornix (see Discussion). Some Wound infection was limited to one case, of this information is gained from morpho- and a post-operative diarrhea was evident logical experimentation, but the majority in two others. These complications were of the more suggestive evidence is based easily controlled. on physiological data. To the authors’ In four of the animals, a dorsal ap- knowledge there is no information avail- proach was used. A craniectomy was ac- able with respect to the monkey about this complished over the motor area. Then a fiber system, except that quoted above small portion of the medial cerebral cortex (Votaw, ’60a). In an attempt to investi- was removed until, with gentle retraction, gate more fully the existence of such a the corpus callosum could be visualized. fiber system in monkeys, and, if possible, A small opening was next made in the to describe some of its relations, the pres- corpus callosum exposing the easily iden- ent study was undertaken. tified bodies of the fornices. Using a spe- cially prepared hooked knife, the bodies of MATERIALS AND METHOD the fornices were lifted away from the Fourteen monkeys (Macaca mulatta) were used in these experiments. The ani- 1This work was supported by National Science Foundation Grant 6-11433 for which the authors mals were unselected with respect to sex, express their appreciation.

195 196 CHARLES L. VOTAW AND EDWARD W. LAUER thalamus, and then sectioned. In one ad- further clarification was necessary, they ditional animal, the opening was made were mounted and examined. more anteriorly, and the lesion produced Precautions were taken to eliminate the by aspirating the septal area and the artifacts which are inherent in the Marchi anterior columns of the fornix. technique. However, a small proportion of In two cases, electrodes were placed normal fibers were stained, and there was stereotaxically in the septal areas. Follow- the appearance of some “Marchi dust” ing stimulation of the septal area for an- in the sections. The only degeneration to other study, electrocoagulation was carried be described will be that which could be out at the tip of the electrode placements. followed serially from the lesion, and In each of the remaining seven animals, which had the characteristic black irregu- a craniectomy was performed over the su- lar shapes of Marchi degeneration. Termi- perior part of the temporal lobe of the nal degeneration is not revealed by this brain. After exposure of the superior and method, and, accordingly, nothing but the middle temporal gyri, an opening was most tentative suggestion will be made made through these gyri, using the su- about the exact termination of the fibers perior temporal fissure as a guide. This to be described. Primary concern in this procedurc was followed until the temporal study is with the make-up of the fornix. horn of the lateral ventricle was opened Therefore, detailed descriptions of the de- and the hippocampus could be seen as the generation in the projections of the fornix ventromedial wall of the ventricle. Elec- system which have been presented else- trodes were then placed visually and the where (Votaw, ’60b) will not be given hippocampus explored by electrical stimu- here, and reference will be made to those lation, with results which will be presented results only for comparison. in another paper. Following this proce- The term hippocampal formation (or dure, lesions were made in the hippocam- hippocampal complex), as it is used in pus using either an aspirator or a special the literature, includes the dentate gyrus, dissection instrument. This procedure was cornu Ammonis, and subiculum. At times carried out in two stages, one for each the term has been extended to include side, with an interval of from 2-11 weeks. the presubiculum and even the entorhinal In four of these seven animals, this was area. The term hippocampus is usually the only procedure. In the othcr three, used to indicate the cornu Ammonis alone the procedure just described was preceded although some earlier authors have used by a fornicotomy as outlined previously. it as synonymous with hippocampal for- Thus, in four animals there were bilateral mation. In this study, the term hippo- lesions of Ammon’s horn, and in three ani- campal formation (or hippocampal com- mals there were bilateral lesions of Am- plex or hippocampal area) will be used mon’s horn plus an additional section of to indicate the complex of dentate gyrus, the fornix bilaterally. Ammon’s horn and subiculum. The pre- After survival periods of from 21 days subiculum and entorhinal area will be to 3.5 months, the animals were killed by treated as separate entities. The terms an overdose of evipal and perfused intra- hippocampus and cornu Ammonis (Am- cardially with 10% formalin. The brains mon’s horn) will be used interchangeably were removed immediately and addition- ally fixed for a period of not less than two RESULTS weeks in 10% formalin. They were then Lesions of septal nuclei. In one case sectioned in 2 mm slices which were left there was complete removal of the medial in the fixative fluid for an additional week. and lateral septal nuclei of the left side. They were then stained according to the This lesion also interrupted the column of Marchi technique as modified by Swank the homolateral fornix and slightly in- and Davenport (’35). Sections were cut jured the column of the other side. The at 30 to 50 CI depending on the friability anterior commissure was involved and of the tissues. Every fifth section was degeneration was observed within it. In mounted and studied microscopically. The addition, there were the lesions in the intermediate sections were stored, and if corpus callosum and in the frontal lobe. HIPPOCAMPAL AFFERENT FIBERS 197

In the other two cases there were small tion should not be involved at all, have a circumscribed lesions bilaterally, but asym- large amount of such dust. Therefore, it metrically located in the medial and lat- is not known to what position in the hippo- eral septal nuclei. The electrode paths campal formation the fibers under study could be followed to the area of the lesion, go. Since there was no degeneration be- but the amount of degeneration about yond the region of the hippocampal for- them was insignificant. The anterior com- mation (subiculum, presubiculum, hippo- missure and the anterior columns of the campal gyrus), it is assumed that the fi- fornix were not involved in these lesions. bers terminated somewhere in that forma- The degeneration resulting from these tion. lesions tended to be largely ipsilateral, as In all three of these cases, but espe- can be determined from the first case. cially in the case in which the septal nuclei However, there was a significant number were entirely removed, degeneration can of degenerated fibers on the contralateral be followed in the postcommissural fornix side. Degeneration could be followed in a column to the medial mammillary nucleus posterior direction in the bodies of the for- and other hypothalamic nuclei. No post- nices. This degeneration was somewhat mammillary degeneration could be found. diffusely dispersed throughout the body Some degeneration was found as well in of the fornix but the region of heaviest the diagonal band of Broca and a very few concentration was found to be on the degenerated fibers were observed in the lateral side of the fornix. Anteriorly, where lateral region of the hypothalamus. The the fornix forms a compact, circular bun- latter were thought to be in the medial dle, the degeneration was found in the forebrain bundle but the data at this time lateral half (see Votaw, ’60, fig. 2b). More are insufficient to draw any conclusions. posteriorly, where the fornix begins to Section of the fornix. (Figs. lA, 2A, 3A.) flatten out into a semi-sheet of fibers, the In three cases the bodies of the fornices degeneration was diffuse but more con- were completely severed bilaterally. In two centrated in the lateral part of the bundle. of these, the section was midway between From this position, the degeneration the hippocampal commissure and the sep- continued into the fimbria along the me- tal nuclei. In the third, the lesion de- dial side of the hippocampus. Degenera- stroyed, in addition to the bodies of the tion could be seen over the entire extent fornices, the anterior columns bilaterally of the hippocampus, reaching its most and involved the septal area to some de- anterior portions. The concentration of gree (fig. 1A). To be sure, there are some degenerated tissue, however, was much fibers which cross just posterior to the more dense in the posterior regions, and septal nuclei as described by McLardy gradually decreased as the pes hippo- (’55a); however, it was evident from the campi were reached. At all levels of the examination of these brains that the main hippocampal formation, degenerated fibers intrahippocampal commissure was left in- could be seen to turn out from the fimbria tact. In a fourth animal, the lesion in- and to enter the alveus bundle. Also, on volved only the lateral two-thirds of the careful observation, some degenerated fi- body of the right fornix bers could be seen turning out of the In addition, these brains had a large alveus and entering stratum oriens of lesion in the corpus callosum. However, Ammon’s horn. although the degeneration from this lesion Small granules could be seen through- sometimes extended above the hippocam- out the pyramidal cell layer of Ammon’s pal commissure, at no time was there evi- horn and the polymorphic cell layer of dence of degeneration within this latter the dentate gyrus, as well as a few such commissure. There was, in addition, a uni- granules in the polymorphic cell layer of lateral Iesion in the medial wall of the Ammon’s horn. Many of these granules frontal lobe. These animals had varying looked like Marchi dust. It was also degrees of paralysis of the lower extremity noticed that brains prepared according opposite the side of the lesion. to the same technique but for another Posterior to the section, degenerated fi- study in which the hippocampal forma- bers could be seen on the lateral side of 198 CHARLES L. VOTAW AND EDWARD W. LAUER

Fig. 1 Illustrations of two of the lesions described in the text. (Marchi stain; frontal sections; x 12) (A) Lesion located at the posterior edge of the arch of the postcommissural fornix column. The entire vertical column of the fornix has been interrupted. For. int.: foramen interventriculare; Fornix, ant. col.: fornix, anterior column. (B) Lesion located in the hippocampus, showing the removal of the lateral layers of this structure. Arrows indicate extent of lesion. HIPPOCAMPAL AFFERENT FIBERS 199 the body of the fornix, passing the lateral In contrast, anterior to the lesion de- side of the region of the hippocampal generation was distributed in the medial commissure, and continuing into the fim- regions of the circular body of each fornix. bria and alveus. The place of termination Unlike the more diffusely distributed fi- of these fibers is also in doubt; the findings bers in the fornices posterior to the lesion, were the same as those described above this degeneration was confined to the for the septa1 lesions (figs. 2A, 3A). medial half of the body of each fornix.

Fig. 2 Illustrations of the bodies of the fornices in the two animals whose lesions are shown in figure 1. (Marchi stain; frontal sections; X 27) (A) Fornix, anterior to hippocampal commissure, from the animal whose lesion is shown in figure 1A. Note the degeneration granules are located diffusely throughout the body of the fornix with some in- crease in concentration in the lateral portion of the pathway. (B) Fornix in same location as A, but from the animal whose lesion is shown in figure 1B. Note the localization of the degeneration in the center portion of the fiber bundle. 200 CHARLES L. VOTAW AND EDWARD W. LAUER

Degeneration could be followed anteriorly of which are described elsewhere (Votaw, to the region of the septal nuclei where '60b). degeneration granules were found in the Commissural fibers of the corpus cal- lateral septal nuclei. The large columns losum were degenerated and could be fol- of the fornices which course through the lowed to symmetric points on each side of caudal part of the medial septal nuclei the brain. The pyramidal tract was heavily had a high concentration of degenerated degenerated on the side of the lesion and tissue within them. Terminal degenera- its fibers could be traced through the tion could also be found in the diagonal internal capsule, the peduncles, and the band of Broca, the medial mammillary pyramids. The spinal cords of these ani- body, the hypothalamus, etc., the details mals were not prepared according to the

Fig. 3 Illustrations of the fornices of the same two animals at almost the junction of the body of the fornix with the fimbria. (Marchi stain; frontal sections; x 23.) (A) Fornix from animal with lesion shown in figure 1A. (B) Fornix from animal with lesion shown in figure 1B. Again note the distribution of the degeneration granules. HIPPOCAMPAL AFFERENT FIBERS 20 1

Marchi technique, and thus degeneration These pathways were distinct from those was not examined further than mid- described above. medulla regions. Lesions of Ammon’s horn plus Lesions of Ammon’s horn. (figs. lB, section of the fornix 2B, 3B.) These lesions involved the outer layers of Ammon’s horn (the alveus, the In three animals the fornices were sec- pyramidal cells, and the polymorphic tioned bilaterally. Later examination dem- cells) (fig. 1B) and extended for about onstrated that both fornices had been three-fourths of the length of the hippo- completely sectioned at a point about one- campus. The lesions involved the areas third the distance from the septal nuclei to the hippocampal comrnissure. The com- CA-1, CA-2, CA-3, but only a small part of missure was not involved in any of these the continguous CA-4 and prosubiculum. lesions. After recovery from the fornix The fimbria and the entire dentate gyms section operation, one hippocampus was escaped injury. exposed and a lesion placed in it. A simi- From the borders of the lesions degener- lar destruction was produced in the hippo- ated fibers could be traced into the alveus campus of the other side at a still later (fig. 1B). Some of these fibers went across date. The destruction in the temporal the alveus to the fimbria where they turned lobe and Ammon’s horn was comparable caudally and then superiorly along the to that described in the previous group of length of the hippocampus; others ran at experiments. Again, the dentate gyms and a more oblique angle along the hippo- the fimbria escaped direct injury. campus in the alveus bundle and entered Degeneration anterior to the section of the fornix at the termination of Ammon’s the fornices was similar in all respects to horn. On reaching the arch of the for- that described for the second group of nix, these degenerated fibers could be seen lesions. Similarly, degeneration from the on the medial side of the fornix body temporal lobe lesions in diencephalic and (fig. 3B). In the hippocampal commissure brain stem regions was the same as that there was a high concentration of fibers described in the previous group of lesions. which contained degeneration granules. The hippocampal commissure was mark- The fibers that did not cross proceeded edly, though not completely, degenerated. rostrally through the body of the fornix on In these cases the one significant find- its medial side (fig. 2B). They did not ing was in the body of the fornix posterior appear to be diffusely arranged, and there to the section in this structure. This re- was practically no degeneration of the lat- gion lies between two lesions which have eral side of the fornix body. been described as producing characteris- These fibers continued forward to the tically located degeneration in the bodies septum, where there was a large amount of the fornices. Examination of the fornix of degeneration in the lateral septal nu- anterior to the hippocampal commissure cleus, and some in the medial septal showed that the entire fornix contained nucleus. Some fibers entered the precom- degenerated fibers, with a heavy concen- missural fornix and went to the diagonal tration medially, and somewhat more dif- band of Broca. The presence of degener- fusely distributed degeneration in the lat- ated fibers in the postcommissural fornix, eral portion of this fiber bundle. the stria medullaris thalami, or any of the In summary, (fig. 4) it has been noted thalamic or hypothalamic nuclei could not that lesions of the septal nuclei and the be established. No degenerated fibers anterior part of the bodies of the fornix could be traced into the midbrain from the give rise to degenerated fibers which ap- fornix structures. On the contrary, the pear to run caudally in the lateral portions postcommissural fornix appeared entirely of the fornix. This degeneration is some- normal in these preparations. what diffusely distributed, with heavier Temporotegmental, temporopontine, and concentration in the lateral part of the temporopulvinar fibers could be followed body of the fornix. These fibers continue from the region of the temporal lobe le- past the hippocampal commissure, enter sions to their respective terminations. the fimbria, then the alveus, and appear 202 CHARLES L. VOTAW AND EDWARD W. LAUER

collosum

I I / I

Fig. 4 Diagrammatic illustrations of frontal sections of the fornix at levels (A) of the anterior thalamic nucleus and (B) of the habenula. The right side of the drawings shows the location of degenerating fibers following septal lesions; the left side, the location after hippocampal lesions. to turn into Ammon’s horn and the dentate appears to be restricted to the anterior gyrus at all levels of the hippocampal parts of the limbic lobe. formation. The degeneration is bilateral with a much heavier concentration of de- DISCUSSION generation granules on the side of the There has been an increasing awareness lesion. of the fact that there are hippocampal Lesions in Ammon’s horn which do not afferent fibers in the fornix. Prior to the involve the fimbria or the dentate gyrus time of Ram6n y Cajal, and, indeed, in give rise to degenerated fibers which run many more recent descriptions of the caudally along the hippocampus and then brain, the fornix is described as a band enter the fornix on its medial side, with of fibers running from the diencephalon many of the fibers crossing over in the to the hippocampus. Ram6n y Cajal (’11) hippocampal commissure. Those which do described the course of the fornix fibers not cross run forward in the medial por- as running in the opposite direction - tion of the body of the fornix to terminate toward the mammillary body. It has been in anterior regions of the rhinencephalon, assumed by many, since that time, that the namely, the septal nuclei and the diagonal fornix is the only, and entirely, efferent band of Broca. outlet of the hippocampal formation. It is further noted that lesions in Am- Brodal (’47) repeats this suggestion in man's horn, which do not destroy a part his excellent review of the hippocampus of the dentate gyrus, produce a different and the sense of smell, but leaves room terminal distribution of degeneration than for the possibility that there are septo- that which results from section of the hippocampal fibers, and that there might fornix. When the fornix is sectioned, de- be conduction in a direction opposite from generation occurs in the stria medullaris that which had been generally assumed. thalami, postcommissural fornix, hypo- Crosby (’17) on the basis of her de- thalamic nuclei, etc., but degeneration sec- scription of the normal anatomy of the ondary to lesions in Ammon’s horn alone reptile, suggested that the medial septal HIPPOCAMPAL AFFERENT FIBERS 203 nucleus might be a way station for ascend- tween the septal area and hippocampus of ing impulses to the hippocampus. Loo the rat. ('31), Young ('36), Fox ('43), and Lauer The evidence presented here strongly ('45) all made similar suggestions from suggests a hippocampal afferent fiber sys- the study of normal material of the opos- tem in the fornix of the monkey. From sum, rabbit, cat, and monkey, respectively. previous accounts and from the data pre- However, these were all descriptions of sented here, the origin of these fibers may normal material prepared for microscopic be the septal area (specifically, the medial study, and did not incorporate experimen- septal nuclei) or the hypothalamic nuclei tal proof of the direction of conduction. which give off fibers which pass by way Mettler ('43) noted in the monkey and of the medial forebrain bundle to enter the Rose and Woolsey ('43) in the rabbit that fornix. The lesions of this study were not massive ablations of the cerebral hemi- in the proper location to determine the spheres did not result in any retrograde origin of the hippocampal afferent fibers. cellular damage to the septal nuclei un- McLardy ('55a, '55b) suggested that there less the hippocampus was included. Other are fibers in the fornix, some of which retrograde cell degeneration studies by cross the midline just posterior to the sep- Sprague and Meyer ('50) in the rabbit, tal area, and interrelate the septal nuclei Daitz and Powell ('54) in the rat, and and the isocortex near the anterior part McLardy ('55a) in the monkey showed of the temporal lobe. He also described that lesions of the fornix resulted in rather these fibers as present in the medial part marked retrograde cell degeneration of of the fornix. The fibers extending cau- the medial septal nucleus, implying the dally in the fornix as described here are presence of caudally running fibers in the more characteristically located on its lat- fornix. eral side. However, unlike the rostrally Gerebtzoff ('39, '41) suggested that the running fibers in the fornix, which were dorsal fornix of the guinea pig contained rather definitely located on the medial side fibers having their cells of origin in the of this bundle, the posteriorly coursing diagonal band of Broca and which termi- fibers were more diffuse, being more con- nated in Ammon's horn. Cragg and Ham- centrated on the lateral side, but with lyn ('57, '59) found fibers in the dorsal some degeneration on the medial side. fornix, arising from the septal nuclei of These medial fibers could be the same as the rabbit, and terminating in the pre- those described by McLardy. No degenera- subiculum but not in Ammon's horn. tion was observed in the temporal cortex, Morin ('50), using the guinea pig, noted but this failure may be due to the tech- that lesions in the precommissural system nique used. McLardy's data depended on ventral to the genu produced Marchi de- retrograde cell degeneration. Certainly, the generation in the fornix which went pos- observations presented here do not con- teriorly as far as the hippocampal com- tradict the evidence of a septotemporal missure and the fimbria. Guillery ('57) pathway as described by McLardy, but a and Cragg ('61) placed lesions in the much heavier lateral degeneration in the hypothalamus of the rat, rabbit, and cat fornix, resulting from septal and fornix and then studied the resulting degenera- lesions, passes caudally in the fornix. The tion with the Nauta-Gygax technique. difference between the results of this study They observed fibers ascending in the me- and that of McLardy's may be due to one dial forebrain bundle, some of which by- of technique. A temporoseptal intercon- passed the septal area, entered the fornix, nection has been confinned by Stoll, and went to the hippocampus. Finally, in Ajmone-Marsan, and Jasper ('51 ) using a study of the connections of the cingulum electrophysiological techniques. in the albino rat, White ('59) noted that Electroneurophysiological methods are degeneration appeared in the hippocampus perhaps somewhat more convincing than only when damage to the septal area, for- are the morphological studies in demon- nix, or dorsal psalterium was evident in strating a hippocampal afferent fiber sys- his specimens. Cus and Stern ('61) were tem in the fornix. In a very careful and equally convinced of a relationship be- thorough analysis of the electrical activity 204 CHARLES L. VOTAW AND EDWARD W. LAUER of the hippocampus as a result of stimula- and Hamlyn ('57) found that stimulation tion of the fornix, reticular system, and of the septum in rabbits caused a response peripheral nerves by Green and co-workers in the presubiculum which was then fol- (Green and Morin, '53; Green and Arduini, lowed by a response in the hippocam- '54; Green and Adey, '56; Von Euler, pus. They did not find any direct septo- Green and Ricci, '58), it was shown that hippocampal responses. In this particular there is a response in the hippocampus to species, they found that lesions of the stimulation of the fornix, reticular system, septum produce degeneration (Nauta- and peripheral nerves. They further dem- Gygax technique) in the presubiculum, onstrate that this activity is orthodromi- but they did not report any in the hippo- cally produced. Stimulation of the reticu- campus. Similar morphological results lar system produced a peculiar type of were found later by these authors in the arousal activity in the hippocampus. This same species ('59). Certainly the pre- arousal activity was eliminated by septal ponderance of the data in the literature lesions which also interrupted the fornix suggests that there is conduction from the system. fornix to Ammon's horn and to the entorhi- From a similar study, Fujika and Naka- nal cortex. Blackstad ('56, '58) has re- mura ('61) have concluded that there are ported commissural fibers in the fornix both excitatory and inhibitory afferent that project to the presubiculum and other systems in the fornix going to the hippo- parts of the hippocampal complex, and campus. They do not demonstrate to their perhaps some of these fibers were involved own satisfaction that the excitatory in- in Cragg and Hamlyn's work, although fluence is a strictly orthodromic one. The from the description of the technique in possibility exists that antidromic stimula- the latter investigation, it would not seem tion of the pyramidal cell axons ortho- SO. dromically stimulates the collaterals of Rose and Woolsey ('49) have distin- this axon (Schaeffer collaterals) which, guished the Ammon's horn and entorhinal in turn, can orthodromically cause a post- area as the only cortical fields which do synaptic response. However, Green and not receive thalamic fibers, a fact that his co-workers (quoted above) and re- would set apart this portion of the telen- cently Feldman ('62) have shown that cephalon. That Ammon's horn is not solely stimulation of more distant structures related to olfaction has been well shown (hypothalamus, reticular system, periph- (see Brodal, '47; Votaw, '59, for further eral nerves) results in excitatory influ- references). It has been suggested that ences in the hippocampus which may perhaps the hippocampus has some influ- reach it by way of the fornix. ence on other cortical fields. The ascend- Andersen and co-workers { Andersen, ing influences (hippocampocortical) may Bruland and Kaada, '61a, '61b) and Dun- well rely on the information which the lop ('58) have carried out electrophysio- hippocampus receives through the afferent logic studies on the reactions of the hippo- fibers in the fornix. As the thalamus does campal pyramids secondary to stimulation not appear to project to the hippocampus, of the septum and/or fornix. They find and, from the data in the literature quoted excitatory responses that must be carried here, the hypothalamus and reticular areas along hippocampal afferent fibers. (as well as others) do, the hippocampal Adey, Sunderland and Dunlop ('57) system may well be set apart physiologi- stimulated the fornix and found that, in cally as a separate system with a function the marsupid, they could obtain responses that is poorly understood. in the hippocampus, followed by responses Finally, the difference observed in this in the entorhinal area. These responses study between lesions of the fornix and were easily elicited. In contrast, it was lesions of the layers of Ammon's horn much more difficult to stimulate the ento- where the dentate gyms and fimbria were rhinal area and record a response in the not involved should be noted. Section of fornix (the more classical direction of the fornix produced Marchi degeneration conduction). Frequently, this latter re- in all of the classical terminations of the sponse could not be obtained at all. Cragg fornix system (see Nauta, '56; Brodal, '47; HIPPOCAMPAL AFFERENT FIBERS 205

Valenstein and Nauta, ’59 and Ram6n y cephalic structures and the brain stem. EEG Cajal, ’11). Lesions of Ammon’s horn Clin. Neurophysiol., 9: 309-324. Andersen, P., H. Bruland and B. R. Kaada 1961a which did not include dentate gyrus and Activation of the dentate area by septal stimu- fimbria resulted in Marchi degeneration lation. Acta Physiol. Scand., 51: 17-28. products in the anterior parts of the for- 1961b Activation of the field CA-1 of nix system (septal area, diagonal band of the hippocampus by septal stimulation. Ibid., Broca, commissural fibers) but little evi- 51: 2940. dence could be found for degeneration in Blackstad, T. W. 1956 Commissural connections of the hippocampal region in the rat, the postcommissural parts of the fornix with special reference to their mode of termi- projection system. The data are not com- nation. J. Comp. Neur., 105: 417-538. plete enough to reach any definite con- 1958 On the termination of some af- clusions, but it is suggested that perhaps ferents to the hippocampus and fascia dentata. there are marked differences in the pattern Acta Anat., 35: 202-214. Brodal, A. 1947 The hippocampus and the of termination of the projection fibers of sense of smell. Brain, 70: 179-224. the dentate gyrus as compared to those of Cragg, B. G. 1961 Olfactory and other afferent Ammon’s horn, and that these may be connections of the hippocampus in the rabbit, important from a physiological standpoint. rat, and cat. Exp. Neurol., 3: 588-600. Cragg, B. G., and L. H. Hamlyn 1957 Some In as much as the dentate gyrus was not commissural and septal connexions of the involved in the lesions discussed here, it hippocampus in the rabbit. A combined histo- might be assumed that fibers arising in logical and electrical study. J. Physiol., 135: this gyrus would be found in lateral parts 460-485. of the bodies of the fornix. However, com- 1959 Histologic connections and electrical and autonomic responses evoked by stim- plete section of the fornix bilaterally which ulation of the dorsal fornix in the rabbit. Exp. would have cut such fibers (as the two Neurol., 1: 187-213. cases reported in the second group of ani- Crosby, E. C. 1917 The forebrain of Alligator mals in the results) still did not result in mississippiensis. J. Comp. Neur., 27: 325-402. degeneration of the lateral portion of the Cus, M., and P. Stern 1961 Beitrag zur funk- tionellen Beziehung des Hippocampus und fornix. We must conclude that most of Septum pellucidum bei der Ratte. Z. ges. exp. the lateral portion of the fornix is reserved Med., 134: 443-445. for hippocampal afferent fibers. Daitz, H. M., and T. P. S. Powell 1954 Studies of the connexions of the fornix system. J. CONCLUSIONS Neurol. Neurosurg. Psychiat., 17: 75-82. 1. A study of the components of the Dunlop, C. W. 1958 Viscero-sensory and SO- mato-sensory representation in the rhinen- fornix of the monkey has been accom- cephalon. EEG Clin. Neurophysiol., 10: 297- plished by placing selected lesions in the 304. septal area, fornix, and Ammon’s horn and Feldman, S. 1962 Neurophysiological mecha- investigating the resulting degeneration as nisms modifying afferent hypothalamo-hippo- demonstrated by the Marchi technique. campal conduction. Exp. Neurol., 5: 269-291. Fox, C. A. 1943 The stria terminalis, longi- 2. It was found that there is a hippo- tudinal association bundle and precommissural campal afferent fiber system in the fornix, fornix fibers in the cat, J. Comp. Neur., 79: located on the lateral side of the body of 277-291. the fornix, which arises from anterior Fujika, Y., and Y. Nakamura 1961 Effect of fornical stimulation upon the CA-1 and CA-2 areas (septum suggested) and terminates apical dendrite of rabbit’s hippocampus. Jap. somewhere in the hippocampal formation. J. Physiol., 11: 357368. 3. The well-known hippocampal effer- Gerebtzoff, M. A. 1939 Sur quelques voies ent system arising from Ammon’s horn is $association de l’6corce ckr6bral. J. Belge Neurol., 39: 205-221. located on the medial side of the body of - 1941 Note anatomo-expbrimentale sur the fornix and terminates in the anterior le fornix, la come d’Ammon et leur relations projection field of the fornix system. avec diverse structure enckphaliques notam- 4. The relation of these findings to pre- ment l‘bpiphyse. Ibid., 41: 199-206. Green, J. D., and W. R. Adey 1956 Electro- vious ones in the literature is discussed. physiological studies of hippocampal connections and excitability. EEG Clin. Neurophysiol., LITERATURE CITED 8: 245-262. Adey, W. R., S. Sunderland and C. W. Dunlop Green, J. D., and A. Arduini 1954 Hippocam- 1957 The entorhinal area; electrophysiologi- pal electrical activity in arousal. J.- Neuro- cal studies of its interrelations with rhinen- physiol., 17: 533-554. 206 CHARLES L. VOTAW AND EDWARD W. LAUER

Green, J. D., and F. Morin 1953 Hypothalamic Sprague, J. M., and M. M. Meyer 1950 An electrical activity and hypothalamo-cortical re- experimental study of the fornix in the rabbit. lationships. Amer. J. Physiol., 172: 175-186. J. Anat., 84: 354-368. Guillery, R. W. 1957 Degeneration in the hy- Stoll, J., C. Ajmone-Marsan and H. H. Jasper pothaIamic connexions of the albino rat. J. 1951 Electrophysiological studies of subcor- Anat., 91: 91-116. tical connections of anterior temporal region in Lauer, E. W. 1945 The nuclear pattern and cat. J. Neurophysiol., 14: 305416. fiber connections of certain basal telencephalic Swank, R. L., and H. A. Davenport 1935 centers in the macaque. J. Comp. Neur., 82: Chlorate-osmic-formalin method for staining 215-254. degenerating myelin. Stain Technol., 10: 87-90. Loo, Y. T. 1931 The forebrain of the opossum Valenstein, E. S., and W. J. H. Nauta 1959 A Didelphis virginiana. Part 11. Histology. Ibid., comparison of the distribution of the fornix 52: 1-148. system in the rat, guinea pig, cat, and monkey. McLardy, T. 1955a Observations on the fornix J. Comp. Neur., 113: 337-363. of the monkey. I. Cell studies. Ibid., 103: Von Euler, C., J. D. Green and G. Ricci 1958 305-325. The role of hippocampal dendrites in evoked 1955b Observations on the fornix of responses and after-discharge. Acta Physiol. the monkey. 11. Fiber studies. Ibid., 103: 327- Scand., 87-111. 343. 42: Mettler, F. A. 1943 Extensive unilateral cere- Votaw, C. L. 1959 Certain functional and bral removals in the primate: physiologic ef- anatomical relations of the cornu Ammonis of fects and resultant degeneration. Ibid., 79: the macaque monkey. I. Functional relations. 185-245. J. Comp. Neur., 112: 353382, Morin, F. 1950 An experimental study of hy- - 1960a Study of septa1 stimulation and pothalamic connections in the guinea pig. ablation in the macaque monkey. Neurol., 10: Ibid., 92: 193-213. 202-209. Nauta, W. J. H. 1956 An experimental study 1960b Certain functional and anatomi- of the fornix in the rat. Ibid., 104: 247-272. cal relations of the cornu Ammonis of the Ramh y Cajal, S. 1911 Histologie du sys- macaque monkey. 11. Anatomical relations. tltme nerveux de l'homme et des vert6bri.s. A. J. Comp. Neur., 114: 283-293. Maloine, Paris. White, L. E. 1959 Ipsilateral afferents to the Rose, J. E., and C. N. Woolsey 1943 A study of hippocampal formation in the albino rat. Ibid., thalamo-cortical relations in the rabbit. Bull. 113: 142. Johns Hopkins Hosp., 73: 65-128. Young, M. W. 1936 The nuclear pattern and 1949 Organization of the mammalian fiber connections of the noncortical centers thalamus and its relationships to the cerebral of the telencephalon of the rabbit (Lepus cortex. EEG Clin. Neurophysiol., 1: 391404. cuniculus). Ibid., 65: 295-401.