Citation
The social behavior and aggressiveness of the hamster following the application of chemicals to the septal region of the forebrain

Material Information

Title:
The social behavior and aggressiveness of the hamster following the application of chemicals to the septal region of the forebrain
Creator:
Sodetz, Frank Jack, 1941-
Publication Date:
Language:
English
Physical Description:
vii, 140 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Atropine ( jstor )
Fornix ( jstor )
Hamsters ( jstor )
Hippocampus ( jstor )
Human aggression ( jstor )
Lesions ( jstor )
Limbic system ( jstor )
Norepinephrine ( jstor )
Rats ( jstor )
Septum of brain ( jstor )
Animal behavior ( lcsh )
Brain -- Localization of functions ( lcsh )
Dissertations, Academic -- Psychology -- UF ( lcsh )
Hamsters ( lcsh )
Psychology thesis Ph. D ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida, 1967.
Bibliography:
Bibliography: leaves 121-135.
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
This item is presumed in the public domain according to the terms of the Retrospective Dissertation Scanning (RDS) policy, which may be viewed at http://ufdc.ufl.edu/AA00007596/00001. The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator(ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
022374526 ( ALEPH )
13697283 ( OCLC )

Downloads

This item has the following downloads:


Full Text












THE SOCIAL BEHAVIOR AND
AGGRESSIVENESS OF THE HAMSTER FOLLOWING THE APPLICATION OF
CHEMICALS TO THE SEPTAL REGION OF THE FOREBRAIN




By
FRANK JACK SODETZ, JR.


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY










UNIVERSITY OF FLORIDA
August, 1967










ACKNOWLEDGMENTS


The author wishes to express his appreciation to his supervisory committee chairman, Dr. B. N. Bunnell, for his support and encouragement throughout the past four years. A debt of gratitude is owed also to the members of the committee, Drs. D. C. Goodman, J. A. Horel, F. A. King, P. Satz, and W. B. Webb, whose guidance and letters made the completion of this dissertation possible.

The untiring efforts of Mr. J. A.Jones and Mr. C. P. Vega in assisting with the collection and processing of the data are most gratefully acknowledged. The services of the University of Florida Computing Center are also acknowledged. The manuscript could not have been completed without the assistance of Mrs. M. A. Harrington and Miss T. R. Karell who looked after the illustrations and photography and Mrs. E. Marable and Miss B. Watson who typed the manuscript.


- ii -












TABLE OF CONTENTS


page
ACKNOWLEDGMENTS................................................ ii

LIST OF TABLES................................................. v

LIST OF FIGURES................................................ vi

INTRODUCTION................................................... 1

General Background
The problem
Anatomical considerations

The Limbic System: Structure
The limbic system in submammalian forms
Amphioxus
Cyclostomes
Chondrichthyes
Teleosts
Amphibians
Reptiles Summary
The mammalian limbic system
Anatomical data
The septal region
Electrical stimulation studies
Chemical mapping of the limbic system
Definition of the mammalian limbic system

The Limbic System: Function
The septal region
Emotion
Avoidance learning
Other learning paradigms
Feeding, drinking, and sleep
Related structures
Emotion Learning
Some relevant pharmacological studies
The limbic system and species-specific behavior

METHOD......................................................... 48

Subjects
Apparatus
Response inventory
Surgery


- iii -










Procedure Histology
Subjects dropped from the study
Saline controls
Norepinephrine group
Intraventricular norepinephrine group
Phenoxybenzamine group
Intraventricular phenoxybenzamine group
Phenoxybenzamine - fornix resection group
Phenoxybenzamine - amygdalectomy group
Carbachol group
Atropine group
Summary of groups
Statistical analysis

RESULTS......................................................... 60

DISCUSSION...................................................... 100

SUMMARY......................................................... 118

REFERENCES...................................................... 121

APPENDIX A...................................................... 137

APPENDIX B...................................................... 139

BIOGRAPHICAL SKETCH. ............................................ 140


- iv -












LIST OF TABLES


Table Page

1. A brief description of the responses which can be seen
in the interaction of two adult male hamsters............ 50

2. A summary of the groups used in the study................ 57


-v -











LIST OF FIGURES


Figure Page

1. The differences in the distribution of proportions of
the saline group (N=18) and that of their normal opponents (N=18) expressed as deflections from the baseline
which represents the distribution of the normal Ss.
The total number of responses observed was 7,214
(Saline = 3,739; Opponents = 3,575)...................... 62

2. The differences between the distributions of proportions
of both the carbachol (N=4) and atropine (N=4) groups and the saline group (N=18) expressed as deflections from the
baseline which represents the distribution of the saline group. The total number of responses observed was 5,304
(Carbachol = 751; Atropine = 814; Saline = 3,739)........ 65

3. The differences between the distributions of proportions
of both the norepinephrine (N=6) and intraventricular
norepinephrine (N=3) groups and that of the saline group
(N=18) which is represented by the baseline. The total
number of responses observed was 5,431 (Norepinephrine =
1,167; Intraventricular norepinephrine = 525; Saline =
3,739)................................................... 68

4. The differences between the distributions of proportions
of both the phenoxybenzamine (N=9) and intraventricular
phenoxybenzamine (N=4) groups and that of the saline group
(N=18) which is represented by the baseline. The total
number of responses observed was 5,474 (Phenoxybenzamine =
1,735; Intraventricular phenoxybenzamine = 688; Saline =
3,739)................................................... 70

5. The differences between the distributions of proportions
of the norepinephrine (N=6) and phenoxybenzamine (N=9)
groups and that of the saline group (N=18) which is represented by the baseline. The total number of responses
observed was 6,641 (Norepinephrine = 1,167; Phenoxybenzamine - 1,735; Saline = 3,739)............................ 73

6. The differences between the distributions of proportions
of the fornix resection group (N=4) obtained with and
without the application of phenoxybenzamine to the septal region and that of the phenoxybenzamine group (N=9) which is represented by the baseline. The total number of responses observed was 2,673 (Fornix resection with septal phenoxybenzamine = 699; Fornix resection alone =
239; Phenoxybenzamine = 1,735). The fornix resection
alone condition was run for only three trials............ 76


- vi -









7. The differences between the distribution of the amygdalectomy-phenoxybenzamine group (N=4) and that of the
phenoxybenzamine group (N=9) represented by the baseline.
the total number of responses observed was 2,058 (Amygdalectomy-phenoxybenzamine group = 353; Phenoxybenzamine group = 1,735)...................................... 79

8. The differences between the distributions of proportions of both the carbachol (N=4) and the norepinephrine (N=6)
groups and that of the saline group (N=18) represented by the baseline. The total number of responses observed was
5,657.(Carbachol group = 751; Norepinephrine group = 1,167;
Saline group = 3,739) .................................... 81

9. The differences between the distributions of proportions of both the atropine (N=4) and phenoxybenzamine (N=9)
groups and that of the saline group (N=18) represented
by the baseline. The total number of responses observed
was 6,288 (Atropine group = 814; Phenoxybenzamine group =
1,735; Saline group = 3,739)............................. 83

10. Sections from representative Ss showing ventricular and
rostral septal placements................................ 86

11. Representative sections from Ss with rostral and typical
septal placements........................................ 88

12. Representative sections from Ss with more extreme variations in septal placements............................... 91

13. Representative sections from two Ss with placements
which were unlike any others in the study and sections
from two Ss from the lesion groups illustrating their
lesions.................................................. 93

14. A summary of the effects obtained in the study. All
groups are compared with the saline group. Differences of less than 34% were ignored. In considering the data
presented in this figure, it should be noted that the mean number of responses observed for each S in each
group was as follows: Saline, 199; Carb., 188; NE Sept.,
194; NE Vent., 175; Phen. Sept., 193; Phen. Vent., 172;
Fornix, 175; Amyg., 78................................... 96


- vii -















INTRODUCTION


General Background

The problem

The present study was designed to further evaluate the role of the septal region of the forebrain in the aggressiveness and social behavior of the hamster. The need for further evaluation arose out of an attempt to cope with a problem in one of the first studies dealing directly with the role of the septal region in these speciesspecific behaviors. The study (Bunnell, Bemporad, and Flesher, 1966) demonstrated that septal ablation increased the social rank of hooded rats. However, social rank was defined on the basis of the effectiveness of the animal in a competitive situation and, while this is not uncommon in such studies, the artificiality of the testing situation may have favored the septal animals. Pelligrino (1967) has since verified this work using a somewhat different competitive test. Bunnell and Smith (1966) studied the social behavior of cotton rats following septal ablation in a more natural setting and found both increased aggressive and escape behavior in these animals. The social behavior of the cotton rat was so disrupted following septal ablation that a number of the animals were killed by their cagemates. At this point, there seemed to be a clear need to study the effect of septal ablation on species-specific behavior using a technique which would permit explicit statements about resulting behavioral changes and yet would


- 1 -









not have the undesirable attributes of an artificial testing situation. The selection of an animal was critical in this regard. The hamster seemed ideally suited for this purpose in that it engages in social interaction without special training or inducement from the experimenter. Furthermore, the hamster uses a discrete number of readily identifiable behaviors in this interaction. A number of studies (Sodetz, 1965; Bunnell, Sodetz, and Shallaway, 1965; Sodetz and Bunnell, 1967a; Sodetz and Bunnell, 1967b) have now shown the septal region of the hamster to be involved in the regulation of such species-specific behaviors as hoarding, nest building, and aggressive interaction. In avoiding the artificial testing situation, a problem was created. These studies sampled a broad range of behaviors. Some of these behaviors diminished in their frequency while others increased in frequency. Previous social experience appeared to determine whether septal ablation would make the hamster more aggressive or more submissive. It also appeared that the behavior of the opponent of the septal animal was of diminished importance in determining the behavior of the septal.

Bunnell, et al. (1965) have suggested that a uniprocess model may not be able to account for the changes in species-specific behavior resulting from septal ablation. Such a model would also have to account for the appearance of "septal rage" in some species (Brady and Nauta, 1953; King, 1959; Brown and Slotnick, 1966) and not in others (Sodetz, Matalka, and Bunnell, 1967). Changes in water intake which follow ablation (Harvey and Hunt, 1965) and stimulation (Fisher and Coury, 1962; Grossman, 1964b) of the septal region would have to be considered in such a model, as would learning deficits (King, 1958; McCleary, 1961; Ellen, Wilson, and Powell, 1964).


- 2 -






-3-


The variety of the behavioral changes seen following ablation and stimulation of the septal region suggested that it would be profitable to investigate the possibility that more than one deficit might be responsible for the results obtained in the studies of species-specific behavior. However, although the septal region consists of a number of anatomically distinct nuclei (Andy and Stephan, 1961) which differ in their connections to other regions of the brain, little evidence has been found for a functional differentiation based upon anatomically distinct regions (Harrison and Lyon, 1957). Burkett and Bunnell (1966) did find evidence that the medial septal region may be responsible for the DRL deficit seen in septal rats (Ellen, Wilson, and Powell, 1964).

While a differentiation in function based upon anatomical data has been difficult to confirm, the work of Hernandez Peon, Ibarra, Morgane, and Timo-laria (1963) which demonstrated cholinergic sites for sleep and rage throughout the septal region and Grossman's (1964a) success in eliciting both eating and drinking from the same anatomical locus in the amygdala with two different neurotransmitter substances, suggested that damage to chemically distinct subsystems might be responsible for the deficits seen following ablation of the septal region.

More specifically, the present study was designed to test the

possibility that a multiplicity of functions might underlie the changes in species-specific behavior which result from septal ablation. The technique of selective chemical stimulation and blocking of chemically distinct systems through indwelling cannulae seemed to offer the best hope in this regard.

Anatomical considerations

The septal region has traditionally been considered to be part of






-4-


the limbic system. One would expect its functions to be intimately related to those of the other structures in this system. However, one has difficulty finding a consensus as to what these structures are and, therefore, inferences about septal function based upon the function of related structures are difficult to make.

Discussions of the limbic system are numerous in the recent literature. Each paper presents an anatomical definition which differs in some aspect, usually according to the function being emphasized by the author, from definitions outlined in similar papers. For Broca (1878) the "grande lobe limbic" was a prosencephalic system common to all mammals. Papez (1937) found it to be an essentially "closed" system lying within the prosencephalon but he tended to emphasize its diencephalic components and proposed that such a system might offer the reverberating circuitry thought necessary to the maintenance of emotional apprehension. The work of Bard (1939) and Bard and Mountcastle (1948) provided data to partially support this view of the limbic system. At the present time, the limbic system of Pribram and Kruger (1954) is among the best known and most widely accepted of a number of limbic system definitions. This acceptance, however, does not extend throughout the community of those involved in limbic system research. Within this group it is recognized that a wealth of close relationships exists between this limbic system and other parts of the brain. The decision to include some or all of these other structures and connections has been based on behavioral data suggesting very close functional relations between structures included in the Pribram and Kruger definition and neocortical, reticular, hypothalamic and extra-pyramidal structures. In opposition to this extension of the limbic system has been the easily






-5-


justifiable argument that one can readily connect any structure within the central nervous system to any other structure by a route which would include very few synaptic relays., By this reasoning, the limbic system could readily be extended to include most of the nervous system. While most arguments both for and against the inclusiveness and exclusiveness of the limbic system have some merit, it should be noted that the limbic system referred to above does not derive its integrity from a common histogenetic origiNnor has a consensus been found regarding some unitary functional designation which can be applied to this system. It would appear that adherence to the concept of a limbic system without neocortical or mesencephalic components is propagated as much for historical reasons as any others. The limbic system has been shown to have numerous complex connections with the mesencephalon (e.g., Nauta and Kuypers, 1958) and is only one relay removed from the efferent nuclei of the medulla. In the rostral cerebrum, the frontogranular cortex has been shown to have direct connections with the limbic system (e.g., Nauta, 1964). Brady (1958a) in attempting to anatomically delimit the limbic system expresses his concern, and that of many others, for the adequacy of any such attempt. What seems to be required is some anatomical and functional redefinition of the limbic system if one is to adequately characterize the function of one of its prominent parts. What follows below is an attempt to review some of the data from which such a definition may someday emerge.


The Limbic System: Structure

The limbic system in submammalian forms

Amphioxus. Elements of a very primitive limbic system can be






-6-


traced back through phylogeny below vertebrates. The chordate Amphioxus, which looks much like a fat worm and lacks eyes, has a very primitive forebrain. A telencephalon and diencephalon cannot be discriminated in this species. Amphioxus does have a single, unpaired olfactory nerve which enters the forebrain and distributes to the homologue of the septal region. The course of fibers originating in the septal region is not known.1

Cyclostomes. In primitive vertebrates an olfactory related limbic apparatus is clearly present. In cyclostomes, the telencephalic portion of this Tstem occupies the mediobasal portion of the hemispheres and consists of the septum. The septum is a major component of the archistriatum from which some of the other components of the limbic system will arise in higher forms. The dorsal convexity of the cerebrum is formed by the archipallium, from which additional components of the limbic system later differentiate. The dorso-medial archipallium is generally considered to be primordial hippocampus. The preoptic nucleus also appears to be present in this form. The epithalamus, consisting mainly of the habenular nuclei, appears to receive much of the output of these centers via its primary afferent pathway, the stria medullaris, and other olfacto-habenular bundles. The olfactory areas also project to most of the hypothalamus. The epithalamus discharges via the fasciculus retroflexus to the nucleus interpeduncularis. From the nucleus interpeduncularis, which is of considerable size in lower forms, fibers distribute through the medulla



iMuch of the phyletic data on the limbic system was obtained from: Ariens Kappers, C.V., Huber, G.C. and Crosby, E.C. The Comparative Anatomy of the Nervous System of Vertebrates including Man. Hafner Publishing Co., 1960, III Volumes.










oblongata to efferent centers. A hypothalamic bulbar fiber tract courses along the extent of the interpeduncular nuclei and also distributes in the medulla.

Notable in the discussion of the primitive limbic structures of these forms is the absence of the reciprocal interconnections seen in higher vertebrates. In cyclostomes limbic pathways are for the most part directed rostrocaudally. One is also impressed by the sparse limbic-hypothalamic interrelationship even though both serve the same medullary efferent centers. This relative simplicity of limbic organization is suggestive of a rigid translation of olfactory input into effector activity.

Chondrichthyes. The brain of the shark shows an increased complexity in olfactory connections and it is at this level in phylogeny that it appears possible to suggest that olfactory-related telencephalic structures may take on other functions. In chondrichthyes there are found a number of commissural limbic connections. In addition, the hypothalamus in these animals sends fibers to the habenulae and receives fibers from both the septal and primordial hippocampal regions. The existence of such connections suggests that considerable interaction may take place between limbic system and hypothalamus. This may be in contrast to the rather direct olfacto-effector relations seen in cyclostomes. Another correlation appears in the telencephalon of chondrichthyes. The dorsal thalamus sends fibers, presumably somato-sensory in origin, to the very small non-olfactory portions of the archistriatum which in turn have connections to the lateral olfactory area or paleopallium. Both of these areas will eventually contribute to the mammalian amygdala, the basolateral components being paleopallial in







-8-


origin and the corticomedial arising from archistriatum. The globus pallidus and pyriform areas also have their origin in these regions. More importantly, what is seen in these forms is an opportunity for the interaction of olfactory and other senses at the telencephalic level.

Teleosts. In more modern fishes the medial olfactory region, much of which will be the septal and preoptic regions in mammals, consists of several identifiable nuclei. These include the nucleus precommissuralis superior and inferior and the nucleus olfactorius anterior pars precommissuralis. Inferior to the caudal border of the decussation of the anterior commissure lies the preoptic nucleus. A large area of the telencephalon may be designated primordial hippocampus. The olfactosomatic paleopallial region noted in the chondrichthyes is larger in the teleosts. Perhaps the most notable development in relationship to the limbic system is a separation of the hypothalamus from primary olfactory fibers and enrichment of the connections between precommissural septal structures and the hypothalamus. By way of the medial forebrain bundle, both afferent and efferent interaction are possible between hypothalamic and limbic regions, including the primordial hippocampus. Another precommissural septal-hypothalamic connection runs ventrocaudally from the septal region and probably represents the precommissural septal contribution to the columns of the fornix in higher forms. The epithalamic-limbic relationship may be of diminished significance in these forms for the habenulae are reduced in size and far fewer fibers distribute to this center than to the hypothalamus.

Amphibians. In the amphibians, all of the structures commonly

considered to be part of the limbic system are seen, with the exception of the cingulate cortex which is neopallial in origin. It is in these






-9-


forms that the archistriatal cortico-medial amygdala first appears. A primordial pyriform lobe can also be identified as lying adjacent to the amydaloid region. The septal region may be divided into medial and lateral septal nuclei. In amphibians, the hippocampus remains primordial. The limbic system of amphibians evidences more definite differentiation of nuclei and a greater interrelatedness with nonolfactory pallial and striatal structures. The interrelationships both within the system and with other systems are further characterized by the reciprocity which is seen in higher forms.

Reptiles. The limbic system of reptilian forms is more differentiated in its nuclear structure than is that of the amphibian. Most of the limbic structures seen in mammals are present and well differentiated. This differentiation extends to the hippocampus, which appears to have the character of a true cortical structure. The reptilian striatal regions are also more highly differentiated. A small region of general cortex can be found in the pallium of the reptile and is the forerunner of the general neocortex of mammals.

Summary. The avian forebrain develops in a line independent of mammalian forms and it is sufficient to note that the limbic system of the bird is similar, though more complex, to that of reptilians.

Throughout the vertebrate phyla the limbic system has undergone development from a rather direct olfacto-effector system, with interaction from other systems arising primarily in the regions nearest effector sites, to an intermediate stage at which interaction with other areas arose in the more rostral regions of the brain. This intermediate stage foreshadowed the development of more numerous ascending influences from other systems to the telencephalic limbic






- 10 -


components. In mammals the limbic system has emerged with massive intrinsic and extrinsic connections and apparently little solitary control over any effector system. While it is true that increasing complexity and interrelatedness is a general phenonemon applicable to the entire nervous system in its ascent through phylogeny, the limbic system differs from other systems which are sensory in origin. While other sensory systems increase their complexity through differentiation and the proliferation of connections with other systems, they do not lose their identityas sensory systems. In the visual system, for example, the occipital neocortex is easily identifiable as a component in a system related to the processing of visual stimuli. The visual neocortex need not be traced back to its origin in phylogeny to locate some ancient link to the optic tectum or thalamus to support its inclusion in the visual system. However, this is not the case with the limbic system, for in lower forms it is clearly sensory-related. This relationship gradually dissipates as one ascends through phylogeny. Rather than maintaining a close association with its olfactory origin, the bulk of the limbic system retreats from its sensory status. The development of this system does not parallel that of other sensory systems in the enhancement of the capacity to elaborate stimuli but reflects the increasing importance, through phylogeny, of some other function or functions. These functions are most certainly related to both sensory and motor systems, but, in a strict sense, are probably neither. An appealing hypothesis presents itself in the notion that the limbic system, which somehow must deal with both sensory and motor systems, has undergone its development in response to the need to deal with a sensorium capable of processing increasingly complex sensory







- 11 -


data and an effector system capable of an enormous versatility of response. In the very lowest vertebrates, the cyclostomes, the source of a stimulus, almost certainly olfactory, is either affixed to and slowly devoured or ignored or avoided. In mammals countless choices exist for countless combinations of stimulus inputs. A much more complex limbic system may be necessary to cope with this wealth of data. The mammalian limbic system

Anatomical data. No attempt will be made in this section to consider in great detail the connections of various structures to be included in the limbic system. Each structure could, in itself, be the basis for a number of papers. Details of the anatomy of the limbic system and related structures can be found in papers by Andy and Stephan (1961), Cowan, Raisman, and Powell (1964; 1965), Gloor (1960), Green (1964), Nauta (1956; 1958; 1961; 1964), Nauta and Kuypers (1956), Papez (1937;1958), Powell (1963; 1964), Pribram (1960), Pribram and Kruger (1954), Valenstein and Nauta (1959), Raisman (1966), Raisman, Cowan, and Powell (1966), and Powell, Cowan, and Raisman (1965).

Perhaps the most profitable approach would be to consider first those structures commonly considered to be part of the limbic system. There is one exception to this, that is, the first system of Pribram and Kruger (1954). The structures included in this system all receive primary olfactory connections. There would appear to be little reason for continuing to consider these areas as part of the limbic system.

The septal region. The septal region will be considered first since it is the primary concern of this paper. The septal region itself can be divided into four nuclear groups (Andy and Stephan, 1961);






- 12 -


however, it is more common to discuss it in terms of a medial and lateral septal region. The lateral and medial septal nuclei differ with respect to their afferents andefferents. The medial septal nuclei send fibers to the hippocampus and receive fibers from the hypothalamus. The lateral septal nuclei receive fibers from the hippocampus and send fibers to the lateral hypothalamus. The septal region projects to, and receives fibers from,the dorsomedial thalamic nucleus (Valenstein and Nauta, 1959) which in turn supplies fibers to the frontogranular cortex (Akert, 1964; Nauta, 1964). Other thalamic projections of the septal area include those to the intralaminar nuclei and the anteroventral thalamic group. These connections are established via the mammillary bodies, although all fibers do not synapse there. Although it is not clear whether they arise in the septal region, the hippocampus, or both, fibers are found which project to the nucleus lateralis dorsalis in primates (Valenstein and Nauta, 1959). Such fibers are barely detectable in lower forms and suggest a relationship to parietal association cortex. Nauta (1953) does report, however, that the dorsolateral nucleus receives a massive projection from the caudal cingulate gyrus in species below primates. An epithalamic projection from the medial septal region to the habenula is found in all vertebrates. From the habenula fibers project to the nucleus interpeduncularis via the fasciculus retroflexus. From this nucleus fibers arise which project to the visceral efferent nuclei of the medulla.

Nauta (1956; 1958) provides an excellent discussion of septalhypothalamic relations via the fornix, while Andy and Stephan (1961)






- 13 -


emphasize the medial forebrain bundle. The connections of the septal region to the mammillary bodies have already been noted. The septal region sends fibers via the diagonal band to the lateral preoptic area. From here, fibers enter the diencephalon and distribute throughout the extent of the lateral hypothalamic area. While septal connections to the lateral hypothalamus are the most prominent, fibers from the columns of the fornix and fibers coursing through the medial preoptic region supply connections to more medial hypothalamic areas. These connections vary appreciably from species to species. The amygdala, which is directly related to the septal region via both the diagonal band and the stria terminalis also contributesto the medial forebrain bundle following relay in the lateral preoptic region. The amygdala, however, also has direct connections to the lateral hypothalamus via ventrofugal fibers described by Nauta (1961). The hippocampus, which receives afferents from the medial septal region and projects efferents to the lateral septal region, also contribute fibers to these hypothalamic areas which are served by the septal region. The medial forebrain bundle also carries fibers from the frontogranular cortex. Fibers from all of the above sources distribute in the mesencephalic tegmentum (Nauta, 1964; Nauta and Kuypers, 1958). These fibers generally distribute to the dorsomedial tegmentum at the lateral border of the central gray.

The septal region is connected to the cingulate gyrus via the anteroventral thalamic nucleus, while the cingulate gyrus projects to the supracommissural septal region via fibers which perforate the corpus callosum and enter the fornix dorsalis (Cragg and Hamlyn, 1959; White, 1965).






- 14 -


Other notable connections are those of the hippocampus to entorhinal cortex and from there to the mesencephalic reticular system, and the frontogranular cortex to the periamydaloid cortex and presumably the amygdala.

The number of interrelationships among these structures would permit the assembly of a number of different systems and favor no one system over any other. The functional data to be considered below may provide some useful basis for grouping these structures into systems.

Electrical Stimulation Studies. Maps of limbic and related

structures have been compiled on the basis of similiarities in obtained results (Delgado, 1964; DeMolina and Hunsperger, 1962; Hess, 1957; MacLean, 1963; Olds, 1956; 1958). These maps have generally linked the septal region, amygdaloid complex, preoptic area, anteromedial hypothalamus, lateral hypothalamus, and mesencephalic tegmentum in a system generally related to both reward and, in a broad sense, affective display. Kaada (1951) and others (e.g., Feindel and Gloor, 1954) have shown that stimulation of limbic areas leads to activation of other limbic sites and cortical association areas. This activation differs from reticular activation in that it does not extend throughout the entire cortical mantle. Gloor (1955) has demonstrated that amygdaloid stimulation produces changes in the activity of the mediodorsal, centromedian, and intralaminar nuclei of the thalamus. This is also true of stimulation of the cingulate gyrus (see Pribram, 1961). Activity can also be noted in the subthalamus and mesencephalon (French, Hernandez Peon, and Livingston, 1955; Gloor, 1955). These stimulation studies correlate very well with anatomical data suggesting a close






- 15 -


relationship between what is usually considered to be the limbic system and the frontogranular cortex, hypothalamus, anterior and intralaminar thalamus, and mesencephalic reticular formation. In reviewing the stimulation data, Pribram (1961) has emphasized the lack of topographical specificity in the effects obtained in limbic stimulation studies. The lack of intralimbic topographical specificity aid the extensiveness of the stimulation effects, however, should not be taken to mean that the limbic system subserves some general arousal or suppressive function. While such a position has some theoretical appeal (Routtenberg, 1966) and electrical stumulation studies have induced sleep from the septal region (Rosvold and Delgado, 1956) and adjacent areas (Sterman and Clemente, 1962) and arousal from the intralaminar nuclei (Akert, 1961), the specificity of the effects of limbic stimulation militate against any functional interpretation based upon a general arousal or inhibitory concept. The work related to the hippocampal theta rhythm further supports this view. Hippocampal theta is relatively slow, synchronous activity seen clearly in the rabbit, but which is less obvious in the monkey (Green and Arduini, 1954). Petsche, Stumpf, and Gogolak (1962) have shown that a group of cells in the post-commissural septal region act as pacemakers in producing hippocampal theta. Following the destruction of these cells, it is not possible to produce theta either by chemical or electrical stimulation (Stumpf, 1965); however, stimulation of the reticular formation will still elicit normal low voltage fast activity in the hippocampus (Mayer and Stumpf, 1958). These data suggest that general arousal is still possible, even within the limbic system, following damage to its components.






- 16 -


Studies of the effects of electrical stimulation of the limbic system have provided support for the anatomical data, suggesting a close relationship between limbic structures. These same studies strongly suggest that the intralaminar thalamic nuclei and the region of the dorsomedial midbrain tegmentum adjoining the periaquaductal gray be included in a functional, if not anatomical, limbic system. Further support for this position can be found in neurochemical studies of the central nervous system.

Chemical mapping of the limbic system. The fluorohistochemical technique of Falck, Hellarp, Thilma, and Torp (1962) has made possible the easy identification of catecholamines in the central nervous system. Of these, norepinephrine, serotonin (5-Hydroxytryptamine), and dopamine are generally considered to be related to synaptic transmission. As yet, no such technique exists for the identification of acetylcholine, the other widely accepted transmitter substance. Acetycholine concentrations are not measured directly as are those of the catecholamines, but are generally inferred from variations in regional concentrations of specific acetylcholinesterases. The inherent dangers of inferring the level of acetylcholine from measures of the enzyme required to destroy it are well known; however, when acetylcholine has been measured direcly, there has been good correspondence between the direct measures and the estimates based upon specific acetylcholinesterases (Keolle, 1954). The appeal of the direct observation of catecholamines has produced a considerable number of studies related to catecholaminergic systems in the brain whereas studies dealing with cholinergic systems, as opposed to regional concentrations, are seen less often.






- 17 -


Techniques for the fluorometric differentiation of the catecholamines have only recently been developed (Glowinski and Baldessarini, 1966); therefore, one has difficulty in distinguishing between norepinephrine and serotonin concentrations in the limbic system. Just as these two substances have been difficult to differentiate fluorometrically, studies using various anticatecholaminergic substances are plagued by the fact that these chemicals lack specificity in their effects and can generally be considered to interfere with both serotonin and norepinephrine.

The distribution of catecholamines in the central nervous system and more specifically, the limbic system, is remarkably similar in species ranging from reptiles (Quay and Wilhoft, 1964) to man (Bertler, 1961). Bertler and Rosengren (1959) have found that, in the six mammalian forms they studied, dopamine concentrations were high in areas where the concentration of norepinephrine was low and norepinephrine concentrations were high in areas where the dopamine levels were low. This may suggest a functional dissociation between these two catecholamines,for serotonin and norepinephrine are often found in high concentration within the same regions of the brain.

A number of studies have demonstrated a limbic-hypothalamicmidbrain catecholaminergic system (Anden, Dahlstrom, Fuxe, Larsson, Olson, and Ungerstedt, 1966; Aghajanian and Bloom, 1967; Bodganski, Weissbach, and Udenfriend, 1957; Dahlstrom, Fuxe, Olson, and Ungerstedt, 1962; Harvey, Heller, and Moore, 1962; 1963; Heller and Moore, 1965; Heller, Seiden, Porcher, and Moore, 1966; Hillarp, Fuxe, and Dahlstrom, 1966; Moore and Heller, 1967; Moore, Wong, and Heller, 1965;






- 18 -


Paasonen, MacLean, and Giarman, 1957; Reivitch and Glowinski, 1966). While these are just a few of many such studies, they are representative in their agreement on the catecholamine content of the limbic system and related structures. They agree that a high concentration of catecholamines is to be found in the limbic midbrain region described by Nauta and Kuypers (1958). This system appears to be both ascending and descending in the medial forebrain bundle and lateral hypothalamus to the preoptic and septal regions. Beyond this, little is known of the origins and connections of this important limbic circuit for an as yet unexplainable phenomenon occurs with lesions of the medial forebrain bundle. Heller, et al. (e.g. 1966) and others in their group, have studied changes in regional concentration of catecholamines following a variety of subcortical lesions. They have demonstrated that transection of the medial forebrain bundle produces a rapid drop in catecholamine levels in the limbic septal and midbrain regions. However, coincident with this rapid drop, is a slow, relatively steady, decrease in norepinephrine levels throughout the forebrain (Moore and Heller, 1965). This effect is obtained in structures many synapses removed from the septal-medial forebrain bundle-midbrain circuit interdicted by the lesion. No explanation has been found for this transsynaptic depletion of norepinephrine. It is of interest to note that in the rat the depletion continues for approximately fourteen days. Septal ablation also produces a depletion in norepinephrine. The results of a study by Pirch and Norton (1965) are of special interest if considered with the results of the work reported above, even though they are not quite appropriate to this






- 19 -


section. Pirch and Norton (1965) treated septal rats, in which all evidence of hyperemotionality had disappeared, with Beta-Phenylisopropy1hydrazine which produces a temporary elevation in brain norepinephrine levels. They observed a reinstatement of septal hyperemotionality which was delayed for about five hours and which gradually diminished during the ensuing seven hours. They followed the change in brain norepinephrine levels during this period. Their results indicate a gradual increment in norepinephrine levels during the first five hours, following the administration of Beta-Phenylisopropylhydrazine followed by a falling phase which paralleled the dissipation of hyperemotionality. The observed septal rage was maximal at the initiation of the falling phase of the norepinephrine concentration and it had disappeared when the concentration reached the chronic low level characteristic of the septal rat. While these authors do not specifically relate their work to the observation of Moore and Heller (1967) of the fourteen-day falling phase of norepinephrine levels produced by lesions in the catecholamine component of the limbic system of the rat, their data, together with those on transsynaptic catecholamine depletion, hae implications for the understanding of the septal rage phenomenon and possibly other transitory and delayed deficits. These changes could represent the gradual transsynaptic depletion of substances important to the normal functioning of the nervous system.

Electrical stimulation of the amygdala has also been shown to

deplete all catecholamines, with the exception of dopamine, throughout the forehrain (Fuxe and Gunne, 1964); however, this depletion does not extend to the limbic mesencephalic region. Hypothalamically induced defense reactions also deplete forebrain catecholamines, excluding dopamine (Gunne and Lewander, 1966).






- 20 -


The hippocampus, unlike subcortical limbic structures, is relatively low in its catecholamine concentration. It does, however, contain more catecholamines than do some non-limbic regions other than the hypothalamus. Cholinergic mechanisms have been implicated in hippocampal function, both through the application of cholinergic substances (MacLean, 1957; Grossman and Mountford, 1964) and through the studies of the hippocampal theta system. Petsche, et al. (1962) and Stumpf (1965) have demonstrated striking changes in hippocampal activity following treatment with the anticholinergic agent, scopalamine.

Fibers from the hippocampus can be followed in the stria medullaris to the lateral habenular nucleus. This nucleus, the tractus retroflexus, and the interpeduncular nuclei have been shown to contain very large quantities of specific acetylcholinesterases (Koelle, 1954). The medial habenula, which receives no fibers from the hippocampus (Nauta, 1956), but does receive a strong septal projection, shows little evidence of specific acetylcholinesterases. The amygdala appears to have a high concentration of both cholinergic (Koelle, 1954) and catecholaminergic (see Paasonen, et al., 1957) substances. As one might expect, the medial forebrain bundle contains a good deal of acetylcholinesterase as do parts of the mesencephalic reticular formation.

Summary. If the limbic system is defined to include portions of

the midbrain tegmentum, it would appear that two systems course throughout its entire extent. It can be tentatively suggested that the limbic pallial-midbrain system may be primarily cholinergic in its activity, while the striatal-midbrain system is catecholaminergic. While this is most certainly an over-simplification, it may offer a profitable way of organizing some of the structures in the limbic system. The limbic






- 21 -


thalamic nuclei vary in their specific cholinesterase content. Those nuclei related to the pallial components of the limbic system, e.g., hippocampus and cingulate, appear to contain large amounts of specific acetylcholinesterase. These are the anteroventral and lateral nuclei. The dorsomedial nucleus contains only a moderate amount of acetylcholinesterase and is closely related to the striatal septal region. There is evidence that the intralaminar nuclei of the thalamus are at least partially served by a cholinergic mechanism (Grossman and Peters, 1966; Grossman, Peters, Friedman, and Willer, 1965). Chemically distinct systems which somehow underlie sleep (Hernandez Peon, et. al., 1963; Vellute and Hernandez Peon, 1963), arousal (Anderson and Curtis, 1964; Cordeau, 1962; Stumpf, 1965), and feeding and drinking (Epstein, 1960; Fisher and Coury, 1962; Grossman, 1960; Grossman, 1964a; 1964b) have been identified through the local application of cholinergic and catecholaminergic substances. These systems closely parallel similar systems identified through anatomical, ablation, and electrical stimulation studies. While maps based upon chemical studies do not in themselves provide the answer to the problem of adequately characterizing the structure and function of the limbic system, they provide direction for the systematist who might otherwise be lost in a maze of interconnected morphologically indistinguishable fibers and provide links between seemingly unconnected behavioral deficits.

Definition of the mammalian limbic system. Broca's choice of the

word "limbic" as a name for the structures surrounding the rostral brainstem was fortunate in that the term was without functional connotation (MacLean, 1958). The fact that a group of structures in the mammalian brain are in proximity to one another is not sufficient reason for






- 22 -


assuming a close functional relationship between them any more than their separation would justify the assumption that they were not functionally related. The limbic system should not be defined in terms of the spatial relationship of its component structures, but rather in terms of common functional characteristics. Definite functional similarities between the frontogranular cortex, the limbic system, corpus striatum, and the mesencephalic tegmentum have been noted repeatedly, yet there is reluctance to extend the functional limbic system to include these areas and appropriate thalamic structures. Unfortunately, this reluctance is justified in that so far we have been unable to adequately conceptualize these functions. A more adequate characterization of brain functions is necessary before their underlying anatomical substrates can be defined. However, this may be accomplished sooner if emphasis is shifted from the morphological limbic system and is directed at the similarities and differences in the characteristics of the functional limbic system which apparently has telencephalic, diencephalic, and mesencephalic components.

The mammalian limbic system can be defined to include the frontogranular cortex as its most recent telencephalic addition. Its most caudal extent can be found in the dorsomedial mesencephalic tegmentum at the rostral border of the pons. A notable exclusion from the system should be the first system of Pribram and Kruger (1954), that is, the olfactory system.

The telencephalic limbic regions should include the frontogranular cortex in species having such a structure, the cingulate gyrus, entorhinal cortex, periamygdaloid temporal cortex, and Ammon's formation at the cortical levels. The remaining telencephalic structures include






- 23 -


the septal region, basolateral amygdalae, globus pallidus, tail of the caudate, and preoptic areas at the diencephalic level, the habenulae, dorsomedial thalamic nuclei, portions of the dorsolateral thalamic nuclei, the anteroventral thalamic nuclei, and the intralaminar group. The mammillary bodies and portions of the subthalamus may also be included in this group. The mesencephalic component lies in the dorsomedial tegmentum at the lateral border of the periaqueductal gray.


The Limbic System - Function

The septal region

Emotion. In 1929, Fulton and Ingrahamn described angry behavior in dogs following prechiasmal lesions. Spiegel, Miller, and Oppenheimer (1940) reported rage following septal ablation in the cat. Brady and Nauta (1953; 1955) found that rats became hyperemotional following septal ablation; however, this deficit was found to dissipate spontaneously in about two weeks. King and Meyer (1958) found that amygdaloid ablation could attenuate this deficit. King (1959) also found genetic differences to be a factor in determining its magnitude. Harrison and Lyon (1957) attempted to find a correlation between damage to specific components of the fornix and septal region and the appearance of septal rage. They could find no correlation. Teitelbaum (1964) found that cortical spreading depression could reinstate septal rage for the duration of its effect. Yutsey, Meyer, and Meyer (1964) demonstrated that anterior neodecortication coupled with septal ablation in the rat produced septal rage which did not dissipate spontaneously. They found it necessary to handle and gentle their animals in order to alleviate this deficit.






- 24 -


Sodetz, Matalka, and Bunnell (1967) could find no evidence for

septal rage in the hamster and they cite unpublished reports of failure to find septal rage in other species. Buddington, King, and Roberts (1967) have failed to find septal rage in the squirrel monkey and they support Votaw's (1960) similar finding for the Rhesus monkey. Bond, Randt, Bidder, and Rowland (1957) were unable to detect septal rage in their cats. More recently, Moore (1965) has reported finding septal rage in but one of eleven cats used in his study. While the evidence is suggestive of septal rage being a phenomenon specific to the rat, the problem is not that simple. Brown and Slotnick (1966) have reported that female mice demonstrate some of the elements of septal rage following septal rage and McMullen and Slotnick (1967) explicitly state that septal rage is reliably obtainable from the mouse.

In species which respond to septal ablation with a transient

hyperemotionality, there is evidence to suggest that brain catecholamines are involved in the mediation of this behavior. Pirch and Norton (1965) have been able to correlate the occurrence of septal rage with a fall in the concentration of norepinephrine. Raitt, Nelson, and Tye (1961), and Pirch and Norton (1966), have demonstrated that the administration of chlorpromazine to the septal rat attenuates septal hyperemotionality. It is commonly held that chlorpromazine occupies or interferes with activity at catecholaminergic post synaptic sites (Glowinski and Baldessarini, 1966). This is inferred from the fact that chlorpromazine exerts its effect without producing any change in norepinephrine levels in the brain (Gey and Pletscher, 1961). Apparently, existing norepinephrine remains bound in its storage sites.






- 25 -


Chlorpromazine might be expected to retard or eliminate the gradual depletion of norepinephrine which has been correlated with the onset of septal hyperemotionality. Horowitz, Furginelle, Brannick, Burke, and Craver (1963) have demonstrated that pharmacological depression of the amygdala through the administration of antidepressants attenuates septal hyperemotionality. This work supports the earlier work of King and Meyer (1958) who reported a similar result following ablation of the amygdaloid complex.

Conditioned suppression or the conditioned emotional response (CER) has been used as an operational definition of emotion (see Brady 1958a; 1958b; 1961). Brady and Nauta (1953; 1955) found impaired retention of a preoperatively conditioned CER following lesions of the septal region, habenulae, and hippocampus of the rat. This effect persisted after all evidence of septal rage had disappeared. It would appear that this loss of what is usually considered to be a fear response is inconsistent with the reports of hyperemotionality. However, this is so only if one views these data in isolation from data related to other deficits seen to follow septal ablation.

Avoidance learning. The septal region has been shown to be related to the performance of an avoidance response. King (1958) demonstrated that septal ablation resulted in significantly superior performance in the acquisition of a two-way shuttle box avoidance response. McCleary (1961) differentiated active and passive avoidance in testing the behavior of septal cats. He found, as did King (1958), that the lesion facilitated two-way shuttle box avoidance. However, in contrast to the active avoidance problem, septal cats were deficient in acquiring the passive avoidance task. This task required that the






- 26 -


animals withhold an approach response in order to avoid shock. McCleary interpreted these results to be consistent with the stimulation data of Kaada (1951) which had shown the subcallosal area to be involved in the inhibition of motor activity. The passive avoidance deficit was considered to result from the lesioned animal's inability to inhibit a response. Kaada, Rasmussen, and Kviem (1962) verified McCleary's (1961) observation of a passive avoidance deficit in the cat with rats. This early work has since been replicated a number of times (Gurowitz and Lubar, 1966; Lubar, 1964). The enhancement of two-way active avoidance learning has also been demonstrated in succeeding studies (Kenyon and Krieckhaus, 1965; Krieckhaus, Simmons, Thomas, and Kenyon, 1964; Schwartzbaum, Green, Beatty, and Thompson, 1967).

Other learning paradigms. Many of the studies of learning deficits following septal ablation appear to have been influenced by the response disinhibition hypothesis of McCleary (1961). Ellen and Powell (1962a; 1962b) demonstrated that septal rats differed from normal animals in their performance on a fixed interval reinforcement schedule. The septal rats took nearly twice as long as the normal animals to learn to concentrate their responses in the last part of the interreinforcement interval. When they finally acquire this characteristic behavior pattern, septal rats show a higher than normal terminal response rate. Both of these deficits appear to be consistent with a response disinhibition hypothesis. However, Harvey and Hunt (1965) have raised the question of motivational changes by demonstrating that septal rats will respond more often than normals for water reward. Studies to be discussed below (Fisher and Coury, 1962; Grossman, 1964; Vilar, Gentil, and Covian, 1967) have implicated the septal area in the mechanism






- 27 -


underlying drinking behavior. Rich and Thompson (1965) and Thompson and Langer (1963) have found deficits in conditioned avoidance retention and position reversal learning. Douglas and Isaacson (1966) have shown septal ablation to eliminate spontaneous alternation. These results are not consistent with a heightened emotional state, but rather suggest a more response-related mechanism. Zucker (1965), measured food intake in his animals and could find no change; yet, his animals were deficient on both successive discrimination learning and position reversal learning. Gurowitz and Lubar (1966) have shown that septal cats showing evidence of a passive avoidance deficit related to withholding of a motivated approach response to an electrified food cup consumed no more food than did normals in the same situation. The septals differed from normals in the number of responses per gram of food consumed. The septals responded many more times than did the normals. Ellen, Wilson and Powell (1964) and Burkett and Bunnell (1966) have both reported septal ablation to produce deficits on a DRL schedule. Septal animals respond more often than normals during the period in which responses should be withheld.

Feeding, drinking, and sleep. Fisher and Coury (1962), Grossman (1964), Quarterman and Miller (1966), and others have used stimulation of the septal-preoptic region to elicit drinking behavior. These studies seem to indicate a cholinergic substrate underlying this behavior since the application of carbochol elicits drinking while catecholaminergic substances do not. The effect may not be simple, however, for lesions of the septal region also result in increased water intake (Harvey and Hunt, 1965; Vilar, Gentil and Covian, 1967). Chemical stimulation has generally been considered to be the functional opposite






- 28 -


of ablation. Since the data indicate that both stimulation and ablation result in an increment in the amount of water ingested, the possibility of the same behavior being produced by two independent functional systems must be entertained. It may be that the rats in Grossman's (1964) study drank because a drinking mechanism had been activated and that the rats in Harvey and Hunt's (1965) study drank because of a general deficit in dealing with their response tendencies. Carey (1967) has suggested that the drinking mechanism may be anatomically distinct from that which controls response inhibition. However, his histology suggests the need for further examination of this hypothesis. Gurowitz and Lubar (1966) have shown that the food intake of septal rats is no different than that of normals; however, there is a tendency for the septal to make more responses to acquire the food. Sodetz (1965) has observed a weight increase of nearly 20 per cent following septal ablation in the hamster; however, measures of food intake were not taken.

The septal-preoptic region has been cholinergically implicated in the production of sleep (Hernandez Peon, et al., 1963; Velluti and Hernandez Peon, 1963). Although the possibility of producing sleep by the electrical stimulation of this region had been debated for some time (Delgado, 1964), specific evidence for such an effect was lacking. The demonstration of a cholinergic mechanism was the first convincing demonstration of a limbic sleep circuit. Recent electrical stimulation studies have confirmed the presence of this circuit (Sterman and Clemente, (1962).

Related structures

The studies to be considered in this section are representative of






- 29 -


those which may provide some insight into the systematic relationships of structures which have been defined as part of the functional limbic system.

Emotion. Petsche, et al. (1962) have convincingly demonstrated the close functional relationship of the septal region and the hippocampus; however, ablation studies have generally failed to produce modifications in emotionality though the destruction of either the hippocampus or the fornix (Allen, 1948; Rothfield and Harman, 1954). Changes in affective reactivity have been reported following destruction of the amygdala in a variety of species (e.g., Schreiner and Kling, 1956). This effect is generally characterized as a loss of fear, or ferocity but operationally what has been observed to follow amygdalectomy is a willingness on the part of the organism to approach objects which would have been avoided preoperatively and a failure to engage in intra and interspecific aggression. (Plotnik, 1966; Rosvold, Mirsky, and Pribram, 1954; Weiskrantz, 1956). Cingulate ablation has been shown to produce affective changes in rats and primates, but there has been little agreement as to the nature of these changes. Kennard (1955) found cingulectomized cats to be aggressive fighters, while Mirsky, Rosvold, and Pribram (1957) found their cingulectomized monkeys to be less fearful of man but unchanged in their reactions to conspecifics. Pechtel, McAvoy, Levitt, Kling, and Masserman (1958) reported their animals to be slightly more aggressive toward both man and conspecifics. Spiegel and Wycis (1949) and Wycis and Spiegel (1951) have reported that ablation of the dorsomedial thalamic nucleus produces a decrement in emotional reactivity. Destruction of the frontogranular cortex was often used in the treatment of emotional disorders (Fulton, 1951).






- 30 -


Heller, et al. (1962) have reported that lesions in the dorsomedial mesencephalic tegmentum produced an increment in the aggressiveness of the rat toward conspecifics without any overt evidence of hyperemotionality to human handling or any other notable behavioral changes.

DeMolina, and Hunsperger (1962), electrically stimulated various hypothalamic sites and have mapped medial and lateral regions which somehow underly flight and defensive reactions. These areas extend throughout the rostrocaudal axis of the hypothalamus and are at least partially under the control of the amygdala (Hilton and Zybrozya, 1963; Zybrozya, 1960; Fernandez DeMolina and Garcia-Sanchez, 1967). The diversity of behavioral changes associated with stimulation and ablation of limbic structures suggests that the observed emotional disorders may be phenotypic byproducts of more specific functional disorders and that the limbic system is not a mechanism for the elaboration and apprehension of emotional states, but rather a more general system dealing with external and internal events. Emotional anomalies would follow damage to this system because the organism is no longer able to process data in an adaptive manner.

Learning. Of the structures under consideration, the hippocampus most closely parallels the septal region with regard to the observable properties of learning deficits produced by its destruction. This should not be taken to mean that these areas subserve similar or even closely related functions, for the analysis of the learning process has not reached such.a level of sophistication that it can be said with any certainty that similarities in observable deficits reflect the disruption of similar neural processes. It is conceivable that the septal region and the hippocampus each contribute to the learning process in






- 31 -


such a way that, in many instances, the destruction of one or both would have the same behavioral outcome. Isaacson, Douglas, and Moore (1961) have demonstrated superior two-way active avoidance learning and deficient passive avoidance performance following bilateral hippocampectomy in the rat. Burkett and Bunnell (1966), and Ellen, Wilson and Powell (1964) have demonstrated a deficit in DRL performance following septal ablation in the rat and Clark and Isaacson (1965) and Schmaltz and Isaacson (1966) have demonstrated a similar deficit in hippocampectomized rats. There is some suggestion that the production of this deficit depends upon the method of training used in shaping the animal's DRL performance. If the animal is shaped with continuous reinforcement and then shifted gradually over to the DRL schedule or if it is trained from the beginning on a DRL schedule, the deficit is not as severe as it is when the animal is shifted from continuous reinforcement to the DRL schedule in one step (Schmaltz and Isaacson, 1966). An important distinction between the deficits produced by septal ablation and those which are obtained from hippocampectomy may be found in the work of Douglas and Raphelson (1966b) and Sodetz (1965). Douglas and Raphelson (1966b) found septal rats to be no more active in an exploratory situation and found nearly a 50 per cent decrement in activity as:mease ured in an activity wheel. Sodetz (1965) found septal hamsters to decrease their activity over trials on an open field activity board at the same daily rate as normal hamsters. Sodetz (unpublished observations) found that septal hamsters averaged approximately 50 per cent fewer revolutions per night in an activity wheel. Hippocampectomized rats, however, do not exceed normal rats in the initial level of






- 32 -


activity observed in an exploratory setting. However, they persist in this behavior over trials and do not show the activity decrement associated with habituation (Roberts, Dember, and Brodwick, 1962). While septal and hippocampal ablation (Douglas and Isaacson, 1964; Douglas and Raphelson, 1966a) both reduce the tendency of the rat to alternate its responses, septal ablation does not produce a deficit in maze learning (Thomas, Moore, Harvey, and Hunt, 1959) which can be attributed to this deficit, whereas, hippocampal ablation does (Douglas and Pribram, 1966).

Damage to the cingulate cortex has been shown to result in a deficit in two-way active avoidance (McCleary, 1961; Peretz, 1960), and normal performance in passive avoidance (Lubar, 1964). It seems likely, however, that significant improvement in passive avoidance learning would be difficult to detect since normal animals learn this type of task very quickly. Lubar (1964) included a group with combined cingulate and septal ablations. These animals showed no deficit in passive avoidance. The fact that cingulate ablation counteracts the effect of septal lesions suggests that this region has a role in this behavior which is not discernable in comparisons with normal animals. It seems likely that the cingulate gyrus serves a function in avoidance learning which is phenotypically antagonistic to that of the septal region and the hippocampus.

Thomas and Slotnick (1962) could find no evidence of a maze

learning deficit in the rat following cingulate ablations. However, they did find a deficit in conditioned avoidance responding. They suggested (Thomas and Slotnick, 1963) that this deficit was produced






- 33 -


by the tendency of the animal to "freeze" in response to the CS. Hypothesizing that food deprivation would tend to increase the activity level of the cingulectomized animals and thereby minimize freezing, they added this seemingly irrelevant drive to their paradigm and found the performance of the animals to compare favorably with that of their controls. Cingulate lesions, therefore, differ from hippocampal ablation in their effects on maze learning and active and passive avoidance learning. Septal and cingulate lesions appear to be similar in their failure to adversely affect maze learning.

Horwath (1963) has found that bilateral amygdalectomy produces a deficit in one-way active, two-way active, and passive avoidance. The effect on passive avoidance, though slight, was statistically significant. The generality of this deficit is suggestive of a disruption of a function critical to the general learning process. The work of Kluver and Bucy (1937) supports this view insofar as the behavior of their monkeys can be characterized as being relatively insensitive to the stimulus characteristics of their environment. Schwartzbaum (1960) has demonstrated that the amygdalectomized monkey is relatively insensitive to changes in the magnitude of reward. Hi-low shifts and low-hi shifts had little effect on their response rate. Perhaps, the contradictory character of nearly every study which has investigated amygdaloid function (See Goddard, 1964) is not a product of the convergence of a great number of functional systems in the region of the amygdala, but is the result of the disruption of the mechanism which makes lawful decisions according to the rules of reinforcement. If this is the case, then the failure of studies of the amygdala to agree may be because the behavior of the amygdaloid animal is being interpreted in terms of the






- 34

laws which govern the behavior of the normal animal while the animal is behaving according to a different set of laws. Douglas and Pribram (1966) have suggested that the amygdalectomized animal is incapable of responding in terms of reward and have found evidence consistent with the hypothesis that the animal conducts its affairs entirely in terms of non-reward.

Thompson and Langer (1963) and Rich and Thompson (1965) have

selectively ablated various limbic, thalamic, and hypothalamic sites in an attempt to delimit a functional system underlying position reversal learning and avoidance conditioning. Their studies have implicated the septal region, hippocampus, mammillary bodies, anterior thalamic, and midline thalamic nuclei in a system which somehow subserves functions necessary to the retention of these habits. Roberts and Cary (1956) have reported that lesions of the dorsomedial thalamic nucleus interfere with the acquisition of a conditioned fear response. Thompson and Rich (1961) have found lesions of the interpeduncular nuclei to temporarily disrupt performance of an avoidance test. Hatton (1965) has confirmed their result, but reports no evidence of interference with a discrimination problem.

Some relevant pharmacological studies. Pharmacological studies

offer indirect support for many of the results sited above. The septal region and hippocampus have been shown to be important to the maintenance of spontaneous alteration (Roberts, et al., 1962; Douglas and Raphelson, 1966a). In addition, there is evidence that cholinergic mechanisms may mediate at least part of their function. If this is so, then the administration of centrally acting anticholinergic substances might be expected to produce results very similar to those obtained in lesion






- 35 -


studies. It should be emphasized, however, that the systemic administration of these agents precludes any exact localization of their site of action. Any attempt at relating these effects to a specific area of the brain must be viewed with some reservation.

The systemic administration of scopalamine produces a reversible loss of spontaneous alternation behavior (Douglas and Isaacson, 1966; Meyers and Domino, 1964). Scopalamine has also been shown to produce a reversible passive avoidance deficit (Meyers,1965) and to attenuate a conditioned fear response (Vogel, Hughes, and Carlton, 1967). Atropine sulfate, another anticholinergic agent, disrupts maze learning, while methyl atropine, a related substance with little central effect and a strong peripheral action, does not interfere with maze learning (Whitehouse, Lloyd, and Fifer 1964). Chalmers and Erickson (1964) have found cholinergic agents to be effective in producing both an acquisition and retention deficit in a shuttle-box avoidance task and in a lever pressing conditioned avoidance problem. They related this deficit to the tendency of their animals to freeze in response to the conditioned stimulus. This was essentially the observation made by Thomas and Slotnick (1963) who tested their hypothesis by adding an irrelevant drive to elevate the activity level of their cingulectomized animals. Both Chalmers and Erickson (1964) and Thomas and Slotnick (1963) report the observation of behavior suggesting that their animals responded with fear to the conditioned stimulus, but simply failed to initiate the avoidance response.

A study by Hearst (1964) demonstrated that both scopalamine, an anticholinergic, and amphetamine, an agent which induces the release of catecholamines, flattened the generalization gradient of the monkey






- 36 -


performing during the acquisition of an eight choice shock avoidance discrimination problem. Monkeys treated with either of these substances tended to distribute their responses more widely over the possible stimuli, while normal monkeys produced the typical curves associated with generalization to closely related stimuli. In this case, two chemical systems may subserve phenotypically similar functions but it may be that these functions are genotypically distinct.

Catecholaminergic agents have been found to be related to behavioral changes usually associated with the limbic system. Andersson and Larsson (1957) demonstrated that surgical separation of the frontogranular cortex of the dog blocked the reduction of food and water intake associated with the administration of amphetamines. Hunt (1957) and Raitt, et al. (1961) reported meprobamate and chlorpromazine to be effective in attenuating septal hyperemotionality. Randall (1961) has shown chlordiazepoxide (Librium) also to be effective in this regard.

Similarities and interactive effects between limbic ablations and the manipulation of pharmacological substances related to cholinergic and catecholaminergic mechanisms have been demonstrated. Studies such as these may provide a basis for the identification of genotypically distinct functional systems in instances where the observable end products of their activity are phenotypically identical. The limbic system and species-specific behavior

Species-specific behaviors are responses common to a given class of organisms. No assumption is made as to whether such behaviors are learned or innate. Furthermore, no distinction is made between those behaviors which are structured by internal mechanisms and those which may be imposed upon the organism by the properties of its ecological






- 37 -


niche. Species-specific behaviors are those response patterns which are phenotypically identical in all members of a species which have been exposed to the natural selection factors most characteristic of those operating on the species as a whole. Members of a species which are not exposed to these same environmental factors may differ with respect to many of the learned components of these behaviors while they may or may not retain the innate behaviors common to their species. Consequently, studies of the species-specific behaviors of laboratory animals may produce results which are limited in their generality. This factor should be given consideration when evaluating the significance of such studies. Another possible limiting factor rests in the conditions under which the behavior of the organism is observed. The more restrictions the experimenter places upon the behavior of the animal, the more likely it is that the obtained results will not be representative of the species under consideration.

The problem still remains as to how a study may be identified as being related to species-specific behavior. The criteria for delimiting species-specific behaviors include its presence throughout a representative sample of the species and the fact that it is idiosyncratic to the species under consideration. More subtle distinctions are made on the basis of the restrictions placed upon the organism. It can be said that the study of the behavior of the rat on an open field activity board is species-specific if no restrictions are placed upon the behavior of the animal. The response pattern of a rat in such a situation might differ markedly from that of a pigeon under the same circumstances. If, however, the experimenter measures only the number of squares entered by the animal, it is not likely that his observations






- 38 -


can be characterized as dealing with species-specific behavior. Similarly, if a rat is placed alone in an empty box, with a drinking tube protruding from the wall, it may be observed to drink following the application of some chemical to the hypothalamus. This behavior cannot be characterized as species-specific, however, for the nature of the apparatus restricts the organism to a discrete set of responses. In such a study, it might be found that the introduction of nesting materials into the box might result in the animal ceasing its drinking behavior and initiating nest building, or perhaps going to sleep.

The study of species-specific behavior requires that internal and environmental variables be left relatively free to operate in the experimental setting. A measure of control is lost in such a design and it becomes difficult to make explicit statements about the properties of the functional systems which underlie the observed behavior. However, it is likely that the organism employs most of its faculties in the conduct of its affairs and the sampling of a broad range of behaviors may produce evidence of deficits which might take far longer to detect in a more restrictive experimental setting.

MacLean (1958) has suggested that the limbic system is the mechanism responsible for those behaviors which are adaptive in the preservation of the individual and the species. This functional characterization of the limhic system may be correct insofar as all adaptive behavior is, by definition, consistent with the preservation of the individual and the species. It is reasonable to assume that many of the same laws which govern the behavior of the organism in a laboratory learning task also control the naturally occurring behaviors of the free-living organism.






- 39 -


Studies of the limbic system,in the laboratory setting, have been discussed in preceding sections. This section will deal with observations of the effects of manipulation of the limbic system on species-specific behavior. Recent systematic observations of changes in species-specific behavior following damage to the limbic system (Brody and Rosvold, 1952; Bunnell, Friel, and Flesher, 1966; Bunnell, et al., 1967; Bunnell and Smith, 1966; Bunnell, et al., 1965; Clemente, Green, and deGroot, 1957; Delgado, 1964, 1966; Fisher, 1956; Fuller, Rosvold, and Pribram, 1957; Green and Kling, 1966; Kennard, 1955; Kling, 1962; Mirsky, et al., 1957; Pechtel, et al., 1958; Plotnick, 1966; Rosvold, et al., 1954; Schreiner and Kling, 1953; Sodetz, 1965; Sodetz and Bunnell, 1967a; 1967b; Stamm, 1954; 1955; Weiskrantz, 1956) have repeatedly demonstrated that this system and related structures are closely related to the mediation and control of these behaviors. Amygdalectomy has been shown to result in the reduction of the aggressiveness of the cat (Schreiner and Kling, 1953), monkey (Plotnik, 1966; Rosvold, et al., 1954), dog (Fuller, et al., 1957), hamster (Bunnell, et al., 1965), and rat (Bunnell, 1966) toward conspecifics. In contrast to this, the same lesion produces a taming effect in a variety of nondomesticated species (Schreiner and Kling, 1956) and a willingness on the part of the monkey to approach man (Weiskrantz, 1956) which might be interpreted as an increased interspecific aggressiveness. Rosvold, et al. (1964) suggested that amydalectomy in the monkey resulted in the animal decreasing in rank in a dominance hierarchy. Plotnik (1966) and Bunnell, et al. (1965) found a decrease in the frequency of social






- 40 -


interaction following amygdalectomy in the monkey and hamster. In the hamster (Bunnell, et al., 1965) a general depression of social activity was observed which transcended both aggressive and submissive behavioral categories. This data is suggestive of a tendency on the part of the amygdaloid animal to become asocial rather than submissive. Cingulate ablation has also been related to social behavior. Kennard (1955) found that cingulectomized cats fought more than normals. Mirsky,et al. (1957) could find no change in the social behavior of the cingulectomized monkey, but like the amygdaloid monkey these animals appeared to be less fearful of man. Pechtel, et al. (1958) noted the discrepancies in the results of earlier studies, but could only add that their animals appeared to be somewhat more aggressive toward both man and conspecifics. They did, however, observe that their animals failed to care for and nurse their young. Stamm (1955) had earlier observed a similar deficit in the maternal behavior of the cingulectomized rat. These rats failed to care for their young following birth. If, however, foster pups, which were old enough to attach themselves to the female, were placed with the animal, it tolerated their nursing, following which the female began to show more normal pup retrieving behavior. Such animals also failed to build nests and, in another study, Stamm (1954) found them to be deficient in hoarding. A nest building and hoarding deficit has been observed to result from septal ablation in the hamster (Sodetz, 1965), but not from ablation of the amydala (Bunnell, et al., 1965). Hippocampal ablation also produces some deficit in hoarding (Bunnell, et al., 1967); however, it is less severe than that seen






- 41 -


in the septal hamster. Both pre-frontal (Brody and Rosvold, 1952) and mesencephalic tegmental (Heller, et al., 1962) damage have been shown to affect the aggressive behavior of the animals studied. Prefrontal lobotomy produced changes in the behavior of monkeys very similar to those seen following cingulectomy or amygdalectomy. In the case of the pre-frontal animals, however, the persistent approach response to previously avoided animals appeared to disappear within a few months and the dominance heirarchy of the monkeys stabilized in a form which differed little from that seen preoperatively. Rats with dorsomedial mesencephalic tegmental lesions (Heller, et al., 1962) were found to be placid by human handlers, but these same animals engaged their normal cagemates in almost continuous aggression throughout the 35 days they remained in the study. Sodetz (1965) and Sodetz and Bunnell (1967a; 1967b) found ablation of the septal region of the hamster to produce changes in aggressive behavior which were consistent with the social history of the animal. Septal hamsters which had no social history of defeat by normal opponents became dominant and displayed more aggressive behavior in post-operative trials. Septal hamsters which had experienced defeat before surgery became more submissive following ablation. Bunnell, et al. (1967) found hippocampectomized hamsters which had no preoperative social experience to be submissive in postoperative trials unless virtually unopposed by their opponents. They also observed that ablation of the hippocampus in animals which had a history of defeat produced an increment in submissive behavior similar to that observed to follow septal ablation in submissive animals. Their comparison of the observed sequences of behavior of hippocampal and septal hamsters






- 42 -


indicated that the two groups, though showing similar changes in the frequency of submissive responses, differed in terms of the sequential arrangement of these behaviors. In this regard, septal animals were very similar to normals while hippocampal hamsters differed from both groups.

Pharmacological studies of the species-specific behavior of a number of forms have been done. Apparently, alterations in speciesspecific behavior provide a very sensitive instrument for the screening of drugs. Horowitz, et al. (1964) report that mouse-killing by rats was precluded by the administration of anti-depressant drugs, whereas dozens of tranquilizers tested were without effect. It was their hypothesis that anti-depressants exert their effect by selectively depressing the amygdala. Karli (1960) has shown that septal ablation increases the likelihood of a given rat killing mice. Tranquilizers have been shown to diminish septal hyperemotionality (Raitt, et al., 1961). This may be evidence for a functional distinction between emotionality changes produced by limbic damage and other behaviors .such as mouse-killing which do not appear to be affected by the same class of pharmacological agents. Bignami (1964) has shown that anticholinergic substances interfere with the mating behavior of the male rat. He suggests that this is due to a disinhibition of the orienting reflex. This hypothesis appears to be consistent with the evidence for cholinergic mediation of hippocampal theta rhythm (Stumpf, 1965) and current concepts of its functional significance (See Douglas, 1967). DeVanzo, Daughterty, Ruckart, and Kang (1966) have demonstrated that monoamineoxidase inhibitors reduce






- 43 -


fighting behavior in mice. They could not, however, detect any difference between the brain norepinephrine levels of fighters and nonfighters, which lead to the conclusion that the effect must be mediated through one of the other catecholamines. Serotonin could be a likely possibility because emotionally reactive and non-reactive strains of rats have been shown to differ in their brain serotonin concentrations (Sudack and Maas, 1964); however, Heimstra and Saller (1967) have recently demonstrated that mice chronically treated for a period following weaning with an agent which increases brain norepinephrine levels will be more likely to dominate their opponents in aggressive encounters as adults. Silverman (1966) found that chlorpromazine, which depresses the activity of adrenergic systems, decreased aggressive social interaction in laboratory rats, tended to increase flight in response to threat and attack, and had no effect on non-social behavior. Amobarbital, which is a CNS depressant, increased social behavior in general and threat responses in particular. Norton (1957) studied the effects of a number of pharmacological agents on the social behavior of monkeys, hamsters, and cats. She found consistent effects across species. Chlorpromazine increased social behavior in general, while decreasing behavior associated with agitation, defense, and attack. Behavior such as grooming, which she took to be indicative of contentment was seen more frequently in animals treated with chlorpromazine. Amphetamine, which excites cateholaminergic systems, was found to decrease sociability, increase attack and defense, and increase evidence of excitement.

Both cholinergic and catecholaminergic systems appear to be involved in species-specific behavior. The fact that most limbic






- 44 -


structures appear to be served by both systems permits the possible number of combinations of deficits which could result from ablation of just one of these structures to be large. It is not surprising that pharmacological studies which have employed the systemic administration of psychotropic agents do not present a consistent pattern of results. There is no way to be certain just where in the CNS these agents are acting, although the lesion data strongly support the view that they are acting on limbic structures. The work of Sodetz (1965) and Sodetz and Bunnell (1967a; 1967b) has shown septal ablation to produce changes in both the social behavior and non-social behavior of the hamster. These effects have not been amenable to description in terms of any uniprocess model of septal function (e.g., McCleary, 1961). Furthermore, the studies of Bunnell, et al. (1965) and Bunnell, et al. (1967) have demonstrated differences and similarities in the effects of septal and hippocampal ablation and differences in the effects of septal and amygdaloid ablation. The pharmacological data suggest that both cholinergic and catecholaminergic systems in these structures may be operative in the mediation of species-specific behavior. A septal lesion would of necessity interdict both types of systems. Therefore, it is likely that the failure to develop an appropriate uniprocess model to account for the observed changes in behavior is due to interference with more than one chemically distinct functional system. The present study was designed to attempt to identify chemically distinct functional systems subserving species-specific behavior through the application of cholinergic and catecholaminergic agents and their antagonists to the septal region of the forebrain.












METHOD


The present study was conducted as a series of experiments each of which differed in the particulars of their methodology. What follows is a general description of the elements common to the entire study. Conditions unique to a given experiment will be discussed in separate sections devoted to each individual experiment.


Subjects. One hundred fifty Manor Farms random bred adult male hamsters (Mesocricetus auratus) were used in the study. All of the Ss were obtained as weanlings at an age of about 27 days. Upon receipt at the laboratory, all Ss were transferred to single cages and housed individually until reaching an age of from 90 to 130 days. At that time, samples were drawn from the subject population and assigned to individual experiments. While in the laboratory, all Ss had paper nesting material, food, and water continuously available to them. Handling of the Ss was kept at a minimum.

The Ss were raised in social isolation in an attempt to maximize the likelihood of their engaging in aggressive social interaction. Hamsters which have been raised in a group are less likely to aggressively engage another similarly experienced animal. Social isolation appears to be an effective and legitimate means of avoiding this difficulty. Hamsters which have experienced defeat in aggressive interaction cease to initiate such activities. In a group of hamsters, one animal characteristically dominates all others, therefore only the dominant animal from a given group would readily engage an opponent. Since isolation-reared animals have had little

- 45 -






- 46 -


experience with defeat, a far greater proportion of them will engage opponents.

In each experiment the Ss drawn from the subject population were assigned to pairs matched by weight and age. Differences in age and weight within a given pair did not exceed five days and five grams, respectively. Differences of this magnitude appear to in no way affect the outcome of an aggressive encounter. Following their assignment to pairs, all Ss were transferred from the single cages in which they had been living to the dominance cages to be described below. They remained in these cages throughout the experiment. No S was ever used in more than one experiment.

Apparatus. The three primary pieces of apparatus used in the study were the dominance cages, recording apparatus, and cannulae.

The dominance cages, in each of which a pair of Ss were both housed and tested, were constructed from double rat cages (45 cm. x 24 cm. x 20 cm.). They were divided into three equal compartments by two metal partitions. The floors of the cages were lined with 1/4 inch hardware cloth. One animal lived in each of the two end compartments which were separated by the third center compartment. The center compartment served as a "neutral" area and was included in the design to permit the operation of territorial effects. Access to all compartments could be made available by raising a guillotine door in each of the partitions. This permitted unrestricted interaction between members of a pair.

The recording apparatus was designed to provide a continuous record of the occurrence of 27 behaviors which can be observed in the interaction of two adult male hamsters. The apparatus consisted






- 47 -


of a Gerbrands multi-channel event recorder, one channel of which was assigned to each animal. A third channel was connected to a Kramer clock, which activated the channel once each second.

A code symbol was assigned to each of the 27 behaviors. Two

trained observers, one for each animal in a pair, sequentially listed the behaviors they observed thereby producing a written record of all the behaviors of each animal for a given five minute observation period. As each behavior was noted, the observer also closed a switch assigned to his S, which entered the occurrence of the behavior on the output of the recorder. The output of the recorder, including the time base, was later collated by computer with the written response record to yield data on response frequencies, durations, sequences, and inter-animal response contingencies.

The double-walled cannulae used for chemical stimulation were

constructed from three different gauges of stainless steel hypodermic tubing. These were selected because, when placed concentrically within one another, each formed a close fit within the next larger. The outer cannula consisted of a piece of 22 gauge tubing 11 mm. in length. This piece formed a guide for the inner cannula and was the portion inserted into the brain. Each of the inner cannulae consisted of a piece of 28 gauge tubing soldered within a piece of 18 gauge tubing approximately 3 mm. in length. To complete the cannula, the inner component was inserted into the outer guide with the 18 gauge tubing forming a snugly fitting cap. When fully inserted the tip of the inner cannula fit flush with the tip of the guide. A Hamilton microliter syringe fitted with a 28 gauge needle and a guide






- 48 -


to limit insertion to exactly 11 mm. was used to inject chemicals into the brain. This was accomplished by removing the inner cannula, and in its place, inserting the needle of the syringe. Following administration of the chemical, the needle was withdrawn and the inner cannula replaced.

Response inventory. Table 1 presents a list and a brief description of each of the behaviors seen in the interaction of two adult male hamsters. The behaviors can be organized into eight categories such as hoarding and fighting. These response categories can be grouped into two classes, the social activities which in some way involve another animal and the non-social categories which bear no relationship to the other animal.

Surgery. Following assignment to pairs and transfer to the

dominance cages, one member of each pair of Ss underwent surgery to implant a single cannula and, in some cases, to produce a lesion in addition to the placement of the cannula. Surgery was performed under sodium pentobarbital anesthetic (90 mg/kg). Atropine sulfate was administered intraperitoneally approximately 30 minutes before administration of the anesthetic. The S's head was shaved and the S was mounted in a stereotaxic instrument. Following incision of the scalp, the skull was cleaned and air-dried. The head was leveled between the parietal-occipital and the frontonasal skull suture and a trephine hole 3 mm. in diameter was bored over the midline about

6 mm. rostral to bregma. Three small holes, one in the dorsal aspect of each of the parietal bones and one on the midline in the nasal bones just rostral to the frontonasal suture, were bored in the skull




































Table 1. A brief description of the responses which can be seen in the interaction of two adult male hamsters.






- 50 -


RESPONSE CATEGORIES

NON-SOCIAL


HOARDING


M = Middle or Neutral Area HC = Home Cage OC = Opponent Cage E = Exploration of Apparatus and Contents


UA = Pouch Paper HA = Hoard Paper UO = Pouch Food HO = Hoard Food


SOCIAL


SOCIAL INVESTIGATORY


N = Sniff Opponent (Requires Contact) N'= Sniff (Analgenital)


DEFENSE


P = Defense Posture (Upright) P'= Defense Posture (Side)


SUBMISSION


AGGRESSION


A = Attack Posture (Side) A'=Attack Posture (Underneath) B = Bite
C = Chase


FIGHTING


S = Spar F = Fight x/y = Pin

MISCELLANEOUS


T = Tail Lift T'=Tail Lift (With Adduction
of Hindlimb) Z = Freeze

ESCAPE

W = Attempted Escape from Cage L = Flee


I = Orient Toward Opponent V = Vocalize G = Groom


ACTIVITY






- 51 -


and a small stainless steel machine screw was inserted in each of them. This was done to provide binding sites for the dental cement which would hold the cannula in place. The skull was then cleaned and air-dried again. The cannula was stereotaxically placed through a slit in the dura into the septal region. An attempt was made to place the tip of the cannula into the midline of the septal region at a point midway in its rotrocaudal extent. It was hoped that such a placement would maximize the likelihood of the chemical bilaterally affecting most of the septal region. Placement of the tip of the cannula into the midline required that the cannula be inserted at an angle so as to avoid the bleeding associated with rupture of the saggital sinus. The stereotaxic coordinates used were 6 mm. (+ .3 mm.) anterior to bregma, 1 mm. lateral to the midline, and 4.25 mm. ventral to the surface of the cortex. The angle of penetration was 13.5 degrees. The dried and exposed area of the skull was then covered over with an application of Caulk "Grip" dental cement. When the cement was dry, any rough or sharp edges were removed with a dental burr and an antibiotic salve was applied to the area of the incision. The animal was then returned to its compartment in the dominance cage and permitted to recover for a period of from 11 to 14 days.

A number of Ss received surgical treatment which differed somewhat from that described above. One group of such Ss had the cannula placed in the lateral ventricle at a point 4.25 mm. rostral to bregma. Another group received either bilateral RF thermocoagulation of the amygdala or RF thermocoagulation of the body of the fornix in addition to having a cannula placed in the septal region.






- 52 -


Procedure. Following the post-operative recovery period testing was begun. Fresh solutions of the appropriate chemical dissolved in physiological saline were prepared daily in a concentration of

2 mg/ml. An operated S and its normal opponent were transported in their dominance cage to the testing room. Here the operated S was removed from the cage and placed in a closed plastic container. Nitrous oxide gas was introduced into the container, lightly anesthetizing the animal. The S was then removed from the container and during the 5 to 10 second period during which the S remained anesthetized the inner cannula was withdrawn and 1 microl. of either the chemical in solution or physiological saline was administered to the animal via the outer cannula. The inner cannula was then cleaned and replaced. During the initial stage of each experiment an animal either always received the chemical or always received saline. While the operated S was being injected with the chemical, an assistant anesthetized the S's normal opponent with nitrous oxide gas. Both Ss were then returned to their respective compartments within the dominance cage. The Ss were left undisturbed for 10 minutes. At the end of this period the test trial was begun. A trial consisted of 5 minutes of unrestricted interaction between members of a pair. Two trained observers, one for each animal, recorded the behavior of the animals using a response inventory of 27 behaviors which can be observed in the interaction of two adult male hamsters. Concurrently, each written response was also entered into the record of a multi-channel event recorder. One channel served as a time base. A computer later collated the output of the recorder with the written response record. Subjects were never run more than one trial a day.






- 53 -


Histology. Following the completion of testing, all operated Ss were anesthetized and perfused with saline followed by formalin. Their brains were then removed, fixed in formalin, and embedded in celloidin. The brains were sectioned at 30 micra and every fourth section was stained with cresyl violet.

Subjects dropped from the study. The difficulty in maintaining chronic indwelling cannulae over a period of several weeks resulted in the loss of a number of Ss. The rule applied throughout the study called for excluding the data from these Ss if they had completed fewer than three of five trials. If they had completed at least three trials, their brains were recovered and the data obtained from them was retained. The same rule was applied to several Ss which died during the course of the study.

If histological examination of the brain of a S showed the

cannula to be placed somewhere other than within the septal region, the data from that S was treated separately and is presented in Appendix A.

Saline controls. Eighteen pairs of Ss in the group which served as saline controls survived to be included in the study. In each pair one of the Ss had a single cannula placed within the septal region. Ten minutes before the beginning of each trial, each S received 1 microl. of sterile isotonic saline administered via the cannula. This group served as a control against which to compare the remaining groups.

Norepinephrine group. In this group, six pairs of Ss met the criterion for remaining in the study. All six operated Ss received






- 54 -


2 microgms. of norepinephrine (D-L-Arteronol) dissolved in 1 microl. of isotonic saline, 10 minutes before the beginning of each trial.

Intraventricular norepinephrine group. This group, consisting of three pairs, received the same dose of norepinephrine noted above. However, the cannulae in these Ss were directed into the lateral ventricle. This group was included to control for any general effect which might have resulted from the administration of norepinephrine.

Phenoxybenzamine group. Nine pairs of Ss survived to be included in this group. The operated S in each pair had a cannula placed in the septal region. Phenoxybenzamine (Dibenzyline) was administered via the cannula 10 minutes before each trial. Each dose consisted of

2 microgms. of phenoxybenzamine dissolved in 1 microl. of isotonic saline. Phenoxybenzamine is known to block the action of norepinephrine by occupying the sites at which it acts. Therefore, it does not have the disadvantage of many adrenergic blocking agents which cause an initial release of norepinephrine by occupying its storage sites (Goodman and Gilman, 1964).

Intraventricular phenoxybenzamine group. There were four pairs of Ss in this group. They differed from the phenoxybenzamine group only in that the cannulae in these Ss were directed into the lateral ventricle.

Phenoxybenzamine - fornix resection group. Four pairs of Ss survived in this group. One S in each pair had a cannula placed within the septal region. In addition, the S underwent bilateral RF thermocoagulation of the fornix during the operation to implant the cannula. Ten minutes before each trial, these Ss received the same dose of phenoxybenzamine as did the other groups in the series.






- 55 -


Phenoxybenzamine - amygdalectomy group. One member of each of the four pairs in this group had a cannula placed within the septal region. In addition, the S underwent bilateral RF thermocoagulation of the amygdaloid complex during the operation to implant the cannula. Each operated S received the same dose of phenoxybenzamine 10 minutes before each trial, as did the other Ss in the series.

Carbachol group. Four pairs of Ss survived in this group. The cholinergic substance used in this series was carbachol which has a stronger and more lasting effect than acetylcholine (Goodman and Gilman, 1964). Each operated S received 2 microgms. of carbachol dissolved in 1 microl. of isotonic saline 10 minutes before each trial.

Atropine group. The operated S in each of the four pairs in this group received 2 microgms. of atropine sulfate dissolved in

1 microl. of isotonic saline 10 minutes before the beginning of each trial. Atropine sulfate is an anticholinergic agent.

Summary of groups. A summary of the groups used in the study and the number of operated Ss in each is presented in Table 2.

Statistical analysis. The data collected in the study presented a special problem in that no complete statistical analysis of the data could be made. The behavior of the hamster at any instant during an observation period is dependent upon the behavior which occurred in the immediately preceding instant and perhaps back through a chain of as many as eight or ten responses. Similarly, the occurrence of the given behavior determines with a specific probability level what the succeeding responses will be. These




































Table 2. A summary of the groups used in the study.
















Number of Ss Operated


Number of Ss Surviving in Study


Cholinergic


Catecholaminergic


Carbachol (C) Atropine (A)


Norepinephrine (N) Intraventricular
Norepinephrine (IN) Phenoxybenzamine (P) Intraventricular
Phenoxybenzamine (IP) Phenoxybenzamine
Fornix (PF)

Phenoxybenzamine Amygdalectomy (PA)


4


Saline Saline (S) 25 18 TOTAL 75 56


- 57 -


Group






- 58 -


internally generated sequential dependencies are further complicated by the impact of the behavior of its opponent upon the animal. It can be said with some certainty that the responses observed in the study are sequentially dependent on one another. The majority of parametric and non-parametric statistical tests require the assumption that the events in question be independent, that is, the occurrence of event A on trial 1 in no way affects the probability of its occurrence on trial 2 nor does it affect the probability of the occurrence of events B,C,D,...K. It is not possible to make this assumption in connection with the behavior observed in this study. If the animal has pouched food pellets and its cheek pouches are filled, the probability of it continuing to pouch pellets is very small and the probability of its emptying its cheek pouches is much greater than it would be if they were only partially filled or empty. With 10 different experimental conditions, the interaction of two animals, and a behavioral repertoire of 27 responses to consider over as many as five trials, the complexity of the sequential dependencies are such that they defy the application of any conventional statistical test. A Markov chain analysis can handle data such as these, for Markov chains were developed to deal with sequential dependencies. The complexity of the dependencies in the present study, however, demand a very sophisticated appreciation of Markovian analyses and the generation of thousands of responses so as to permit the specification of the exact probabilities of each response following any other response under all possible relevant conditions. Such an analysis is beyond the capabilities of the author. The data were






- 59


examined with the aid of statistical tests where possible. However, because of the sequential dependencies present in the data, this analysis was very limited in scope and could not consider the various treatment effects in detail. An effort was made to be conservative in assessing those results which were not amenable to statistical test.














RESULTS


A group of eighteen Ss was used as a saline control against

which to compare the effects of the administration of chemicals to the septal region of the forebrain. The behavior of these Ss was not expected to differ from that of their normal opponents. As many saline (S) animals were expected to be dominant as were expected to be submissive. Of the eighteen Ss in this group, eleven were submissive, as rated by the two observers, and seven were dominant. The response profile shown in Figure 1 reflects this tendency toward submissiveness. Figure 1 represents a comparison between the proportion of responses seen in each response category for the S group, represented by the bars, and the response profile of their normal opponents represented by the base line from which the bars are positively and negatively reflected. The ordinate represents the difference between the observed proportions in a given category. In the exploration (E) category, there was no difference between the S group and their normal opponents. In the social investigatory category, sniffing (N), the proportion of responses falling into this category for the S group was .10 (rounded to the nearest one hundredth). The proportion for their normal opponents was .14 of the total number of responses observed. The negatively reflected bar was obtained by subtracting .14 from .10. The same procedure was followed in all of the other response categories. Figure 1


- 60 -


























Figure 1. The differences in the distribution of proportions of the saline group (N=18) and that of their normal opponents (N=18) expressed as deflections from the baseline which represents the distribution of the normal Ss. The total number of responses observed was 7,214 (Saline = 3,739; Opponents = 3,575).











Saline Control


, ,UO,


UA HA HO N' C A' B

A
N


P

L


P' X T T' , ,F , ,


V

I , i


.04


.02
1"


- F~~ ~ . __________________________


I I I I I I I I I I j


.02 .04


.06


.04,


-.06


.02


.04


.06-


I


� w


. E .






- 63 -


indicates that the S group showed more evidence of escape behavior (L;W) and submissive posturing (T;T';Z) and less aggressive behavior (A;A';B;C) than their normal opponents. They also showed more evidence of defensive behavior (P;P'), attending (I), and vocalization (V) which is consistent with the response profile of the submissive normal hamster. The differences between the response profiles of the S group and their normal opponents tended to bias comparisons between the chemically treated groups and the S group. However, as will be noted below, the bias was not large enough to cancel those chemically produced effects which were large enough to reasonably be considered to be due to the treatment.

Figure 2 compares the effects of the administration of either carbachol or atropine against the effects of the administration of saline. The response profile of the S group is represented by the baseline. Subjects treated with carbachol (N=4) were observed to explore (E) more than the S animals, although they were lower than the S group in most categories. It may not be reasonable to consider the .01 differences observed in a defense posture (P'), freezing (E), vocalization (V), and grooming (G) as being due to the treatment. Atropine sulphate (N=4) has little effect on any behavior except exploration (E), aggression (A;A';B;C), actual fighting (F) as opposed to sparring (S), and attending (I). It may be useful to the reader to pay special attention to the fact that although the carbachol group showed less submissive behavior than the S group, they did not show more aggressive behavior. In this regard, the atropine Ss showed more aggressive behavior, but did not show less submissive behavior.

























Figure 2. The differences between the distributions of proportions of both the carbachol (N=4) and atropine (N=4) groups and the saline group (N=18) expressed as deflections from the baseline which represents the distribution of the saline group. The total number of responses observed was 5,304 (Carbachol = 751; Atropine = 814; Saline = 3,739).










Carbachol Atropine


.08 06 .04 .02





.02


.04 .06 .08


-L


HA UO


2 3 .08


.06


.04






- 66 -


The lack of an effect of carbachol on freezing (Z) and the pronounced effect of atropine on attending (I) should also be noted.

Figure 3 represents the distributions of proportions of both the norepinephrine group (N=6), and the intraventricular norepinephrine group (N=3) compared with that of the S group. The application of norepinephrine to the septal region produced a marked reduction in the exploratory behavior of the treated Ss when compared with the S group. Intraventricular norepinephrine produced an effect in the same direction but it was one-sixth as large as that produced by norepinephrine in the septal region. Neither location had an effect on hoarding behavior (UA;HA;UO;HO), while both may have been associated with a reduction in social investigatory behavior (N;N'). A slight decrease in aggression (A;A';B;C;) was effected at both sites as was an increase in defensive behavior (P;P'). Fighting (F) and sparring (S) were differentially affected. Norepinephrine placed in the septal region may have slightly increased fighting while intraventricular norepinephrine produced a marked increase in sparring (S) and pinning (X), which do not involve biting as does actual fighting. At both sites, norepinephrine decreased escape behavior (L;W), but norepinephrine increased the use of submissive postures (T;T';Z) only following application to the septal region. Attending (I) was also increased following the application of norepinephrine to the septal region. Vocalization (V) constituted .08 of the response total, and was recorded 89 times during the observation of the group treated with norepinephrine in the septal region.

Figure 4 represents the distribution of proportions obtained

























Figure 3. The differences between the distributions of proportions of both the norepinephrine (N=6) and intraventricular norepinephrine (N=3) groups and that of the saline group (N=18) which is represented by the baseline. The total number of responses observed was 5,431 (Norepinephrine = 1,167; Intraventricular norepinephrine = 525; Saline = 3,739).









Norepinephrine Intraventricular Norepinephrine


N N'


M I-


.08 .06


.04


.02


P P'


.06 .08
.12


06

























Figure 4. The differences between the distributions of proportions of both the phenoxybenzamine (N=9) and intraventricular phenoxybenzamine (N=4) groups and that of the saline group (N=18) which is represented by the baseline. The total number of responses observed was 5,474 (Phenoxybenzamine = 1,735; Intraventricular phenoxybenzamine = 688; Saline = 3,739).








E .11
.04


.06 .04 .02


-


.02


.04


.06


.08


Phenoxybenzamine I Intraventricular Phenoxybenzamine


UAHA =i


N




UO HO
,Im. ,A


B





A I.


I
I I I I I Izu ~


..
C


I IUI I U


T T'
W


-I- I


G
G


P'I
F


x'


.06


-.02


U -


.02 .04 .06 .08


I


FI






- 71 -


following the application of the anticatecholaminergic agent, phenoxybenzamine, to the septal region (N=9) and to the lateral ventricle (N=4) compared with the distribution obtained following the application of saline (N=18) to the septal region. The septal and intraventricular effects of the application of phenoxybenzamine closely parallel each other with the exception of social investigation (N) and the aggressive attack (A). Although few differences are to be found in the direction of the effects caused by phenoxybenzamine at these two sites, there appear to be marked quantitative differences. Intraventricular placements produced a strong increase in exploration (E) and hoarding (UA;HA;UO;HO), but septal placement had only a moderate effect. Septal placements had a strong effect on biting (B) and ventricular placements, little or none. The remaining effects on defense, fighting, escape, submission, and other categories are both qualitatively and quantatively similar.

When the norepinephrine and phenoxybenzamine groups are compared with the S group, as shown in Figure 5, it is clear that these two agents generally produce opposite effects on the behavior of the hamster. Phenoxybenzamine produces an increase in exploration, while norepinephrine produces a considerable reduction in exploration. Norepinephrine may also have an opposite effect on hoarding behavior. Phenoxybenzamine produces an increase in social investigatory behavior, while norepinephrine, if it has an effect, may produce a reduction in this behavior. Aggression scores increase following the administration of phenoxybenzamine while the application of norepinephrine produces a reduction of aggressive behaviors. Defense behav-


























Figure 5. The differences between the distributions of proportions of the norepinephrine (N=6) and phenoxybenzamine (N=9) groups and that of the saline group (N=18) which is represented by the baseline. The total number of responses observed was 6,641 (Norepinephrine = 1,167; Phenoxybenzamine = 1,735; Saline = 3,739).










Norepinephrine ..R Phenoxybenzamine I


UA HA


.08


.06 .04 .02


.02 .04 .06 .08
.12


.02 .04


.12






- 74 -


ior is reduced by phenoxybenzamine, but norepinephrine increases the frequency with which these behaviors are observed. Fighting may be unaffected by phenoxybenzamine and norepinephrine, but there is some suggestion that norepinephrine produces a reduction in fighting. Escape behaviors are reduced by both norepinephrine and phenoxybenzamine. Norepinephrine produces a dramatic increase in submissive postures, attending, and vocalization. Phenoxybenzamine does not appear to have an effect on these behaviors.

Figure 6 represents the results of the fornix-resection group (N=4). The distributions were obtained from trials run with and without the application of phenoxybenzamine to the septal region. Both distributions are compared with that of the phenoxybenzamine group which is represented by the baseline. The direction of each effect is the same for both the fornix resection condition and the fornix resection-phenoxybenzamine condition. The presence or absence of phenoxybenzamine alters only the magnitude of the effect. The presence of phenoxybenzamine reduces the magnitude of the increase in exploration produced by fornix resection; however, either with or without phenoxybenzamine, resection of the fornix produces an increase in exploration which exceeds that seen following the application of phenoxybenzamine to the septal region of the intact animal (see Figure 4). Fornix resection reduces hoarding to zero. Phenoxybenzamine alleviates this deficit slightly. Both groups are lower in terms of their social investigatory behavior (N;N'). Phenoxybenzamine produces an increase in the aggressiveness of the intact animal (see Figure 5), while fornix resection alone produces an increment in aggressiveness which approximates that seen in the phenoxybenzamine treated intact
























Figure 6. The differences between the distributions of proportions of the fornix resection group (N=4) obtained with and without the application of phenoxybenzamine to the septal region and that of the phenoxybenzamine group (N=9) which is represented by the baseline. The total number of responses observed was 2,673 (Fornix resection with septal phenoxybenzamine = 699; Fornix resection alone = 239; Phenoxybenzamine = 1,735). The fornix resection alone condition was run for only three trials.









Fornix Resection Alone Fornix Resection and
Phenoxybenzamine


.06


.04 .02






.02 .04 .06 .08
.09


.02


.08 .09






- 77 -


animal. The treatment of the fornix resection group with phenoxybenzamine produces a further increment in aggression (Figure 6). Phenoxybenzamine partially cancels the depression in defensive behavior seen following fornix resection. Phenoxybenzamine, in combination with fornix resection increased the incidence of fighting more than did fornix resection alone. The remainder of the two response profiles are identical in their deviation from the response profile of the phenoxybenzamine group.

In Figure 7 the distribution of proportions of the amygdalectomyphenoxybenzamine group (N=4) is compared with that of the phenoxybenzamine group. The treated group showed increases in both exploration and hoarding, while all of the social categories were depressed. The most marked depression was found in the social investigatory (N) and fighting (B) categories.

Figure 8 compares the effects of the administration of norepinephrine and the effects of the administration of carbachol with the distribution of proportions of the S group. Norepinephrine reduces exploration, while carbachol produces a large increase in the category. Norepinephrine and carbachol both may somewhat reduce hoarding. Carbachol produces a general reduction in social behavior as may be seen from an examination of the remaining categories in Figure 8. The norepinephrine profile does not show this consistency. There is a marked increase in the use of submissive postures (T;T';Z) and the occurrence of vocalization in the norepinephrine group.

Figure 9 compares the distributions of the atropine and phenoxybenzamine groups with that of the S group. Atropine and phenoxybenzamine produce opposite effects on exploration, but have little effect


























Figure 7. The differences between the distribution of the amygdalectomy-phenoxybenzamine group (N=4) and that of the phenoxybenzamine group (N=9) represented by the baseline. The total number of responses observed was 2,058 (Amygdalectomy-phenoxybenzamine group = 353; Phenoxybenzamine group = 1,735).










Phenoxybenzamine and Amygdalectomy


.02 .02

























Figure 8. The differences between the distributions of proportions of both the carbachol (N=4) and the norepinephrine (N=6) groups and that of the saline group (N=18) represented by the baseline. The total number of responses observed was 5,657 (Carbachol group = 751; Norepinephrine group = 1,167; Saline group = 3,739).









Carbachol Norepinephrine


P F


1.12


HA UO


A B


.04


1.06


.1 2



























Figure 9. The differences between the distributions of proportions of both the atropine (N=4) and phenoxybenzamine (N=9) groups and that of the saline group (N=18) represented by the baseline. The total number of responses observed was 6,288 (Atropine group = 814; Phenoxybenzamine group = 1,735; Saline group = 3,739).












Atropine Phenoxybenzamine


P' Z IV G


1.08


.08


.08 .06


C
.o .04
0 0.
O .2 .02
0

CL
O L .02
0


0
0
c .02

u
c L .04


.06


UAHA







- 84 -


on hoarding. Atropine reduces social investigatory behavior while phenoxybenzamine has an opposite, but equal, effect. Atropine uniformly increases aggressive behavior (A;A';B;C). Although phenoxybenzamine appears to increase biting (B), it appears to have little effect on the remaining aggressive behaviors. The direction of the effects of atropine and phenoxybenzamine is similar throughout the rest of the profile although the magnitude of the effect on sparring

(S) and attending (I) is larger in the atropine group.

Histological results. The results of histological examination of the brains of the Ss in the study showed the cannulae to have been well-placed. The majority of the placements were in the midline and within .75 mm. rostral to the decussation of the anterior commissure. Some variability in the placements was intended, therefore, a precise statement cannot be made regarding some specific point within the septal region at which the tips of the cannulae were found. Moreover, it is doubtful that such a statement would be useful since the extent of the action of the chemicals in unknown. If can be said that if the chemical exerted an effect for a radius of 1.5 mm. around the tips of the cannulae, it is likely that most of the septal region was affected.

Figure 10 shows two sections which were cut at the level at which the cannulae penetrated into the lateral ventricle. It also shows the sections from two Ss which had more rostral septal placements. Figure 11 shows another two rostral placements. Subject-3N had the tip of the cannula placed in the anterior limbic cortex rostral to the septal region, while S-11N had its cannula projected into the midline approximately midway in the rostrocaudal extent of





























Figure 10. Sections from representative Ss showing ventricular and rostral septal placements.
































Figure 11. Representative sections from Ss with rostral and typical septal placements.







S9L-S


NC-S


dlGL-S






- 89 -


the septal region. This placement was slightly anterior in the most typical placements. Subject-15P is a very representative animal. The brain of S-15S is included here as an example of a very extreme placement. The cannula was found to be in the lateral septal nucleus. Figure 12 also shows some of the variability in the sites at which the tips of the cannulae could be found. In S-15A the cannula was found to be slightly off the midline but within the medial septal nucleus. Subject-29N had its cannula placed dorsal to, and at the level of, the decussation of the anterior commissure. Subject-9PA was found to have its cannula in the posterior septal region among the fibers of the columns of the post-commissural fornix. In S-27P the cannula was placed in the most caudal part of the medial septal region. Figure 13 shows two atypical placements. In S-9N the cannula had been directed at the septal region, however, it was found to have been placed caudal to the septal region in a position from which the structures adjoining the third ventricle and the bed nucleus of the stria terminalis could have been affected by the chemical. This S showed a behavior pattern which contrasted sharply with the other Ss in the norepinephrine group. The results from this S are discussed in Appendix A. The remaining atypical S had the cannula placed in the hippocampus. This S was supposed to have been a ventricular animal. Although the cannula was not in the ventricle, there was little basis for distinguishing the data from this S from those of the remaining intraventricular Ss. As no clear behavioral difference was observed, this S was not given separate treatment in Appendix A.

The remaining two Ss in Figure 13 represent the lesion groups.

The fornix lesion was relatively complete. No damage was seen in the





























Figure 12. Representative sections from Ss with more extreme variations in septal placements.





























N6Z?-S


,9 Ab-



























Figure 13. Representative sections from two Ss with placements which were unlike any others in the study and sections from two Ss from the lesion groups illustrating their lesions.






- 93 -


f) N




Full Text

PAGE 1

THE SOCIAL BEHAVIOR AND AGGRESSIVENESS OF THE HAMSTER FOLLOWING THE APPLICATION OF CHEMICALS TO THE SEPTAL REGION OF THE FOREBRAIN By FRANK JACK SODETZ, JR. A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLOKIDA August, 1967

PAGE 3

ACKNOWLEDGMENTS The author wishes to express his appreciation to his supervisory conimittee chairman. Dr. B. N. Bunnell, for his support and encouragement throughout the past four years. A debt of gratitude is owed also to the members of the committee, Drs. D. C. Goodman, J. A. Horel, F. A. King, P. Satz, and W. B. Webb, whose guidance and letters made the completion of this dissertation possible. The untiring efforts of Mr. J. A.Jones and Mr. C. P. Vega in assisting with the collection and processing of the data are most gratefully acknowledged. The services of the University of Florida Computing Center are also acknowledged. The manuscript could not have been completed without the assistance of Mrs. M. A. Harrington and Miss T. R. Karell who looked after the illustrations and photography and Mrs. E. Marable and Miss B. Watson who typed the manuscript. il -

PAGE 4

TABLE OF CONTENTS ACKNOWLEDGMENTS page ii LIST OF TABLES V LIST OF FIGURES vl INTRODUCTION 1 General Background The problem Anatomical considerations The Limbic System: Structure The limbic system in submammalian forms Amphioxus Cyclostomes Chondrichthyes Teleosts Amphibians Reptiles Summary The mammalian limbic system Anatomical data The septal region Electrical stimulation studies Chemical mapping of the limbic system Definition of the mammalian limbic system The Limbic System: Function The septal region Emotion Avoidance learning Other learning paradigms Feeding, drinking, and sleep Related structures Emotion Learning Some relevant pharmacological studies The limbic system and species-specific behavior METHOD 48 Subjects Apparatus Response inventory Surgery iii -

PAGE 5

Procedure Histology Subjects dropped from the study Saline controls Norepinephrine group Intraventricular norepinephrine group Phenoxybenz amine group Intraventricular phenoxybenzamine group Phenoxybenzamine fornix resection group Phenoxybenzamine amygdalectomy group Carbachol group Atropine group Summary of groups Statistical analysis RESULTS DISCUSSION j^QQ SUMMARY -^-j^g REFERENCES 2.21 APPENDIX A APPENDIX B -^29 BIOGRAPHICAL SKETCH 240 iv -

PAGE 6

LIST OF TABLES Table Page 1. A brief description of the responses which can be seen in the interaction of two adult male hamsters 50 2. A summary of the groups used in the study 57 V -

PAGE 7

LIST OF FIGURES Figure Page 1. The differences in the distribution of proportions of the saline group (N=18) and that of their normal opponents (N=18) expressed as deflections from the baseline which represents the distribution of the normal S^s. The total number of responses observed was 7 , 214 (Saline = 3,739; Opponents = 3,575) 62 2. The differences between the distributions of proportions of both the carbachol (N=4) and atropine (N=4) groups and the saline group (N=18) expressed as deflections from the baseline which represents the distribution of the saline group. The total number of responses observed was 5,304 (Carbachol = 751; Atropine = 814; Saline = 3,739) 65 3. The differences between the distributions of proportions of both the norepinephrine (N=6) and intraventricular norepinephrine (N=3) groups and that of the saline group (N=18) which is represented by the baseline. The total number of responses observed was 5,431 (Norepinephrine = 1,167; Intraventricular norepinephrine = 525; Saline = 3,739) 68 4. The differences between the distributions of proportions of both the phenoxybenzamine (N=9) and intraventricular phenoxybenzamine (N=4) groups and that of the saline group (N=18) which is represented by the baseline. The total number of responses observed was 5,474 (Phenoxybenzamine = 1,735; Intraventricular phenoxybenzamine = 688; Saline = 3,739) 70 5. The differences between the distributions of proportions of the norepinephrine (N=6) and phenoxybenzamine (N=9) groups and that of the saline group (N=18) which is represented by the baseline. The total number of responses observed was 6,641 (Norepinephrine = 1,167; Phenoxybenzamine 1,735; Saline = 3,739) 73 6. The differences between the distributions of proportions of the fornix resection group (N=4) obtained with and without the application of phenoxybenzamine to the septal region and that of the phenoxybenzamine group (N=9) which is represented by the baseline. The total number of responses observed was 2,673 (Fornix resection with septal phenoxybenzamine = 699; Fornix resection alone = 239; Phenoxybenzamine = 1,735). The fornix resection alone condition was run for only three trials 76 vi -

PAGE 8

7. The differences between the distribution of the amygdalectomy-phenoxybenzamine group (N=4) and that of the phenoxybenzamine group (N=9) represented by the baseline, the total number of responses observed was 2,058 (Amygdalectomy-phenoxybenzamine group = 353; Phenoxybenzamine group = 1,735) 79 8. The differences between the distributions of proportions of both the carbachol (N=4) and the norepinephrine (N=6) groups and that of the saline group (N=18) represented by the baseline. The total number of responses observed was 5,657. (Carbachol group = 751; Norepinephrine group = 1,167; Saline group = 3,739X 81 9. The differences between the distributions of proportions of both the atropine (N=4) and phenoxybenzamine (N=9) groups and that of the saline group (N=18) represented by the baseline. The total number of responses observed was 6,288 (Atropine group = 814; Phenoxybenzamine group = 1,735; Saline group = 3,739) 83 10. Sections from representative Ss showing ventricular and rostral septal placements 86 11. Representative sections from ^s with rostral and typical septal placements 88 12. Representative sections from S^s with more extreme variations in septal placements 91 13. Representative sections from two S^s with placements which were unlike any others in the study and sections from two S^s from the lesion groups illustrating their lesions 93 14. A summary of the effects obtained in the study. All groups are compared with the saline group. Differences of less than 34% were ignored. Iji considering the data presented in this figure, it should be noted that the mean number of responses observed for each in each group was as follows: Saline, 199; Carb., 188; NE Sept., 194; NE Vent., 175; Phen. Sept., 193; Phen. Vent., 172; Fornix, 175; Amyg., 78 96 — vii -

PAGE 9

INTRODUCTION General Background The problem The present study was designed to further evaluate the role of the septal region of the foreb rain in the aggressiveness and social behavior of the hamster. The need for further evaluation arose out of an attempt to cope with a problem in one of the first studies dealing directly with the role of the septal region in these speciesspecific behaviors. The study (Bunnell, Bemporad, and Flesher, 1966) demonstrated that septal ablation increased the social rank of hooded rats. However, social rank was defined on the basis of the effectiveness of the animal in a competitive situation and, while this is not uncommon in such studies, the artificiality of the testing situation may have favored the septal animals. Pelligrino (1967) has since verified this work using a somewhat different competitive test. Bunnell and Smith (1966) studied the social behavior of cotton rats following septal ablation in a more natural setting and found both increased aggressive and escape behavior in these animals. The social behavior of the cotton rat was so disrupted following septal ablation that a number of the animals were killed by their cagemates. At this point, there seemed to be a clear need to study the effect of septal ablation on species-specific behavior, using a technique which would permit explicit statements about resulting behavioral changes and yet would

PAGE 10

~ 2 not have the undesirable attributes of an artificial testing situation. The selection of an animal was critical in this regard. The hamster seemed ideally suited for this purpose in that it engages in social interaction without special training or inducement from the experimenter. Furthermore, the hamster uses a discrete number of readily identifiable behaviors in this interaction. A number of studies (Sodetz, 1965; Bunnell, Sodetz, and Shallaway, 1965; Sodetz and Bunnell, 1967a; Sodetz and Bunnell, 1967b) have now shown the septal region of the hamster to be involved in the regulation of such species-specific behaviors as hoarding, nest building, and aggressive interaction. In avoiding the artificial testing situation, a problem was created. These studies sampled a broad range of behaviors. Some of these behaviors diminished in their frequency while others increased in frequency. Previous social experience appeared to determine whether septal ablation would make the hamster more aggressive or more submissive. It also appeared that the behavior of the opponent of the septal animal was of diminished importance in determining the behavior of the septal. Bunnell, et al. (1965) have suggested that a uniprocess model may not be able to account for the changes in species-specific behavior resulting from septal ablation. Such a model would also have to account for the appearance of "septal rage" in some species (Brady and Nauta, 1953; King, 1959; Brown and Slotnick, 1966) and not in others (Sodetz, Matalka, and Bunnell, 1967). Changes in water intake which follow ablation (Harvey and Hunt, 1965) and stimulation (Fisher and Coury, 1962; Grossman, 1964b) of the septal region would have to be considered in such a model, as would learning deficits (King, 1958; McCleary, 1961; Ellen, Wilson, and Powell, 1964),

PAGE 11

3 The variety of the behavioral changes seen following ablation and stimulation of the septal region suggested that it would be profitable to investigate the possibility that more than one deficit might be responsible for the results obtained in the studies of species-specific behavior. However, although the septal region consists of a number of anatomically distinct nuclei (Andy and Stephan, 1961) which differ in their connections to other regions of the brain, little evidence has been found for a functional differentiation based upon anatomically distinct regions (Harrison and Lyon, 1957). Burkett and Bunnell (1966) did find evidence that the medial septal region may be responsible for the DRL deficit seen in septal rats (Ellen, Wilson, and Powell, 1964). While a differentiation in function based upon anatomical data has been difficult to confirm, the work of Hernandez Peon, Ibarra, Morgane, and Tlmo-Iarla (1963) which demonstrated cholinergic sites for sleep and rage throughout the septal region and Grossman's (1964a) success \ in eliciting both eating and drinking from the same anatomical locus in the amygdala with two different neurotransmitter substances, sug^ gested that damage to chemically distinct subsystems might be responsible for the deficits seen following ablation of the septal region. / More specifically, the present study was designed to test the possibility that a multiplicity of functions might underlie the changes in species-specific behavior which result from septal ablation. The technique of selective chemical stimulation and blocking of chemically distinct systems through Indwelling cannulae seemed to offer the best hope In this regard. Anatomical considerations The septal region has traditionally been considered to be part of

PAGE 12

4 the limbic system. One would expect its functions to be intimately related to those of the other structures in this system. However, one has difficulty finding a consensus as to what these structures are and, therefore, inferences about septal function based upon the function of related structures are difficult to make. Discussions of the limbic system are numerous in the recent literature. Each paper presents an anatomical definition which differs in some aspect, usually according to the function being emphasized by the author, from definitions outlined in similar papers. For Broca (1878) the "grande lobe limbic" was a prosencephalic system common to all mammals. Papez (1937) found it to be an essentially "closed" system lying within the prosencephalon but he tended to emphasize its diencephalic components and proposed that such a system might offer the reverberating circuitry thought necessary to the maintenance of emotional apprehension. The work of Bard (1939) and Bard and Mountcastle (1948) provided data to partially support this view of the limbic system. At the present time, the limbic system of Pribram and Kruger (1954) is among the best known and most widely accepted of a number of limbic system definitions. This acceptance, however, does not extend throughout the community of those involved in limbic system research. Within this group it is recognized that a wealth of close relationships exists between this limbic system and other parts of the brain. The decision to include some or all of these other structures and connections has been based on behavioral data suggesting very close functional relations between structures included in the Pribram and Kruger definition and neocortical, reticular, hypothalamic and extra-pyramidal structures. In opposition to this extension of the limbic system has been the easily

PAGE 13

5 justifiable argument that one can readily connect any structure within the central nervous system to any other structure by a route which would Include very few synaptic relays.. By this reasoning, the limbic system could readily be extended to include most of the nervous system. While most arguments both for and against the inclusiveness and exclusiveness of the limbic system have some merit, it should be noted that the limbic system referred to above does not derive its integrity from a common histogenetic origin, nor has a consensus been found regarding some unitary functional designation which can be applied to this system. It would appear that adherence to the concept of a limbic system without neocortical or mesencephalic components is propagated as much for historical reasons as any others. The limbic system has been shown to have numerous complex connections with the mesencephalon (e.g., Nauta and Kuypers, 1958) and is only one relay removed from the efferent nuclei of the medulla. In the rostral cerebrum, the f rontogranular cortex has been shown to have direct connections with the limbic system (e.g., Nauta, 1964). Brady (1958a) in attempting to anatomically delimit the limbic system expresses his concern, and that of many others, for the adequacy of any such attempt. What seems to be required is some anatomical and functional redefinition of the limbic system if one is to adequately characterize the function of one of its prominent parts. What follows below is an attempt to review some of the data from which such a definition may someday emerge. The Limbic System; Structure The limbic system in submammalian forms Amphioxus . Elements of a very primitive limbic system can be

PAGE 14

6 traced back through phylogeny below vertebrates. The chordate Amphioxus . which looks much like a fat worm and lacks eyes, has a very primitive forebrain. A telencephalon and diencephalon cannot be discriminated in this species. Amphioxus does have a single, unpaired olfactory nerve which enters the forebrain and distributes to the homologue of the septal region. The course of fibers originating in the septal region is not known. 1 Cycles tomes . In primitive vertebrates an olfactory related limbic apparatus is clearly present. In cyclostomes, the telencephalic portion of this g^stem occupies the mediobasal portion of the hemispheres and consists of the septum. The septum is a major component of the archistriatum from which some of the other components of the limbic system will arise in higher forms. The dorsal convexity of the cerebrxim is formed by the archipallium, from which additional components of the limbic system later differentiate. The dorso-medial archipallium is generally considered to be primordial hippocampus. The preoptic nucleus also appears to be present in this form. The epithalamus, consisting mainly of the habenular nuclei, appears to receive much of the output of these centers via its primary afferent pathway, the stria medullaris, and other olf acto-habenular bundles. The olfactory areas also project to most of the hypothalamus. The epithalamus discharges via the fasciculus retroflexus to the nucleus interpeduncularis . From the nucleus interpeduncularis , which is of considerable size in lower forms, fibers distribute through the medulla '"Much of the phyletic data on the limbic system was obtained from: Ariens Kappers, C.V., Huber, G.C. and Crosby, E.C. The Comparative Anatomy of the Nervous System of Vertebrates including Man . Hafner Publishing Co., 1960, III Volumes.

PAGE 15

7 oblongata to efferent centers. A hypothalamic bulbar fiber tract courses along the extent of the interpeduncular nuclei and also distributes in the medulla. Notable in the discussion of the primitive limbic structures of these forms is the absence of the reciprocal interconnections seen in higher vertebrates. In cyclostomes limbic pathways are for the most part directed rostrocaudally . One is also impressed by the sparse limbic-hypothalamic interrelationship even though both serve the same medullary efferent centers. This relative simplicity of limbic organization is suggestive of a rigid translation of olfactory input into effector activity. Chondrichthyes . The brain of the shark shows an increased complexity in olfactory connections and it is at this level in phylogeny that it appears possible to suggest that olfactory-related telencephalic structures may take on other functions. In chondrichthyes there are found a number of commissural limbic connections. In addition, the hypothalamus in these animals sends fibers to the habenulae and receives fibers from both the septal and primordial hlppocampal regions. The existence of such connections suggests that considerable interaction may take place between limbic system and hypothalamus. This may be in contrast to the rather direct olf acto-ef f ector relations seen in cyclostomes. Another correlation appears in the telencephalon of chondrichthyes. The dorsal thalamus sends fibers, presumably somato-sensory in origin, to the very small non-olfactory portions of the archistriatum which in turn have connections to the lateral olfactory area or paleopallium. Both of these areas will eventually contribute to the mammalian amygdala, the basolateral components being paleopallial in

PAGE 16

8 origin and the corticomedial arising from archistriatum. The globus pallidus and pyriform areas also have their origin in these regions. More importantly, what is seen in these forms is an opportunity for the interaction of olfactory and other senses at the telencephalic level. Teleosts . In more modern fishes the medial olfactory region, much of which will be the septal and preoptic regions in mammals, consists of several identifiable nuclei. These include the nucleus preconnnissuralis superior and inferior and the nucleus olfactorius anterior pars precommissuralis. Inferior to the caudal border of the decussation of the anterior commissure lies the preoptic nucleus. A large area of the telencephalon may be designated primordial hippocampus. The olfactosomatic paleopallial region noted in the chondrichthyes is larger in the teleosts. Perhaps the most notable development in relationship to the limbic system is a separation of the hypothalamus from primary olfactory fibers and enrichment of the connections between precommissural septal structures and the hypothalamus. By way of the medial forebrain bundle, both afferent and efferent interaction are possible between hypothalamic and limbic regions, including the primordial hippocampus. Another precommissural septal-hypothalamic connection runs ventrocaudally from the septal region and probably represents the precommissural septal contribution to the columns of the fornix in higher forms. The epithalamic-limbic relationship may be of diminished significance in these forms for the habenulae are reduced in size and far fewer fibers distribute to this center than to the hypothalamus. Amphibians . In the amphibians, all of the structures commonly considered to be part of the limbic system are seen, with the exception of the cingulate cortex which is neopallial in origin. It is in these

PAGE 17

9 forms that the archlstriatal cortico-medial amygdala first appears. A primordial pyriform lobe can also be identified as lying adjacent to the amydaloid region. The septal region may be divided into medial and lateral septal nuclei. In amphibians, the hippocampus remains primordial. The limbic system of amphibians evidences more definite differentiation of nuclei and a greater interrelatedness with nonolfactory pallial and striatal structures. The interrelationships both within the system and with other systems are further character— ized by the reciprocity which is seen in higher forms . Reptiles . The limbic system of reptilian forms is more differentiated in its nuclear structure than is that of the amphibian. Most of the limbic structures seen in mammals are present and well differentiated. This differentiation extends to the hippocampus, which appears to have the character of a true cortical structure. The reptilian striatal regions are also more highly differentiated. A small region of general cortex can be found in the pallium of the reptile and is the forerunner of the general neocortex of mammals. Summary . The avian forebrain develops in a line independent of mammalian forms and it is sufficient to note that the limbic system of the bird is similar, though more complex, to that of reptilians. Throughout the vertebrate phyla the limbic system has undergone development from a rather direct olf acto-ef f ector system, with interaction from other systems arising primarily in the regions nearest effector sites, to an intermediate stage at which interaction with other areas arose in the more rostral regions of the brain. This intermediate stage foreshadowed the development of more ninnerous ascending influences from other systems to the telencephalic limbic

PAGE 18

components. In mammals the limbic system has emerged with massive intrinsic and extrinsic connections and apparently little solitary control over any effector system. While it is true that increasing complexity and interrelatedness is a general phenonemon applicable to the entire nervous system in its ascent through phylogeny, the limbic system differs from other systems which are sensory in origin. While other sensory systems increase their complexity through differentiation and the proliferation of connections with other systems, they do not lose their identity as sensory systems. In the visual system, for example, the occipital neocortex is easily identifiable as a component in a system related to the processing of visual stimuli. The visual neocortex need not be traced back to its origin in phylogeny to locate some ancient link to the optic tectum or thalamus to support its inclusion in the visual system. However, this is not the case with the limbic system, for in lower forms it is clearly sensory-related. This relationship gradually dissipates as one ascends through phylogeny. Rather than maintaining a close association with its olfactory origin, the bulk of the limbic system retreats from its sensory status. The development of this system does not parallel that of other sensory systems in the enhancement of the capacity to elaborate stimuli but reflects the increasing importance, through phylogeny, of some other function or functions. These functions are most certainly related to both sensory and motor systems, but, in a strict sense, are probably neither. An appealing hypothesis presents itself in the notion that the limbic system, which somehow must deal with both sensory and motor systems, has undergone its development in response to the need to deal with a sensorium capable of processing increasingly complex sensory

PAGE 19

11 data and an effector system capable of an enormous versatility of response. In the very lowest vertebrates, the cyclostomes, the source of a stimulus, almost certainly olfactory, is either affixed to and slowly devoured or ignored or avoided. In mammals countless choices exist for countless combinations of stimulus inputs. A much more complex limbic system may be necessary to cope with this wealth of data. The mammalian limbic system Anatomical data . No attempt will be made in this section to consider in great detail the connections of various structures to be included in the limbic system. Each structure could, in itself, be the basis for a number of papers. Details of the anatomy of the limbic system and related structures can be found in papers by Andy and Stephan (1961), Cowan, Raisman, and Powell (1964; 1965), Gloor (1960), Green (1964), Nauta (1956; 1958; 1961; 1964), Nauta and Kuypers (1956), Papez (1937; 1958), Powell (1963; 1964), Pribram (1960), Pribram and Kruger (1954), Valenstein and Nauta (1959), Raisman (1966), Raisman, Cowan, and Powell (1966), and Powell, Cowan, and Raisman (1965). Perhaps the most profitable approach would be to consider first those structures commonly considered to be part of the limbic system. There is one exception to this, that is, the first system of Pribram and Kruger (1954). The structures included in this system all receive primary olfactory connections. There would appear to be little reason for continuing to consider these areas as part of the limbic system. The septal region . The septal region will be considered first since it is the primary concern of this paper. The septal region itself can be divided into four nuclear groups (Andy and Stephan, 1961);

PAGE 20

12 however, it is more common to discuss it in terms of a medial and lateral septal region. The lateral and medial septal nuclei differ with respect to their af ferents and ef f erents. The medial septal nuclei send fibers to the hippocampus and receive fibers from the hypothalamus. The lateral septal nuclei receive fibers from the hippocampus and send fibers to the lateral hypothalamus. The septal region projects tOj and receives fibers from, the dorsomedial thalamic nucleus (Valenstein and Nauta, 1959) which in turn supplies fibers to the frontogranular cortex (Akert, 1964; Nauta, 1964). Other thalamic projections of the septal area include those to the intralaminar nuclei and the anteroventral thalamic group. These connections are established via the mammillary bodies, although all fibers do not synapse there. Although it is not clear whether they arise in the septal region, the hippocampus, or both, fibers are found which project to the nucleus lateralis dorsalis in primates (Valenstein and Nauta, 1959) . Such fibers are barely detectable in lower forms and suggest a relationship to parietal association cortex. Nauta (1953) does report, however, that the dorsolateral nucleus receives a massive projection from the caudal cingulate gyrus in species below primates. An epithalamic projection from the medial septal region to the habenula is found in all vertebrates. From the habenula fibers project to the nucleus interpeduncularis via the fasciculus retroflexus. From this nucleus fibers arise which project to the visceral efferent nuclei of the medulla. Nauta (1956; 1958) provides an excellent discussion of septalhypothalamic relations via the fornix, while Andy and Stephan (1961)

PAGE 21

13 emphasize the medial forebrain bundle. The connections of the septal region to the mammillary bodies have already been noted. The septal region sends fibers via the diagonal band to the lateral preoptic area. From here, fibers enter the diencephalon and distribute throughout the extent of the lateral hypothalamic area. While septal connections to the lateral hypothalamus are the most prominent, fibers from the columns of the fornix and fibers coursing through the medial preoptic region supply connections to more medial hypothalamic areas. These connections vary appreciably from species to species. The amygdala, which is directly related to the septal region via both the diagonal band and the stria terminalis also contributes to the medial forebrain bundle following relay in the lateral preoptic region. The amygdala, however, also has direct connections to the lateral hypothalamus via ventrofugal fibers described by Nauta (1961). The hippocampus, which receives afferents from the medial septal region and projects efferents to the lateral septal region, also contribute fibers to these hypothalamic areas which are served by the septal region. The medial forebrain bundle also carries fibers from the f rontogranular cortex. Fibers from all of the above sources distribute in the mesencephalic tegmentum (Nauta, 1964; Nauta and Kuypers, 1958). These fibers generally distribute to the dorsomedial tegmentum at the lateral border of the central gray. The septal region is connected to the cingulate gyrus via the anteroventral thalamic nucleus, while the cingulate gyrus projects to the supracommissural septal region via fibers which perforate the corpus callosum and enter the fornix dorsalis (Cragg and Hamlyn, 1959; White, 1965).

PAGE 22

14 Other notable connections are those of the hippocampus to entorhinal cortex and from there to the mesencephalic reticular system, and the frontogranular cortex to the periamydaloid cortex and presumably the amygdala. The number of interrelationships among these structures would permit the assembly of a number of different systems and favor no one system over any other. The functional data to be considered below may provide some useful basis for grouping these structures into systems. Electrical Stimulation Studies . Maps of limbic and related structures have been compiled on the basis of similiarities in obtained results (Delgado, 1964; DeMolina and Hunsperger, 1962; Hess, 1957; Maclean, 1963; Olds, 1956; 1958). These maps have generally linked the septal region, amygdaloid complex, preoptic area, anteromedial hypothalamus, lateral hypothalamus, and mesencephalic tegmentum in a system generally related to both reward and, in a broad sense, affective display. Kaada (1951) and others (e.g., Feindel and Gloor, 1954) have shown that stimulation of limbic areas leads to activation of other limbic sites and cortical association areas. This activation differs from reticular activation in that it does not extend throughout the entire cortical mantle. Gloor (1955) has demonstrated that amygdaloid stimulation produces changes in the activity of the mediodorsal, centromedian, and intralaminar nuclei of the thalamus. This is also true of stimulation of the clngulate gyrus (see Pribram, 1961). Activity can also be noted in the sub thalamus and mesencephalon (French, Hernandez Peon, and Livingston, 1955; Gloor, 1955). These stimulation studies correlate very well with anatomical data suggesting a close

PAGE 23

15 relationship between what is usually considered to be the limbic system and the f rontogranular cortex, hypothalamus, anterior and intralaminar thalamus, and mesencephalic reticular formation. In reviewing the stimulation data, Pribram (1961) has emphasized the lack of topographical specificity in the effects obtained in limbic stimulation studies. The lack of intralimbic topographical specificity aid the extensiveness of the stimulation effects, however, should not be taken to mean that the limbic system subserves some general arousal or suppressive function. While such a position has some theoretical appeal (Routtenberg , 1966) and electrical stumulation studies have induced sleep from the septal region (Rosvold and Del— gado, 1956) and adjacent areas (Sterman and Clemente, 1962) and arousal from the intralaminar nuclei (Akert, 1961), the specificity of the effects of limbic stimulation militate against any functional interpretation based upon a general arousal or inhibitory concept. The work related to the hippocampal theta rhythm further supports this view. Hippocampal theta is relatively slow, synchronous activity seen clearly in the rabbit, but which is less obvious in the monkey (Green and Arduini, 1954). Petsche, Stumpf, and Gogolak (1962) have shown that a group of cells in the post-commissural septal region act as pacemakers in producing hippocampal theta. Following the destruction of these cells, it is not possible to produce theta either by chemical or electrical stimulation (Stumpf, 1965); however, stimulation of the reticular formation will still elicit normal low voltage fast activity in the hippocampus (Mayer and Stumpf, 1958). These data suggest that general arousal is still possible, even within the limbic system, following damage to its components.

PAGE 24

16 Studies of the effects of electrical stimulation of the limbic system have provided support for the anatomical data, suggesting a close relationship between limbic structures. These same studies strongly suggest that the intralaminar thalamic nuclei and the region of the dorsomedial midbrain tegmentum adjoining the periaqueductal gray be included in a functional, if not anatomical, limbic system. Further support for this position can be found in neurochemical studies of the central nervous system. Chemical mapping of the limbic system . The f luorohistochemical technique of Falck, Hellarp, Thilma, and Torp (1962) has made possible the easy identification of catecholamines in the central nervous system. Of these, norepinephrine, serotonin (5-Hydroxytryptamine) , and dopamine are generally considered to be related to synaptic transmission. As yet, no such technique exists for the identification of acetylcholine, the other widely accepted transmitter substance. Acetycholine concentrations are not measured directly as are those of the catecholamines, but are generally inferred from variations in regional concentrations of specific acetylcholinesterases. The inherent dangers of inferring the level of acetylcholine from measures of the enzyme required to destroy it are well known; however, when acetylcholine has been measured direcly, there has been good correspondence between the direct measures and the estimates based upon specific acetylcholinesterases (Keolle, 1954). The appeal of the direct observation of catecholamines has produced a considerable number of studies related to catecholaminergic systems in the brain^, whereas studies dealing with cholinergic systems, as opposed to regional concentrations, are seen less often.

PAGE 25

17 Techniques for the fluorometric differentiation of the catecholamines have only recently been developed (Glowinski and Baldessarini, 1966); therefore, one has difficulty in distinguishing between norepinephrine and serotonin concentrations in the limbic system. Just as these two substances have been difficult to differentiate fluorometrically, studies using various anticatecholaminergic substances are plagued by the fact that these chemicals lack specificity in their effects and can generally be considered to interfere with both serotonin and norepinephrine. The distribution of catecholamines in the central nervous system and more specifically, the limbic system, is remarkably similar in species ranging from reptiles (Quay and Wilhoft, 1964) to man (Bertler, 1961). Bertler and Rosengren (1959) have found that, in the six mammalian forms they studied, dopamine concentrations were high in areas where the concentration of norepinephrine was low and norepinephrine concentrations were high in areas where the dopamine levels were low. This may suggest a functional dissociation between these two catecholamines, for serotonin and norepinephrine are often found in high concentration within the same regions of the brain. A number of studies have demonstrated a limbic-hypothalamicmidbrain catecholaminergic system (Anden, Dahlstrom, Fuxe, Larsson, Olson, and Ungerstedt, 1966; Aghajanian and Bloom, 1967; Bodganski, Weissbach, and Udenfriend, 1957; Dahlstrom, Fuxe, Olson, and Ungerstedt, 1962; Harvey, Heller, and Moore, 1962; 1963; Heller and Moore, 1965; Heller, Seiden, Porcher, and Moore, 1966; Hillarp, Fuxe, and Dahlstrom, 1966; Moore and Heller, 1967; Moore, Wong, and Heller, 1965;

PAGE 26

18 Paasonen, MacLean, and Giarman, 1957; Reivitch and Glowinski, 1966). While these are just a few of many such studies, they are representative in their agreement on the catecholamine content of the limbic system and related structures. They agree that a high concentration of catecholamines is to be found in the limbic midbrain region described by Nauta and Kuypers (1958). This system appears to be both ascending and descending in the medial forebrain bundle and lateral hypothalamus to the preoptic and septal regions. Beyond this, little is known of the origins and connections of this important limbic circuit for an as yet unexplainable phenomenon occurs with lesions of the medial forebrain bundle. Heller, et al. (e.g. 1966) and others in their group, have studied changes in regional concentration of catecholamines following a variety of subcortical lesions. They have demonstrated that transection of the medial forebrain bundle produces a rapid drop in catecholamine levels in the limbic septal and midbrain regions. However, coincident with this rapid drop, is a slow, relatively steady, decrease in norepinephrine levels throughout the forebrain (Moore and Heller, 1965). This effect is obtained in structures many synapses removed from the septal-medial forebrain bundle-midbrain circuit interdicted by the lesion. No explanation has been found for this transs3maptic depletion of norepinephrine. It is of interest to note that in the rat the depletion continues for approximately fourteen days. Septal ablation also produces a depletion in norepinephrine. The results of a study by Pirch and Norton (1965) are of special interest if considered with the results of the work reported above, even though they are not quite appropriate to this

PAGE 27

19 section. Pirch and Norton (1965) treated septal rats, in which all evidence of hyperemotionality had disappeared, with Beta-Phenylisopropylhydrazine which produces a temporary elevation in brain norepinephrine levels. They observed a reinstatement of septal hyperemotionality which was delayed for about five hours and which gradually diminished during the ensuing seven hours. They followed the change in brain norepinephrine levels during this period. Their results indicate a gradual increment in norepinephrine levels during the first five hours, following the administration of Beta-Phenylisopropylhydrazine followed by a falling phase which paralleled the dissipation of hyperemotionality. The observed septal rage was maximal at the initiation of the falling phase of the norepinephrine concentration and it had disappeared when the concentration reached the chronic low level characteristic of the septal rat. While these authors do not specifically relate their work to the observation of Moore and Heller (1967) of the fourteen-day falling phase of norepinephrine levels produced by lesions in the catecholamine component of the limbic system of the rat, their data, together with those on transsynaptic catecholamine depletion, hare implications for the understanding of the septal rage phenomenon and possibly other transitory and delayed deficits. These changes could represent the gradual transsynaptic depletion of substances important to the normal functioning of the nervous system. Electrical stimulation of the amygdala has also been shown to deplete all catecholamines, with the exception of dopamine, throughout the forebrain (Fuxe and Gunne, 1964); however, this depletion does not extend to the limbic mesencephalic region. Hypothalamically induced defense reactions also deplete forebrain catecholamines, excluding dopamine (Gunne and Lewander, 1966).

PAGE 28

20 The hippocampus, unlike subcortical limbic structures, is relatively low in its catecholamine concentration. It does, however, contain more catecholamines than do some non-limbic regions other than the hypothalamus. Cholinergic mechanisms have been implicated in hippocampal function, both through the application of cholinergic substances (MacLean, 1957; Grossman and Mountford, 1964) and through the studies of the hippocampal theta system. Petsche, et al. (1962) and Stumpf (1965) have demonstrated striking changes in hippocampal activity following treatment with the anticholinergic agent, scopalamine. Fibers from the hippocampus can be followed in the stria medullaris to the lateral habenular nucleus. This nucleus, the tractus retroflexus and the interpeduncular nuclei have been shown to contain very large quantities of specific acetylcholinesterases (Koelle, 1954). The medial habenula, which receives no fibers from the hippocampus (Nauta, 1956), but does receive a strong septal projection, shows little evidence of specific acetylcholinesterases. The amygdala appears to have a high concentration of both cholinergic (Koelle, 1954) and catecholaminergic (see Paasonen, et al., 1957) substances. As one migh-t expect, the medial forebrain bundle contains a good deal of acetylcholinesterase as do parts of the mesencephalic reticular formation. Svmmiary . If the limbic system is defined to include portions of the midbrain tegmentum., it would appear that two systems course throughout its entire extent. It can be tentatively suggested that the limbic pallial-midbrain system may be primarily cholinergic in its activity, while the striatal-midbrain system is catecholaminergic. While this is most certainly an over-simplification, it may offer a profitable way of organizing some of the structures in the limbic system. The limbic

PAGE 29

21 thalamic nuclei vary in their specific cholinesterase content. Those nuclei related to the pallial components of the limbic system, e.g., hippocampus and cingulate, appear to contain large amounts of specific acetylcholinesterase. These are the anteroventral and lateral nuclei. The dorsomedial nucleus contains only a moderate amount of acetylcholinesterase and is closely related to the striatal septal region. There is evidence that the intralaminar nuclei of the thalamus are at least partially served by a cholinergic mechanism (Grossman and Peters, 1966; Grossman, Peters, Friedman, and Wilier, 1965). Chemically distinct systems which somehow underlie sleep (Hernandez Peon, et. al., 1963; Vellute and Hernandez Peon, 1963), arousal (Anderson and Curtis, 1964; Cordeau, 1962; Stumpf, 1965), and feeding and drinking (Epstein, 1960; Fisher and Coury, 1962; Grossman, 1960; Grossman, 1964a; 1964b) have been identified through the local application of cholinergic and catecholaminergic substances. These systems closely parallel similar systems identified through anatomical, ablation, and electrical stimulation studies. While maps based upon chemical studies do not in themselves provide the answer to the problem of adequately characterizing the structure and function of the limbic system, they provide direction for the systematist who might otherwise be lost in a maze of interconnected morphologically indistinguishable fibers and provide links between seemingly unconnected behavioral deficits. Definition of the mammalian limbic system . Broca's choice of the word "limbic" as a name for the structures surrounding the rostral brainstem was fortunate in that the term was without functional connotation (MacLean, 1958). The fact that a group of structures in the mammalian brain are in proximity to one another is not sufficient reason for

PAGE 30

22 assuming a close functional relationship between them any more than their separation would justify the assumption that they were not functionally related. The limbic system should not be defined in terms of the spatial relationship of its component structures, but rather in terms of common functional characteristics. Definite functional similarities between the frontogranular cortex, the limbic system, corpus striatum, and the mesencephalic tegmentum have been noted repeatedly, yet there is reluctance to extend the functional limbic system to include these areas and appropriate thalamic structures. Unfortunately, this reluctance is justified in that so far we have been unable to adequately conceptualize these functions. A more adequate characterization of brain functions is necessary before their underlying anatomical substrates can be defined. However, this may be accomplished sooner if emphasis is shifted from the morphological limbic system and is directed at the similarities and differences in the characteristics of the functional limbic system which apparently has telencephalic , diencephalic, and mesencephalic components . The mammalian limbic system can be defined to include the frontogranular cortex as its most recent telencephalic addition. Its most caudal extent can be found in the dorsomedial mesencephalic tegmentum at the rostral border of the pons. A notable exclusion from the system should be the first system of Pribram and Kruger (1954), that is, the olfactory system. The telencephalic limbic regions should include the frontogranular cortex in species having such a structure, the cingulate gyrus, entorhinal cortex, periamygdaloid temporal cortex, and Ammon's formation at the cortical levels. The remaining telencephalic structures include

PAGE 31

23 the septal region, basolateral amygdalae, globus pallldus, tail of the caudate, and preoptic areas at the diencephalic level, the habenulae, dorsomedial thalamic nuclei, portions of the dorsolateral thalamic nuclei, the anteroventral thalamic nuclei, and the intralaminar group. The mammillary bodies and portions of the subthalamus may also be included in this group. Tlie mesencephalic component lies in the dorsomedial tegmentum at the lateral border of the periaqueductal gray. The Limbic System Function The septal region Emotion . In 1929, Fulton and Ingrahamn described angry behavior in dogs following prechiasmal lesions. Spiegel, Miller, and Oppenhelmer (1940) reported rage following septal ablation in the cat, Brady and Nauta (1953; 1955) found that rats became hyper emotional following septal ablation; however, this deficit was found to dissipate spontaneously in about two weeks. King and Meyer (1958) found that amygdaloid ablation could attenuate this deficit. King (1959) also found genetic differences to be a factor in determining its magnitude. Harrison and Lyon (1957) attempted to find a correlation between damage to specific components of the fornix and septal region and the appearance of septal rage. They could find no correlation. Teitelbaum (1964) found that cortical spreading depression could reinstate septal rage for the duration of its effect. Yutsey, Meyer, and Meyer (1964) demonstrated that anterior neodecortication coupled with septal ablation in the rat produced septal rage which did not dissipate spontaneously. They found it necessary to handle and gentle their animals in order to alleviate this deficit.

PAGE 32

24 Sodetz, Matalka, and Bunnell (1967) could find no evidence for septal rage in the hamster and they cite unpublished reports of failure to find septal rage in other species. Buddington, King, and Roberts (1967) have failed to find septal rage in the squirrel monkey and they support Votaw's (1960) similar finding for the Rhesus monkey. Bond, Randt, Bidder, and Rowland (1957) were unable to detect septal rage in their cats. More recently, Moore (1965) has reported finding septal rage in but one of eleven cats used in his study. While the evidence is suggestive of septal rage being a phenomenon specific to the rat, the problem is not that simple. Brown and Slotnick (1966) have reported that female mice demonstrate some of the elements of septal rage following septal rage and McMullen and Slotnick (1967) explicitly state that septal rage is reliably obtainable from the mouse . M In species which respond to septal ablation with a transient hyperemotionality, there is evidence to suggest that brain catecholamines are involved in the mediation of this behavior. Pirch and Norton (1965) have been able to correlate the occurrence of septal rage with a fall in the concentration of norepinephrine. Raitt, Nelson, and Tye (1961) , and Pirch and Norton (1966) , have demonstrated that the administration of chlorpromazine to the septal rat attenuates septal hyperemotionality. It is commonly held that chlorpromazine occupies or interferes with activity at catecholaminergic post synaptic sites (Glowinski and Baldessarini, 1966). This is inferred from the fact that chlorpromazine exerts its effect without producing any change in norepinephrine levels in the brain (Gey and Pletscher, 1961). Apparently, existing norepinephrine remains bound in its storage sites.

PAGE 33

25 Chlorpromazine might be expected to retard or eliminate the gradual depletion of norepinephrine which has been correlated with the onset of septal hyperemotionality . Horowitz, Furginelle, Brannick, Burke, and Graver (1963) have demonstrated that pharmacological depression of the amygdala through the administration of antidepressants attenuates septal hyperemotionality. This work supports the earlier work of King and Meyer (1958) who reported a similar result following ablation of the amygdaloid complex. Conditioned suppression or the conditioned emotional response (CER) has been used as an operational definition of emotion (see Brady 1958a; 1958b; 1961). Brady and Nauta (1953; 1955) found impaired retention of a preoperatively conditioned CER following lesions of the septal region, habenulae, and hippocampus of the rat. This effect persisted after all evidence of septal rage had disappeared. It would appear that this loss of what is usually considered to be a fear response is inconsistent with the reports of hyperemotionality. However, this is so only if one views these data in isolation from data related to other deficits seen to follow septal ablation. Avoidance learning . The septal region has been shown to be related to the performance of an avoidance response. King (1958) demonstrated that septal ablation resulted in significantly superior performance in the acquisition of a two-way shuttle box avoidance response. McCleary (1961) differentiated active and passive avoidance in testing the behavior of septal cats. He found, as did King (1958), that the lesion facilitated two-way shuttle box avoidance. However, in contrast to the active avoidance problem, septal cats were deficient in acquiring the passive avoidance task. This task required that the

PAGE 34

26 animals withhold an approach response In order to avoid shock. McCleary interpreted these results to be consistent with the stimulation data of Kaada (1951) which had shown the subcallosal area to be involved in the inhibition of motor activity. The passive avoidance deficit was considered to result from the lesioned animal's inability to inhibit a response. Kaada, Rasmussen, and Kviem (1962) verified McCleary's (1961) observation of a passive avoidance deficit in the cat with rats. This early work has since been replicated a number of times (Gurowitz and Lubar, 1966; Lubar, 1964). The enhancement of two-way active avoidance learning has also been demonstrated in succeeding studies (Kenyon and Krieckhaus, 1965; Krieckhaus, Simmons, Thomas, and Kenyon, 1964; Schwartzbaum, Green, Beatty, and Thompson, 1967). Other learning paradigms . Many of the studies of learning deficits following septal ablation appear to have been influenced by the response disinhibition hypothesis of McCleary (1961). Ellen and Powell (1962a; 1962b) demonstrated that septal rats differed from normal animals in their performance on a fixed interval reinforcement schedule. The septal rats took nearly twice as long as the normal animals to learn to concentrate their responses in the last part of the interreinforcement interval. When they finally acquire this characteristic behavior pattern, septal rats show a higher than normal terminal response rate. Both of these deficits appear to be consistent with a response disinhibition hypothesis. However, Harvey and Hunt (1965) have raised the question of motivational changes by demonstrating that septal rats will respond more often than normals for water reward. Studies to be discussed below (Fisher and Coury, 1962; Grossman, 1964; Vilar, Gentil, and Covian, 1967) have implicated the septal area in the mechanism

PAGE 35

27 underlying drinking behavior. Rich and Thompson (1965) and Thompson and Langer (1963) have found deficits in conditioned avoidance retention and position reversal learning. Douglas and Isaacson (1966) have shown septal ablation to eliminate spontaneous alternation. These results are not consistent with a heightened emotional state, but rather suggest a more responserelated mechanism. Zucker (1965), measured food intake in his animals and could find no change; yet, his animals were deficient on both successive discrimination learning and position reversal learning. Gurowitz and Lubar (1966) have shown that septal cats showing evidence of a passive avoidance deficit related to withholding of a motivated approach response to an electrified food cup consumed no more food than did normals in the same situation. The septals differed from normals in the number of responses per gram of food consumed. The septals responded many more times than did the normals. Ellen, Wilson and Powell (1964) and Burkett and Bunnell (1966) have both reported septal ablation to produce deficits on a DRL schedule. Septal animals respond more often than normals during the period in which responses should be withheld. Feeding, drinking, and sleep . Fisher and Coury (1962) , Grossman (1964) , Quarterman and Miller (1966) , and others have used stimulation of the septal-preoptic region to elicit drinking behavior. These studies seem to indicate a cholinergic substrate underlying this behavior since the application of carbochol elicits drinking while catecholaminergic substances do not. The effect may not be simple, however, for lesions of the septal region also result in increased water intake (Harvey and Hunt, 1965; Vllar, Gentil and Covian, 1967). Chemical stimulation has generally been considered to be the functional opposite

PAGE 36

28 of ablation. Since the data indicate that both stimulation and ablation result in an increment in the amount of water ingested, the possibility of the same behavior being produced by two independent functional systems must be entertained. It may be that the rats in Grossman's (1964) study drank because a drinking mechanism had been activated and that the rats in Harvey and Hunt's (1965) study drank because of a general deficit in dealing with their response tendencies. Carey (1967) has suggested that the drinking mechanism may be anatomically distinct from that which controls response inhibition. However, his histology suggests the need for further examination of this hypothesis. Gurowitz and Lubar (1966) have shown that the food intaHe of septal rats is no different than that of normals; however, there is a tendency for the septal to make more responses to acquire the food. Sodetz (1965) has observed a weight increase of nearly 20 per cent following septal ablation in the hamster; however, measures of food intake were not taken . The septal-preoptic region has been cholinergically implicated in the production of sleep (Hernandez Peon, et al., 1963; Velluti and Hernandez Peon, 1963). Although the possibility of producing sleep by the electrical stimulation of this region had been debated for some time (Delgado, 1964), specific evidence for such an effect was lacking. The demonstration of a cholinergic mechanism was the first convincing demonstration of a limbic sleep circuit. Recent electrical stimulation studies have confirmed the presence of this circuit (Sterman and Clemente, (1962) . Related structures The studies to be considered in this section are representative of

PAGE 37

those which may provide some insight into the systematic relationships of structures which have been defined as part of the functional limbic system. Emotion . Petsche, et al. (1962) have convincingly demonstrated the close functional relationship of the septal region and the hippocampus; however, ablation studies have generally failed to produce modifications in emotionality though the destruction of either the hippocampus or the fornix (Allen, 1948; Rothfield and Harman, 1954). Changes in affective reactivity have been reported following destruction of the amygdala in a variety of species (e.g., Schreiner and Kling, 1956). This effect is generally characterized as a loss of fear, or ferocity but operationally what has been observed to follow amygdalectomy is a willingness on the part of the organism to approach objects which would have been avoided preoperatively and a failure to engage in intra and interspecific aggression. (Plotnik, 1966; Rosvold, Mirsky, and Pribram, 1954; Weiskrantz, 1956). Cingulate ablation has been shown to produce affective changes in rats and primates, but there has been little agreement as to the nature of these changes. Kennard (1955) found cingulectomized cats to be aggressive fighters, while Mirsky, Rosvold, and Pribram (1957) found their cingulectomized monkeys to be less fearful of man but unchanged in their reactions to conspecif ics . Pechtel, McAvoy, Levitt, Kling, and Masserman (1958) reported their animals to be slightly more aggresave toward both man and conspecif ics . Spiegel and Wycis (1949) and Wycis and Spiegel (1951) have reported that ablation of the dorsomedial thalamic nucleus produces a decrement in emotional reactivity. Destruction of the f rontogranular cortex was often used in the treatment of emotional disorders (Fulton, 1951) .

PAGE 38

30 Heller, et al. (1962) have reported that lesions in the dorsomedial mesencephalic tegmentum produced an Increment in the aggressiveness of the rat toward conspecifics without any overt evidence of hyperemotionality to human handling or any other notable behavioral changes. ^•-e? DeMolina, and Hunsperger (1962), electrically stimulated various hypothalamic sites and have mapped medial and lateral regions which somehow underly flight and defensive reactions. These areas extend throughout the rostrocaudal axis of the hypothalamus and are at least partially under the control of the amygdala (Hilton and Zybrozya, 1963; Zybrozya, 1960; Fernandez DeMolina and Garcia-Sanchez , 1967). The diversity of behavioral changes associated with stimulation and ablation of limbic structures suggests that the observed emotional disorders may be phenotypic byproducts of more specific functional disorders and that the limbic system is not a mechanism for the elaboration and apprehension of emotional states, but rather a more general system dealing with external and internal events. Emotional anomalies would follow damage to this system because the organism is no longer able to process data in an adaptive manner. Learning . Of the structures under consideration, the hippocampus most closely parallels the septal region with regard to the observable properties of learning deficits produced by its destruction. This should not be taken to mean that these areas subserve similar or even closely related functions, for the analysis of the learning process has not reached such a level of sophistication that it can be said with any certainty that similarities in observable deficits reflect the disruption of similar neural processes. It is conceivable that the septal region and the hippocampus each contribute to the learning process in

PAGE 39

31 such a way that, In many instances, the destruction of one or both would have the same behavioral outcome. Isaacson, Douglas, and Moore (1961) have demonstrated superior two-way active avoidance learning and deficient passive avoidance performance following bilateral hippocampectomy in the rat. Burkett and Bunnell (1966), and Ellen, Wilson and Powell (1964) have demonstrated a deficit in DRL performance following septal ablation in the rat and Clark and Isaacson (1965) and Schmaltz and Isaacson (1966) have demonstrated a similar deficit in hippocampectomized rats. There is some suggestion that the production of this deficit depends upon the method of training used in shaping the animal's DRL performance. If the animal is shaped with continuous reinforcement and then shifted gradually over to the DRL schedule or if it is trained from the beginning on a DRL schedule, the deficit is not as severe as it is when the animal is shifted from continuous reinforcement to the DRL schedule in one step (Schmaltz and Isaacson, 1966). An important distinction between the deficits produced by septal ablation and those which are obtained from hippocampectomy may be found in the work of Douglas and Raphelson (1966b) and Sodetz (1965) . Douglas and Raphelson (1966b) found septal rats to be no more active in an exploratory situation and found nearly a 50 per cent decrement in activity as measf ured in an activity wheel. Sodetz (1965) found septal hamsters to decrease their activity over trials on an open field activity board at the same daily rate as normal hamsters. Sodetz (unpublished observations) found that septal hamsters averaged approximately 50 per cent fewer revolutions per night in an activity wheel. Hippocampectomized rats, however, do not exceed normal rats in the initial level of

PAGE 40

32 activity observed in an exploratory setting. However, they persist in this behavior over trials and do not show the activity decrement associated with habituation (Roberts, Dember, and Brodwick, 1962). While septal and hippocampal ablation (Douglas and Isaacson, 1964; Douglas and Raphelson, 1966a) both reduce the tendency of the rat to alternate its responses, septal ablation does not produce a deficit in maze learning (Thomas, Moore, Harvey, and Hunt, 1959) which can be attributed to this deficit, whereas, hippocampal ablation does (Douglas and Pribram, 1966) . Damage to the cingulate cortex has been shown to result in a deficit in two-way active avoidance (McCleary, 1961; Peretz, 1960), and normal performance in passive avoidance (Lubar, 1964). It seems likely, however, that significant improvement in passive avoidance learning would be difficult to detect since normal animals learn this tjrpe of task very quickly. Lubar (1964) included a group with combined cingulate and septal ablations. These animals showed no deficit in passive avoidance. The fact that cingulate ablation counteracts the effect of septal lesions suggests that this region has a role in this behavior which is not discernable in comparisons with normal animals. It seems likely that the cingulate gyrus serves a function in avoidance learning which is pheno typically antagonistic to that of the septal region and the hippocampus. Thomas and Slotnick (1962) could find no evidence of a maze learning deficit in the rat following cingulate ablations. However, they did find a deficit in conditioned avoidance responding. They suggested (Thomas and Slotnick, 1963) that this deficit was produced

PAGE 41

33 by the tendency of the animal to "freeze" in response to the CS. Hypothesizing that food deprivation would tend to increase the activity level of the cingulectomized animals and thereby minimize freezing, they added this seemingly irrelevant drive to their paradigm and found the performance of the animals to compare favorably with that of their controls. Cingulate lesions, therefore, differ from hippocampal ablation in their effects on maze learning and active and passive avoidance learning. Septal and cingulate lesions appear to be similar in their failure to adversely affect maze learning. Horwath (1963) has found that bilateral amygdalectomy produces a deficit in one-way active, two-way active, and passive avoidance. The effect on passive avoidance, though slight, was statistically significant. The generality of this deficit is suggestive of a disruption of a function critical to the general learning process. The work of Kluver and Bucy (1937) supports this view insofar as the behavior of their monkeys can be characterized as being relatively insensitive to the stimulus characteristics of their environment. Schwartzbaum (1960) has demonstrated that the amygdalectomized monkey is relatively insensitive to changes in the magnitude of reward. Hi-low shifts and low-hi shifts had little effect on their response rate. Perhaps, the contradictory character of nearly every study which has investigated amygdaloid function (See Goddard, 1964) is not a product of the convergence of a great number of functional systems in the region of the amygdala, but is the result of the disruption of the mechanism which makes lawful decisions according to the rules of reinforcement. If this is the case, then the failure of studies of the amygdala to agree may be because the behavior of the amygdaloid animal is being interpreted in terms of the

PAGE 42

34 laws which govern the behavior of the normal animal while the animal is behaving according to a different set of laws. Douglas and Pribram (1966) have suggested that the amygdalectomized animal is incapable of responding in terms of reward and have found evidence consistent with the hypothesis that the animal conducts its affairs entirely in terms of non-reward. Thompson and Langer (1963) and Rich and Thompson (1965) have selectively ablated various limbic, thalamic, and hypothalamic sites in an attempt to delimit a functional system underlying position reversal learning and avoidance conditioning. Their studies have implicated the septal region, hippocampus, mammillary bodies, anterior thalamic, and midline thalamic nuclei in a system which somehow subserves functions necessary to the retention of these habits. Roberts and Gary (1956) have reported that lesions of the dorsomedial thalamic nucleus interfere with the acquisition of a conditioned fear response. Thompson and Rich (1961) have found lesions of the interpeduncular nuclei to temporarily disrupt performance of an avoidance test. Hatton (1965) has confimed their result, but reports no evidence of interference with a discrimination problem. Some relevant pharmacological studies . Pharmacological studies offer indirect support for many of the results sited above. The septal region and hippocampus have been shown to be important to the maintenance of spontaneous alteration (Roberts, et al., 1962; Douglas and Raphelson, 1966a) . In addition, there is evidence that cholinergic mechanisms may mediate at least part of their function. If this is so, then the administration of centrally acting anticholinergic substances might be expected to produce results very similar to those obtained in lesion

PAGE 43

35 studies. It should be emphasized, however, that the systemic administration of these agents precludes any exact localization of their site of action. Any attempt at relating these effects to a specific area of the brain must be viewed with some reservation. The systemic administration of scopalamine produces a reversible loss of spontaneous alternation behavior (Douglas and Isaacson, 1966; Meyers and Domino, 1964). Scopalamine has also been shown to produce a reversible passive avoidance deficit (Meyers^ 1965) and to attenuate a conditioned fear response (Vogel, Hughes, and Carlton, 1967). Atropine sulfate, another anticholinergic agent, disrupts maze learning, while methyl atropine, a related substance with little central effect and a strong peripheral action, does not interfere with maze learning (Whitehouse, Lloyd, and Fifer 1964). Chalmers and Erickson (1964) have found cholinergic agents to be effective in producing both an acquisition and retention deficit in a shuttle-box avoidance task and in a lever pressing conditioned avoidance problem. They related this deficit to the tendency of their animals to freeze in response to the conditioned stimulus. This was essentially the observation made by Thomas and Slotnick (1963) who tested their hypothesis by adding an irrelevant drive to elevate the activity level of their cingulectomized animals. Both Chalmers and Erickson (1964) and Thomas and Slotnick (1963) report the observation of behavior suggesting that their animals responded with fear to the conditioned stimulus, but simply failed to initiate the avoidance response. A study by Hearst (1964) demonstrated that both scopalamine, an anticholinergic, and amphetamine, an agent which induces the release of catecholamines, flattened the generalization gradient of the monkey

PAGE 44

36 performing during the acquisition of an eight choice shock avoidance discrimination problem. Monkeys treated with either of these substances tended to distribute their responses more widely over the possible stimuli, while normal monkeys produced the typical curves associated with generalization to closely related stimuli. In this case, two chemical systems may subserve phenotypically similar functions but it may be that these functions are genotypically distinct. Catecholaminergic agents have been found to be related to behavioral changes usually associated with the limbic system. Andersson and Larsson (1957) demonstrated that surgical separation of the frontogranular cortex of the dog blocked the reduction of food and water intake associated with the administration of amphetamines. Hunt (1957) and Raitt, et al. (1961) reported meprobamate and chlorpromazine to be effective in attenuating septal hyperemotionality . Randall (1961) has shown chlordiazepoxide (Librium) also to be effective in this regard. Similarities and interactive effects between limbic ablations and the manipulation of pharmacological substances related to cholinergic and catecholaminergic mechanisms have been demonstrated. Studies such as these may provide a basis for the identification of genotypically distinct functional systems in instances where the observable end products of their activity are phenotypically identical. The limbic system and species-specific behavior Species-specific behaviors are responses common to a given class of organisms. No assumption is made as to whether such behaviors are learned or innate. Furthermore, no distinction is made between those behaviors which are structured by internal mechanisms and those which may be imposed upon the organism by the properties of its ecological

PAGE 45

37 niche. Species-specific behaviors are those response patterns which are phenotypically identical in all members of a species which have been exposed to the natural selection factors most characteristic of those operating on the species as a whole. Members of a species which are not exposed to these same environmental factors may differ with respect to many of the learned components of these behaviors while they may or may not retain the innate behaviors common to their species. Consequently, studies of the species-specific behaviors of laboratory animals may produce results which are limited in their generality. This factor should be given consideration when evaluating the significance of such studies. Another possible limiting factor rests in the conditions under which the behavior of the organism is observed. The more restrictions the experimenter places upon the behavior of the animal, the more likely it is that the obtained results will not be representative of the species under consideration. The problem still remains as to how a study may be identified as being related to species-specific behavior. The criteria for delimiting species-specific behaviors include its presence throughout a representative sample of the species and the fact that it is idiosyncratic to the species under consideration. More subtle distinctions are made on the basis of the restrictions placed upon the organism. It can be said that the study of the behavior of the rat on an open field activity board is species-specific if no restrictions are placed upon the behavior of the animal. The response pattern of a rat in such a situation might differ markedly from that of a pigeon under the same circumstances. If, however, the experimenter measures only the number of squares entered by the animal, it is not likely that his observations

PAGE 46

38 can be characterized as dealing with species-specific behavior. Similarly, if a rat is placed alone in an empty box, with a drinking tube protruding from the wall, it may be observed to drink following the application of some chemical to the hypothalamus . This behavior cannot be characterized as species-specific, however, for the nature of the apparatus restricts the organism to a discrete set of responses. In such a study, it might be found that the introduction of nesting materials into the box might result in the animal ceasing its drinking behavior and initiating nest building, or perhaps going to sleep. The study of species-specific behavior requires that internal and environmental variables be left relatively free to operate in the experimental setting. A measure of control is lost in such a design and it becomes difficult to make explicit statements about the properties of the functional systems which underlie the observed behavior. However, it is likely that the organism employs most of its faculties in the conduct of its affairs and the sampling of a broad range of behaviors may produce evidence of deficits which might take far longer to detect in a more restrictive experimental setting. MacLean (1958) has suggested that the limbic system is the mechanism responsible for those behaviors which are adaptive in the preservation of the individual and the species. This functional characterization of the limbic system may be correct insofar as all adaptive behavior is, by definition, consistent with the preservation of the individual and the species. It is reasonable to assume that many of the same laws which govern the behavior of the organism in a laboratory learning task also control the naturally occurring behaviors of the free-living organism.

PAGE 47

39 Studies of the limbic system, in the laboratory setting, have been discussed in preceding sections. This section will deal with observations of the effects of manipulation of the limbic system on species-specific behavior. Recent systematic observations of changes in species-specific behavior following damage to the limbic system (Brody and Rosvold, 1952; Bunnell, Friel, and Flesher, 1966; Bunnell, et al., 1967; Bunnell and Smith, 1966; Bunnell, et al., 1965; Clemente, Green, and deGroot, 1957; Delgado, 1964, 1966; Fisher, 1956; Fuller, Rosvold, and Pribram, 1957; Green and Kling, 1966; Kennard, 1955; Kling, 1962; Mirsky, et al., 1957; Pechtel, et al., 1958; Plotnick, 1966; Rosvold, et al., 1954; Schreiner and Kling, 1953; Sodetz, 1965; Sodetz and Bunnell, 1967a; 1967b; Stairan, 1954; 1955; Weiskrantz, 1956) have repeatedly demonstrated that this system and related structures are closely related to the mediation and control of these behaviors. Amygdalectomy has been shown to result in the reduction of the aggressiveness of the cat (Schreiner and Kling, 1953), monkey (Plotnik, 1966; Rosvold, et al., 1954), dog (Fuller, et al., 1957), hamster (Bunnell, et al., 1965), and rat (Bunnell, 1966) toward conspecif ics . In contrast to this, the same lesion produces a taming effect in a variety of nondomesticated species (Schreiner and Kling, 1956) and a willingness on the part of the monkey to approach man (Weiskrantz, 1956) which might be interpreted as an increased interspecific aggressiveness. Rosvold, etal. (1964) suggested that amydalectomy in the monkey resulted in the animal decreasing in rank in a dominance hierarchy. Plotnik (1966) and Bunnell, et al. (1965) found a decrease in the frequency of social

PAGE 48

40 , interaction following amygdalectomy in the monkey and hamster. In the hamster (Bunnell, et al., 1965) a general depression of social activity was observed which transcended both aggressive and submissive behavioral categories. This data is suggestive of a tendency on the part of the amygdaloid animal to become asocial rather than submissive. Cingulate ablation has also been related to social behavior. Kennard (1955) found that cingulectomized cats fought more than normals. Mir sky, et al. (1957) could find no change in the social behavior of the cingulectomized monkey, but like the amygdaloid monkey these animals appeared to be less fearful of man. Pechtel, et al. (1958) noted the discrepancies in the results of earlier studies, but could only add that their animals appeared to be somewhat more aggressive toward both man and conspecif ics . They did, however, observe that their animals failed to care for and nurse their young. Stamm (1955) had earlier observed a similar deficit in the maternal behavior of the cingulectomized rat. These rats failed to care for their young following birth. If, however, foster pups, which were old enough to attach themselves to the female, were placed with the animal, it tolerated their nursing, following which the female began to show more normal pup retrieving behavior. Such animals also failed to build nests and, in another study, Stamm (1954) found them to be deficient in hoarding. A nest building and hoarding deficit has been observed to result from septal ablation in the hamster (Sodetz, 1965), but not from ablation of the amydala (Bunnell, et al., 1965) . Hippocampal ablation also produces some deficit in hoarding (Bunnell, et al., 1967); however, it is less severe than that seen

PAGE 49

in the septal hamster. Both pre-f rental (Brody and Rosvold, 1952) and mesencephalic tegmental (Heller, et al., 1962) damage have been shown to affect the aggressive behavior of the animals studied. Prefrontal lobotomy produced changes in the behavior of monkeys very similar to those seen following cingulectomy or amygdalectomy . In the case of the pre-frontal animals, however, the persistent approach response to previously avoided animals appeared to disappear within a few months and the dominance heirarchy of the monkeys stabilized in a form which differed little from that seen preoperatively . Rats with dorsomedial mesencephalic tegmental lesions (Heller, et al., 1962) were found to be placid by human handlers, but these same animals engaged their normal cagemates in almost continuous aggression throughout the 35 days they remained in the study. Sodetz (1965) and Sodetz and Bunnell (1967a; 1967b) found ablation of the septal region of the hamster to produce changes in aggressive behavior which were consistent with the social history of the animal. Septal hamsters which had no social history of defeat by normal opponents became dominant and displayed more aggressive behavior in post-operative trials. Septal hamsters which had experienced defeat before surgery became more submissive following ablation. Bunnell, et al. (1967) found hippocampectomized hamsters which had no preoperative social experience to be submissive in postoperative trials unless virtually unopposed by their opponents. They also observed that ablation of the hippocampus in animals which had a history of defeat produced an increment in submissive behavior similar to that observed to follow septal ablation in submissive animals. Their comparison of the observed sequences of behavior of hippocampal and septal hamsters

PAGE 50

indicated that the two groups, though showing similar changes in the frequency of submissive responses, differed in terms of the sequential arrangement of these behaviors. In this regard, septal animals were very similar to normals while hippocampal hamsters differed from both groups. Pharmacological studies of the species-specific behavior of a niamber of forms have been done. Apparently, alterations in speciesspecific behavior provide a very sensitive instrument for the screening of drugs. Horowitz, et al. (1964) report that mouse-killing by rats was precluded by the administration of anti-depressant drugs, whereas dozens of tranquilizers tested were without effect. It was their hypothesis that anti-depressants exert their effect by selectively depressing the amygdala. Karli (1960) has shown that septal ablation increases the likelihood of a given rat killing mice. Tranquilizers have been shown to diminish septal hyperemotionality (Raitt, et al., 1961). This may be evidence for a functional distinction between emotionality changes produced by limbic damage and other behaviors such as mouse-killing which do not appear to be affected by the same class of pharmacological agents. Bignami (1964) has shown that anticholinergic substances interfere with the mating behavior of the male rat. He suggests that this is due to a disinhibition of the orienting reflex. This hypothesis appears to be consistent with the evidence for cholinergic mediation of hippocampal theta rhythm (Stumpf, 1965) and current concepts of its functional significance (See Douglas, 1967). DeVanzo, Daughterty, Ruckart, and Kang (1966) have demonstrated that monoamineoxidase inhibitors reduce

PAGE 51

43 fighting behavior in mice. They could not, however, detect any difference between the brain norepinephrine levels of fighters and nonfighters, which lead to the conclusion that the effect must be mediated through one of the other catecholamines. Serotonin could be a likely possibility because emotionally reactive and non-reactive strains of rats have been shown to differ in their brain serotonin concentrations (Sudack and Maas, 1964); however, Heimstra and Sailer (1967) have recently demonstrated that mice chronically treated for a period following weaning with an agent which increases brain norepinephrine levels will be more likely to dominate their opponents in aggressive encounters as adults. Silverman (1966) found that chlorpromazine, which depresses the activity of adrenergic systems, decreased aggressive social interaction in laboratory rats, tended to increase flight in response to threat and attack, and had no effect on non-social behavior. Amobarbital, which is a CNS depressant, increased social behavior in general and threat responses in particular. Norton (1957) studied the effects of a number of pharmacological agents on the social behavior of monkeys, hamsters, and cats. She found consistent effects across species. Chlorpromazine increased social behavior in general, while decreasing behavior associated with agitation, defense, and attack. Behavior such as grooming, which she took to be indicative of contentment was seen more frequently in animals treated with chlorpromazine. Amphetamine, which excites cateholaminergic systems, was found to decrease sociability, increase attack and defense, and increase evidence of excitement. Both cholinergic and catecholaminergic systems appear to be involved in species-specific behavior. The fact that most limbic

PAGE 52

44 structures appear to be served by both systems permits the possible number of combinations of deficits which could result from ablation of just one of these structures to be large. It is not surprising that pharmacological studies which have employed the systemic administration of psychotropic agents do not present a consistent pattern of results. There is no way to be certain just where in the CNS these agents are acting, although the lesion data strongly support the view that they are acting on limbic structures. The work of Sodetz (1965) and Sodetz and Bunnell (1967a; 1967b) has shown septal ablation to produce changes in both the social behavior and non-social behavior of the hamster. These effects have not been amenable to description in terms of any uniprocess model of septal function (e.g., McCleary, 1961). Furthermore, the studies of Bunnell, et al. (1965) and Bunnell, et al. (1967) have demonstrated differences and similarities in the effects of septal and hippocampal ablation and differences in the effects of septal and amygdaloid ablation. The pharmacological data suggest that both cholinergic and catecholaminergic systems in these structures may be operative in the mediation of species-specific behavior. A septal lesion would of necessity interdict both types of systems. Therefore, it is likely that the failure to develop an appropriate uniprocess model to account for the observed changes in behavior is due to interference with more than one chemically distinct functional system. The present study was designed to attempt to identify chemically distinct functional systems subserving species-specific behavior through the application of cholinergic and catecholaminergic agents and their antagonists to the septal region of the forebrain.

PAGE 53

METHOD The present study was conducted as a series of experiments each of which differed in the particulars of their methodology. What follows is a general description of the elements common to the entire study. Conditions unique to a given experiment will be discussed in separate sections devoted to each individual experiment. Subjects . One hundred fifty Manor Farms random bred adult male hamsters ( Mesocricetus auratus) were used in the study. All of the S^s were obtained as weanlings at an age of about 27 days. Upon receipt at the laboratory, all Ss were transferred to single cages and housed individually until reaching an age of from 90 to 130 days. At that time, samples were drawn from the subject population and assigned to individual experiments. While in the laboratory, all S^s had paper nesting material, food, and water continuously available to them. Handling of the S^s was kept at a minimum. The Ss were raised in social isolation in an attempt to maximize the likelihood of their engaging in aggressive social interaction. Hamsters which have been raised in a group are less likely to aggressively engage another similarly experienced animal. Social isolation appears to be an effective and legitimate means of avoiding this difficulty. Hamsters which have experienced defeat in aggressive interaction cease to initiate such activities. In a group of hamsters, one animal characteristically dominates all others, therefore only the dominant animal from a given group would readily engage an opponent. Since isolation-reared animals have had little 45 -

PAGE 54

46 experience with defeat, a far greater proportion of them will engage opponents . In each experiment the S^s drawn from the subject population were assigned to pairs matched by weight and age. Differences in age and weight within a given pair did not exceed five days and five grams, respectively. Differences of this magnitude appear to in no way affect the outcome of an aggressive encounter. Following their assignment to pairs, all S^s were transferred from the single cages in which they had been living to the dominance cages to be described below. They remained in these cages throughout the experiment. No was ever used in more than one experiment . Apparatus . The three primary pieces of apparatus used in the study were the dominance cages, recording apparatus, and cannulae. The dominance cages, in each of which a pair of S^s were both housed and tested, were constructed from double rat cages (45 cm. x 24 cm. X 20 cm.). They were divided into three equal compartments by two metal partitions. The floors of the cages were lined with 1/4 inch hardware cloth. One animal lived in each of the two end compartments which were separated by the third center compartment. The center compartment served as a "neutral" area and was included in the design to permit the operation of territorial effects. Access to all compartments could be made available by raising a guillotine door in each of the partitions. This permitted unrestricted interaction between members of a pair. The recording apparatus was designed to provide a continuous record of the occurrence of 27 behaviors which can be observed in the interaction of two adult male hamsters. The apparatus consisted

PAGE 55

47 of a Gerbrands multi-channel event recorder, one channel of which was assigned to each animal. A third channel was connected to a Kramer clock, which activated the channel once each second. A code symbol was assigned to each of the 27 behaviors. Two trained observers, one for each animal in a pair, sequentially listed the behaviors they observed thereby producing a written record of all the behaviors of each animal for a given five minute observation period. As each behavior was noted, the observer also closed a switch assigned to his S^, which entered the occurrence of the behavior on the output of the recorder. The output of the recorder, including the time base, was later collated by computer with the written response record to yield data on response frequencies , durations, sequences, and inter— animal response contingencies. The double-walled cannulae used for chemical stimulation were constructed from three different gauges of stainless steel hypodermic tubing. These were selected because, when placed concentrically within one another, each formed a close fit within the next larger. The outer cannula consisted of a piece of 22 gauge tubing 11 mm. in length. This piece formed a guide for the inner cannula and was the portion inserted into the brain. Each of the inner cannulae consisted of a piece of 28 gauge tubing soldered within a piece of 18 gauge tubing approximately 3 mm. in length. To complete the cannula, the inner component was inserted into the outer guide with the 18 gauge tubing forming a snugly fitting cap. When fully inserted the tip of the inner cannula fit flush with the tip of the guide. A Hamilton microliter syringe fitted with a 28 gauge needle and a guide

PAGE 56

to limit insertion to exactly 11 mm. was used to inject chemicals into the brain. This was accomplished by removing the inner cannula, and in its place, inserting the needle of the syringe. Following administration of the chemical, the needle was withdrawn and the inner cannula replaced. Response inventory . Table 1 presents a list and a brief description of each of the behaviors seen in the interaction of two adult male hamsters. The behaviors can be organized into eight categories such as hoarding and fighting. These response categories can be grouped into two classes, the social activities which in some way involve another animal and the non-social categories which bear no relationship to the other animal. Surgery . Following assignment to pairs and transfer to the dominance cages, one member of each pair of ^s underwent surgery to implant a single cannula and, in some cases, to produce a lesion in addition to the placement of the cannula. Surgery was performed under sodium pentobarbital anesthetic (90 mg/kg) . Atropine sulfate was administered intraperitoneally approximately 30 minutes before administration of the anesthetic. The S^'s head was shaved and the ^ was mounted in a stereotaxic instrument. Following incision of the scalp, the skull was cleaned and air-dried. The head was leveled between the parietal-occipital and the frontonasal skull suture and a trephine hole 3 mm. in diameter was bored over the midline about 6 mm. rostral to bregma. Three small holes, one in the dorsal aspect of each of the parietal bones and one on the midl ine in the nasal bones just rostral to the frontonasal suture, were bored in the skull

PAGE 57

Table 1. A brief description of the responses which can be seen in the interaction of two adult male hamsters.

PAGE 58

50 RESPONSE CATEGORIES NON-SOCIAL ACTIVITY HOARDING M = Middle or Neutral Area HC = Home Cage OC = Opponent Cage E = Exploration of Apparatus and Contents UA = Pouch Paper HA = Hoard Paper UO = Pouch Food HO = Hoard Food SOCIAL SOCIAL INVESTIGATORY AGGRESSION N = Sniff Opponent (Requires Contact) N'= Sniff (Analgenital) A = Attack Posture (Side) A'=Attack Posture (Underneath) B = Bite C = Chase DEFENSE FIGHTING P = Defense Posture (Upright) P'= Defense Posture CSide) S = Spar F = Fight x/y = Pin SUBMISSION MISCELLANEOUS T = Tail Lift T'=Tail Lift (With Adduction of Hindlimb) Z = Freeze I = Orient Toward Opponent V = Vocalize G = Groom ESCAPE W = Attempted Escape from Cage L = Flee

PAGE 59

5i and a small stainless steel machine screw was inserted in each of them. This was done to provide binding sites for the dental cement which would hold the cannula in place. The skull was then cleaned and air-dried again. The cannula was stereotaxically placed through a slit in the dura into the septal region. An attempt was made to place the tip of the cannula into the midline of the septal region at a point midway in its rotrocaudal extent . It was hoped that such a placement would maximize the likelihood of the chemical bilaterally affecting most of the septal region. Placement of the tip of the cannula into the midline required that the cannula be inserted at an angle so as to avoid the bleeding associated with rupture of the saggital sinus. The stereotaxic coordinates used were 6 mm. (+ .3 mm.) anterior to bregma, 1 mm. lateral to the midline, and 4.25 mm. ventral to the surface of the cortex. The angle of penetration was 13.5 degrees. The dried and exposed area of the skull was then covered over with an application of Caulk "Grip" dental cement. When the cement was dry, any rough or sharp edges were removed with a dental burr and an antibiotic salve was applied to the area of the incision. The animal was then returned to its compartment in the dominance cage and permitted to recover for a period of from 11 to 14 days . A number of S^s received surgical treatment which differed somewhat from that described above. One group of such Ss had the cannula placed in the lateral ventricle at a point 4.25 mm. rostral to bregma. Another group received either bilateral RF thermocoagulation of the amygdala or RF thermocoagulation of the body of the fornix in addition to having a cannula placed in the septal region.

PAGE 60

52 Procedure . Following the post-operative recovery period testing was begun. Fresh solutions of the appropriate chemical dissolved in physlologdcal saline were prepared daily in a concentration of 2 mg/ml. An operated and its normal opponent were transported in their dominance cage to the testing room. Here the operated was removed from the cage and placed in a closed plastic container. Nitrous oxide gas was introduced into the container, lightly anesthetizing the animal. The was then removed from the container and during the 5 to 10 second period during which the S_ remained anesthetized the inner cannula was withdrawn and 1 microl. of either the chemical in solution or physiological saline was administered to the animal via the outer cannula. The inner cannula was then cleaned and replaced. During the initial stage of each experiment an animal either always received the chemical or always received saline. While the operated ^ was being injected with the chemical, an assistant anesthetized the S^'s normal opponent with nitrous oxide gas. Both Ss were then returned to their respective compartments within the dominance cage. The Ss were left undisturbed for 10 minutes. At the end of this period the test trial was begun. A trial consisted of 5 minutes of unrestricted interaction between members of a pair. Two trained observers, one for each animal, recorded the behavior of the animals using a response inventory of 27 behaviors which can be observed in the interaction of two adult male hamsters. Concurrently, each written response was also entered into the record of a multi-channel event recorder. One channel served as a time base. A computer later collated the output of the recorder with the written response record. Subjects were never run more than one trial a day.

PAGE 61

53 Histology . Following the completion of testing, all operated ^s were anesthetized and perfused with saline followed by formalin. Their brains were then removed, fixed in formalin, and embedded in celloidin. The brains were sectioned at 30 micra and every fourth section was stained with cresyl violet. Subjects dropped from the study . The difficulty in maintaining chronic indwelling cannulae over a period of several weeks resulted in the loss of a number of S^s. The rule applied throughout the study called for excluding the data from these S^s if they had completed fewer than three of five trials. If they had completed at least three trials, their brains were recovered and the data obtained from them was retained. The same rule was applied to several S^s which died during the course of the study. If histological examination of the brain of a ^ showed the cannula to be placed somewhere other than within the septal region, the data from that was treated separately and is presented in Appendix A. Saline controls . Eighteen pairs of £s in the group which served as saline controls survived to be included in the study. In each pair one of the S^s had a single cannula placed within the septal region. Ten minutes before the beginning of each trial, each S received 1 microl. of sterile isotonic saline administered via the cannula. This group served as a control against which to compare the remaining groups . Norepinephrine group . In this group, six pairs of S^s met the criterion for remaining in the study. All six operated S^s received

PAGE 62

54 2 microgms. of norepinephrine (D-L-Arteronol) dissolved in 1 microl. of isotonic saline, 10 minutes before the beginning of each trial. Intraventricular norepinephrine group . This group, consisting of three pairs, received the same dose of norepinephrine noted above. However, the cannulae in these S^s were directed into the lateral ventricle. This group was included to control for any general effect which might have resulted from the administration of norepinephrine. Phenoxybenzamine group . Nine pairs of ^s survived to be included in this group. The operated S_ in each pair had a cannula placed in the septal region. Phenoxybenzamine (Dibenzyline) was administered via the cannula 10 minutes before each trial. Each dose consisted of 2 microgms. of phenoxybenzamine dissolved in 1 microl. of isotonic saline. Phenoxybenzamine is known to block the action of norepinephrine by occupying the sites at which it acts. Therefore, it does not have the disadvantage of many adrenergic blocking agents which cause an initial release of norepinephrine by occupying its storage sites (Goodman and Gilman, 1964). Intraventricular phenoxybenzamine group . There were four pairs of S^s in this group. They differed from the phenoxybenzamine group only in that the cannulae in these £s were directed into the lateral ventricle. Phenoxybenzamine fornix resection group . Four pairs of Ss survived in this group. One in each pair had a cannula placed within the septal region. In addition, the S. underwent bilateral RF thermocoagulation of the fornix during the operation to implant the cannula. Ten minutes before each trial, these Ss received the same dose of phenoxybenzamine as did the other groups in the series.

PAGE 63

55 Phenoxyb en z amine amygdalectomy group . One member of each of the four pairs in this group had a cannula placed within the septal region. In addition, the underwent bilateral RF thermocoagulation of the amygdaloid complex during the operation to implant the cannula. Each operated received the same dose of phenoxybenzamine 10 minutes before each trial, as did the other Ss in the series. Carbachol group . Four pairs of S^s survived in this group. The cholinergic substance used in this series was carbachol which has a stronger and more lasting effect than acetylcholine (Goodman and Oilman, 1964). Each operated received 2 microgms. of carbachol dissolved in 1 microl. of isotonic saline 10 minutes before each trial. Atropine group . The operated £ in each of the four pairs in this group received 2 microgms. of atropine sulfate dissolved in 1 microl. of isotonic saline 10 minutes before the beginning of each trial. Atropine sulfate is an anticholinergic agent. Summary of groups . A summary of the groups used in the study and the number of operated S^s in each is presented in Table 2. Statistical analysis . The data collected in the study presented a special problem in that no complete statistical analysis of the data could be made. The behavior of the hamster at any instant during an observation period is dependent upon the behavior which occurred in the immediately preceding instant and perhaps back through a chain of as many as eight or ten responses. Similarly, the occurrence of the given behavior determines with a specific probability level what the succeeding responses will be. These

PAGE 64

Table 2. A summary of the groups used in the study.

PAGE 65

57 Number of Nimber of S_s Surviving Group Ss Operated in Study Carbachol (C) 6 4 Cholinergic Atropine (A) 6 4 Norepinephrine (N) 10 6 Intraventricular Norepinephrine (IN) 4 3 Phenoxybenzamine (P) 10 9 Catecholaminergic Intraventricular Phenoxybenzamine (IP) 4 4 Phenoxybenzamine Fornix (PF) 5 4 Phenoxybenzamine Amygdalectomy (PA) 5 4 Saline Saline (S) 25 18 TOTAL 75 56

PAGE 66

58 Internally generated sequential dependencies are further complicated by the impact of the behavior of its opponent upon the animal. It can be said with some certainty that the responses observed in the study are sequentially dependent on one another. The majority of parametric and non-parametric statistical tests require the assumption that the events in question be independent, that is, the occurrence of event A on trial 1 in no way affects the probability of its occurrence on trial 2 nor does it affect the probability of the occurrence of events B,C,D,...K. It is not possible to make this assumption in connection with the behavior observed in this study. If the animal has pouched food pellets and its cheek pouches are filled, the probability of it continuing to pouch pellets is very small and the probability of its emptying its cheek pouches is much greater than it would be if they were only partially filled or empty. With 10 different experimental conditions, the interaction of two animals, and a behavioral repertoire of 27 responses to consider over as many as five trials, the complexity of the sequential dependencies are such that they defy the application of any conventional statistical test. A Markov chain analysis can handle data such as these, for Markov chains were developed to deal with sequential dependencies. The complexity of the dependencies in the present study, however, demand a very sophisticated appreciation of Markovian analyses and the generation of thousands of responses so as to permit the specification of the exact probabilities of each response following any other response under all possible relevant conditions. Such an analysis is beyond the capabilities of the author. The data were

PAGE 67

examined with the aid of statistical tests where possible. However, because of the sequential dependencies present in the data, this analysis was very limited in scope and could not consider the various treatment effects in detail. An effort was made to be conservative in assessing those results which were not amenable to statistical test.

PAGE 68

RESULTS A group of eighteen Ss was used as a saline control against which to compare the effects of the administration of chemicals to the septal region of the forebrain. The behavior of these S^s was not expected to differ from that of their normal opponents. As many saline (S) animals were expected to be dominant as were expected to be submissive. Of the eighteen S^s in this group, eleven were submissive, as rated by the two observers, and seven were dominant. The response profile shown in Figure 1 reflects this tendency toward submissiveness . Figure 1 represents a comparison between the proportion of responses seen in each response category for the S group, represented by the bars, and the response profile of their normal opponents represented by the base line from which the bars are positively and negatively reflected. The ordinate represents the difference between the observed proportions in a given category. In the exploration (E) category, there was no difference between the S group and their normal opponents. In the social investigatory category, sniffing (N) , the proportion of responses falling into this category for the S group was .10 (rounded to the nearest one hundredth). The proportion for their normal opponents was .14 of the total number of responses observed. The negatively reflected bar was obtained by subtracting .14 from .10. The same procedure was followed in all of the other response categories. Figure 1 _ 60 _

PAGE 69

4) u •H 4J M-l 4-1 O M-l a O o •rl 4J U * cd 3 m 4-1 •H >-i m T) 4-1 C CO CO to •ri T3 II 00 0) CO r-l u II 4J a (U CO a 4J o c ft 3 0) ft O CO o u (U 00 Cli 0) 0) CO >-l •rl «\ iH ro CO o CO •H II (U •rl rH M-l •rl Cd O r-l C/i 0) CO CO C -ao rH •H CS| 4-> (11 *> M rO 4J P< CO o S cd u o > ft M-4 M-l (U o CO > CI u o (U o •H CO •H 4J 4J O o 3 (U •5 rH CO •H M-l (U >-i (U CO 4J 13 Ci CO o •H CO ft Cfl CO 01 (U XI 1-1 (U 4J CO MH CO o (h (U •H l-l u ft 0) CO 0) (U o g a ^ — \ 0) 00 H rH II Cd m 5 4J 14-1 o •H u •a CO 4J
PAGE 70

62 -

PAGE 71

indicates that the S group showed more evidence of escape behavior (L;W) and submissive posturing (T;T';Z) and less aggressive behavior (A;A';B;C) than their normal opponents. They also showed more evidence of defensive behavior (P;P'), attending (I), and vocalization (V) which is consistent with the response profile of the submissive normal hamster. The differences between the response profiles of the S group and their normal opponents tended to bias comparisons between the chemically treated groups and the S group. However, as will be noted below, the bias was not large enough to cancel those chemically produced effects which were large enough to reasonably be considered to be due to the treatment. Figure 2 compares the effects of the administration of either carbachol or atropine against the effects of the administration of saline. The response profile of the S group is represented by the baseline. Subjects treated with carbachol (N=4) were observed to explore (E) more than the S animals , although they were lower than the S group in most categories. It may not be reasonable to consider the .01 differences observed in a defense posture (P'), freezing (E) , vocalization (V), and grooming (G) as being due to the treatment. Atropine sulphate (N=4) has little effect on any behavior except exploration (E) , aggression (A;A';B;C), actual fighting (F) as opposed to sparring (S) , and attending (I) . It may be useful to the reader to pay special attention to the fact that although the carbachol group showed less submissive behavior than the S group, they did not show more aggressive behavior. In this regard, the atropine Ss showed more aggressive behavior, but did not show less submissive behavior.

PAGE 73

65 H o SI u (J 4) c a o (0 U
PAGE 74

66 The lack of an effect of carbachol on freezing (Z) and the pronounced effect of atropine on attending (I) should also be noted. Figure 3 represents the distributions of proportions of both the norepinephrine group (N=6) , and the intraventricular norepinephrine group (N=3) compared with that of the S group. The application of norepinephrine to the septal region produced a marked reduction in the exploratory behavior of the treated S^s when compared with the S group. Intraventricular norepinephrine produced an effect in the same direction but it was one-sixth as large as that produced by norepinephrine in the septal region. Neither location had an effect on hoarding behavior (UA;HA;UO ;H0) , while both may have been associated with a reduction in social investigatory behavior (N;N'). A slight decrease in aggression (A;A';B;C;) was effected at both sites as was an increase in defensive behavior (P;P'). Fighting (F) and sparring (S) were differentially affected. Norepinephrine placed in the septal region may have slightly increased fighting while intraventricular norepinephrine produced a marked increase in sparring (S) and pinning (X), which do not involve biting as does actual fighting. At both sites, norepinephrine decreased escape behavior (L;W), but norepinephrine increased the use of submissive postures (T;T';Z) only following application to the septal region. Attending (I) was also increased following the application of norepinephrine to the septal region. Vocalization (V) constituted .08 of the response total, and was recorded 89 times during the observation of the group treated with norepinephrine in the septal region. Figure 4 represents the distribution of proportions obtained

PAGE 75

1 3 a 0) CJ M •rl 0) 4J u }-{ c M o •H 4J CO 3 c ^ cd o •H CO CO 4-1 ft (U CO 3 CO • •rl O c T3 >-{ o CT\ bO ft cn (U CO r-f — \ 0) 4-1 ro >< CO II C3 •a II 0) o
PAGE 76

68 I/) -r Q. u CD 0) c a
PAGE 78

^00 ID < I < 3 ^!55555555555 hi i55^5555555 ID O O O CM O o CD O 00 O SUOI^JOdOJd P9U!B;qO Uj QDUQJdHlQ

PAGE 79

following the application of the anticatecholaminergic agent, phenoxybenzamine, to the septal region (N=9) and to the lateral ventricle (N=4) compared with the distribution obtained following the application of saline (N=18) to the septal region. The septal and intraventricular effects of the application of phenoxybenz amine closely parallel each other with the exception of social investigation (N) and the aggressive attack (A) . Although few differences are to be found in the direction of the effects caused by phenoxybenzamine at these two sites, there appear to be marked quantitative differences. Intraventricular placements produced a strong increase in exploration (E) and hoarding (UA;HA;UO;HO) , but septal placement had only a moderate effect. Septal placements had a strong effect on biting (B) and ventricular placements, little or none. The remaining effects on defense, fighting, escape, submission, and other categories are both qualitatively and quantatively similar. When the norepinephrine and phenoxybenzamine groups are compared with the S group, as shown in Figure 5, it is clear that these two agents generally produce opposite effects on the behavior of the hamster. Phenoxybenzamine produces an increase in exploration, while norepinephrine produces a considerable reduction in exploration. Norepinephrine may also have an opposite effect on hoarding behavior. Phenoxybenzamine produces an increase in social investigatory behavior, while norepinephrine, if it has an effect, may produce a reduction in this behavior. Aggression scores increase following the administration of phenoxybenzamine while the application of norepinephrine produces a reduction of aggressive behaviors. Defense behav-

PAGE 80

I cd * o pi o ON CL) 4-1 (U oo H s II CO • (1) pi VO PI •rl II •rl rH CO ^ II •H rQ Pu (U 0) 13 .3 U Q) •H o 4-1 i G rs CO > a> O t-i o pi CO CO (U •H o •rl u •rl o Pj rH o /— \ rH 00 rH II II O 0) — ' pi CO •rl ft >-l o •H O ft •P (U bO Pi •rl •rl 0) ft Pi 0) +J •rl t-l CO rH o •rl CO 13 CO QJ (U rH <} +J 4-J vO •* c ^ o CO 13 4-1 nj +J Cfl 13 0) J3 >C) 4-1 13 0) CO 13 > (U S a cO 0) CO a) CO )^ p, Q 0) pi M-l o CO i u (U •rl 60 CO •a a o ft II CO H (U >-) (U <4-l m PI o •H (U M N 60 a •H 1

PAGE 81

73 suo!;jocjojcl pauie^qo ui sDusja^^iQ

PAGE 82

74 lor is reduced by phenoxybenzamine, but norepinephrine increases the frequency with which these behaviors are observed. Fighting may be unaffected by phenoxybenzamine and norepinephrine, but there is some suggestion that norepinephrine produces a reduction in fighting. Escape behaviors are reduced by both norepinephrine and phenoxybenzamine. Norepinephrine produces a dramatic increase in submissive postures, attending, and vocalization. Phenoxybenzamine does not appear to have an effect on these behaviors. Figure 6 represents the results of the f ornix-resection group (N=4) . The distributions were obtained from trials run with and without the application of phenoxybenzamine to the septal region. Both distributions are compared with that of the phenoxybenzamine group which is represented by the baseline. The direction of each effect is the same for both the fornix resection condition and the fornix resection-phenoxybenzamine condition. The presence or absence of phenoxybenzamine alters only the magnitude of the effect. The presence of phenoxybenzamine reduces the magnitude of the increase in exploration produced by fornix resection; however, either with or without phenoxybenzamine, resection of the fornix produces an increase in exploration which exceeds that seen following the application of phenoxybenzamine to the septal region of the intact animal (see Figure 4). Fornix resection reduces hoarding to zero. Phenoxybenzamine alleviates this deficit slightly. Both groups are lower in terms of their social investigatory behavior (N;N'). Phenoxybenzamine produces an increase in the aggressiveness of the intact animal (s ee Figure 5) , while fornix resection alone produces an increment in aggressiveness which approximates that seen in the phenoxybenzamine treated intact

PAGE 83

Q) O •rl *^ c 03 &^ X O (U o !-< (0 c 60 o J3 (U O J3 a 4-1 ft o to 03 •H Xi 0) • A JJ •p 03 a^ O C CO 0) 13 O CM 03 C ft 0) n) 03 II VI (U C Vi a> o 13 •H •H 14-1 O C 60 O iH (U CO • O VI VI 03 <+-l 0) C iH rH O CO 0) e •H •H U 2 4-1 VI ft O 4J 0) (U y-i 03 rH 03 0) o Cd 0) (U Q) 4-1 Vi Vi m o C 4-1 4J 4J o •H •H O (U C •U u J3 Vi H VI H O (3 O O P< c O • • r> V4 VI 0) ON O a. OJ 13 M-4 N •H >£> C .H O (U 0) II 3 03 Vi CO cd (U (3 (3 03 o O •H CO H (3 0) B IS 4J ^ S 3 4-1 N 13 J3 ft c O •H >^ (U •H Vi M-l ,£3 4J 4J O •H CO 13 !«! 13 •H (3 0) O CJ 13 O 4-1 C o •H 13 (U u 0) 4J 0) J3 J3 cd 03 ft 0) o (U 13 •rl V< r-l o rH ft cd rH 0) ft (U 4-> CO 0) ft VI ft & n) 0) 13 03 03 O 0) 0) •H •H J3 4-> 4-1 J3 4J U 03 a •H 4J •rl 4-1 15 II O a 4-1 Iz; (U U •H T3 V — ^ 03 O T3 13 QJ M-l n) ft VI 0) 3 0) J3 O J3 H +j VI •H H •H 60 13 IS V4 0) O vO 13 |3 |i< 0) •H s-^ in 0) 13 CO M •H ro r-cd N 60 4-1 d vo rH •H (U O J2 CM II

PAGE 84

76 00 ID O O TC\J o CVJ ^ o o T r 8 00 0 0 c 1 I 9) c o < c o +-> u 4; in q: >< 'c (o ll. (D c E OJ N C u ISI r .vwvwwwvvvwvvwvvv H .xvwxxxxx-v> 00 1 o I JO wXXXXXXX < I < 00 xN CO ^ CD O O C\J o CM o o ID o 00 a o c suoi^jodojd paujE^qo ui SDuaja^jiQ

PAGE 85

animal. The treatment of the fornix resection group with phenoxybenzamine produces a further increment in aggression (Figure 6) . Phenoxybenzamine partially cancels the depression in defensive behavior seen following fornix resection. Phenoxybenzamine, in combination with fornix resection increased the incidence of fighting more than did fornix resection alone. The remainder of the two response profiles are identical in their deviation from the response profile of the phenoxybenzamine group. In Figure 7 the distribution of proportions of the amygdalectomyphenoxyb en z amine group (N=4) is compared with that of the phenoxybenzamine group . The treated group showed increases in both exploration and hoarding, while all of the social categories were depressed. The most marked depression was found in the social investigatory (N) and fighting (B) categories . Figure 8 compares the effects of the administration of norepinephrine and the effects of the administration of carbachol with the distribution of proportions of the S group. Norepinephrine reduces exploration, while carbachol produces a large increase in the category. Norepinephrine and carbachol both may somewhat reduce hoarding. Carbachol produces a general reduction in social behavior as may be seen from an examination of the remaining categories in Figure 8. The norepinephrine profile does not show this consistency. There is a marked increase in the use of submissive postures (T;T';Z) and the occurrence of vocalization in the norepinephrine group. Figure 9 compares the distributions of the atropine and phenoxybenzamine groups with that of the S group. Atropine and phenoxybenzamine produce opposite effects on exploration, but have little effect

PAGE 86

-l 1^ bO ** ine o n (1) p< •2 3 N g O C (1) c 60 to O •4J •H c o CO J2 N & (U 1 0) H • o O c i 4-1 o o t)0 C o •u QJ •H S3 d 4-1 Qi •H 3 CO :3 0) >-l N u (X d u (U
PAGE 87

suoj^jodojd p9U!e;qo ui 9duqjq^\q

PAGE 88

iH XI o3 4-* •W O TP 4-* 111 *H H IT* Cg II AT •H rH O i-H CJ iH o CO cd II XI X (U 4-J >-< bO >^ X cu P tC 'U •H (D }-i o +-J r-< rC M P* CD M O >^ 0 •H +J / — V ^4 00 ' ^ O r— 1 r— 1 PII o It (I P4 O 0 0 CO >-l tiO o ^ 1—i •H 0 i-H fi3 cd U •H CO cd P cu J-i CO Co •H 0 l-l J-J CO X! o +J CO cd s (U to Xi CO CO > a) u 0 -i 44 n 00 0) 0 C 11 0) •H }-l P< P< 3 0) 3 00 ^ 0 •H 0 1 M pJ^ c

PAGE 89

81 CVJOOOO I ooo O rr*5^PT r r 1 1 i 1 1 CD 4) C 'l_ u E m a n
PAGE 90

1 0^ b CO H o cd CO r-l II U cu 3 4-1 O J-i o bO ,o -l u &i >i o (U X p4 O o 3 >-( ^ N a) o< GO rS3 r-l PM II o • ' ' CO r-l p. 00 2 •rl o II 4-1 3 bO ft 3 •rl (L) o U |3 •rl bO CO rH •rl td cu 13 CO •rl d) ft ^ o 4-1 u 4-1 ?i M-t (u O ^ ^ 3 4-1 cx) •W td CO (U 4-1 CO a> S CO CJ Cd 2 Oi CO u ft a) 3 0) <4-l O > U-l »-i •rl C>0 cu X) CO 43 2i o II H CO • ' dl CO 3 o> c o •H ft
PAGE 91

83 -

PAGE 92

84 on hoarding. Atropine reduces social investigatory behavior while phenoxybenzamine has an opposite, but equal, effect. Atropine uniformly increases aggressive behavior (A;A';B;C). Although phenoxybenzamine appears to increase biting (B) , it appears to have little effect on the remaining aggressive behaviors. The direction of the effects of atropine and phenoxybenzamine is similar throughout the rest of the profile although the magnitude of the effect on sparring (S) and attending (I) is larger in the atropine group. Histological results . The results of histological examination of the brains of the Ss in the study showed the cannulae to have been well-placed. The majority of the placements were in the midline and within .75 mm. rostral to the decussation of the anterior commissure. Some variability in the placements was intended, therefore, a precise statement cannot be made regarding some specific point within the septal region at which the tips of the cannulae were found. Moreover, it is doubtful that such a statement would be useful since the extent of the action of the chemicals in unknown. If can be said that if the chemical exerted an effect for a radius of 1.5 mm. around the tips of the cannulae, it is likely that most of the septal region was affected. Figure 10 shows two sections which were cut at the level at which the cannulae penetrated into the lateral ventricle. It also shows the sections from two S^s which had more rostral septal placements. Figure 11 shows another two rostral placements. Subject-3N had the tip of the cannula placed in the anterior limbic cortex rostral to the septal region, while S_-11N had its cannula projected into the midline approximately midway in the rostrocaudal extent of

PAGE 93

§ o u 00 •H

PAGE 95

n O CO (d u •H CD u n o u I M-l I O CO 0) 80) •H

PAGE 97

the septal region. This placement was slightly anterior in the most typical placements. Subject-15P is a very representative animal. The brain of S^15S is included here as an example of a very extreme placement. The cannula was found to be in the lateral septal nucleus. Figure 12 also shows some of the variability in the sites at which the tips of the cannulae could be found. In S;-15A the cannula was found to be slightly off the midline but within the medial septal nucleus. Subject-29N had its cannula placed dorsal to, and at the level of, the decussation of the anterior commissure. Subject-9PA was found to have its cannula in the posterior septal region among the fibers of the columns of the post-commissural fornix. In S^-27P the cannula was placed in the most caudal part of the medial septal region. Figure 13 shows two atypical placements. In S^-9N the cannula had been directed at the septal region, however, it was found to have been placed caudal to the septal region in a position from which the structures adjoining the third ventricle and the bed nucleus of the stria terminalis could have been affected by the chemical. This showed a behavior pattern which contrasted sharply with the other S^s in the norepinephrine group. The results from this are discussed in Appendix A. The remaining atypical S_ had the cannula placed in the hippocampus. This ^ was supposed to have been a ventricular animal. Although the cannula was not in the ventricle, there was little basis for distinguishing the data from this from those of the remaining intraventricular S^s. As no clear behavioral difference was observed, this was not given separate treatment in Appendix A. The remaining two S^s in Figure 13 represent the lesion groups. The fornix lesion was relatively complete. No damage was seen in the

PAGE 102

94 hippocampus or septal region. There was damage to the corpus calloseum and cingulate cortex, and minor thalamic damage. This damage, however, did not extend for any distance to the rostrocaudal plane. This is somewhat atypical in that small sections of the fimbria appear to be intact and normal. This may mean that efferent fibers from the posteroventral hippocampus to the lateral septal region may not have been damaged. In the amygdaloid Ss, represented here by S-23PA, bilateral damage was found in the pyriform cortex and in the anterior amygdaloid area. There was also some damage to the corticomedial amygdala. Bilateral damage to the lateral and basal amygdala was very slight, although in every case there was some extensive unilateral damage to these regions. Summary of results . Figure 14 presents a summary of the effects obtained in the study. The behaviors have been collapsed into nine categories. All of the comparisons in the figure were made using the saline group as a reference. The pluses and minuses represent the percentage of difference between the saline group and each other group in every category. The first 33 per cent was always ignored, therefore differences smaller than 34 per cent are not represented. This was done in an attempt to control out some of the differences which may not have been due to the treatments. Exploration was increased by carbachol, but was not affected by atropine. It was slightly decreased by norepinephrine placed in the septal region. Phenoxybenz amine had no effect on exploration when placed in the septal region, but had a slight positive effect when placed in the lateral ventricle. A fornix lesion alone produced a substantial increase in exploration whereas the addition of phenoxybenzamine

PAGE 103

Figure 14. A sunmary of the effects obtained in the study. All groups are compared with the saline group. Differences of less than 34% were ignored. In considering the data presented in this figure, it should be noted that the mean number of responses observed for each. in each group was as follow^: Saline, 199; Carb., 188; NE Sept., 194; NE Vent., 175; Phen. Sept., 193; Phen. Vent., 172; Fornix, 175; Amyg. , 78.

PAGE 104

96 EXPLORE HOARD SOC. INV. AGGRESS DEFENSE SUBPOSTUREl ESCAPE FIGHT MISC. Saline .30 .0 6 .10 .05 .13 .02 .0 6 .1 5 .1 2 Carb. NE Sept .17 I NE Vent. W Atrop. Phen. Sept. H of Phen. Vent. Phen & -omix -esbn Fornix -esion Alone Phen. & Annyg. jesion nn .11 .091 101 1^ >66<' 5J oj oj Ipi 131 101 06 01 * > 120°/o >99<'/o mm >-66°/o > 3000/0 >-33°/c^H >-99°/o V Vocalization

PAGE 105

97 reduced this effect to a more moderate level. Phenoxybenz amine in combination with amygdalectomy produced a strong positive effect on exploration. Hoarding was slightly depressed in the carbachol group and unaffected by norepinephrine and intraventricular norepinephrine. Septal phenoxybenz amine produced a slight increase in hoarding, but intraventricular phenoxybenzamine produced a much stronger positive effect. Septal phenoxybenzamine in combination with amygdalectomy also produced a strong positive effect. Section of the fornix reduced hoarding to zero-; however, the addition of phenoxybenzamine alleviated the deficit slightly. Social investagory behavior was depressed by both carbachol and atropine. Norepinephrine had no effect on social investigatory behavior, but phenoxybenzamine had a moderate positive effect. Phenoxybenzamine had a moderate positive effect on social investigation if placed in the septal region but no effect when placed in the ventricle. Fornix resection depressed these behaviors, but the addition of septal phenoxybenzamine canceled this effect. When septal phenoxybenzamine was combined with amygdalectomy, social investigatory behavior was depressed. Aggression was moderately depressed by carbachol while atropine had a strong positive effect on these behaviors. Norepinephrine, both septal and intraventricular, had a moderate depressive effect on aggres sive behavior. Phenoxybenzamine had a strong positive effect only in the septal region; intraventricularly it was without an effect. Fornix resection produced a strong positive effect, but this was increased by

PAGE 106

98 the addition of septally placed phenoxybenzamine. Aggression was found to be moderately depressed in the phenoxybenzamine amygdalectomy group. Defense postures were depressed by treatment with phenoxybenzamine. Intraventricular phenoxybenzamine did not have an effect. Phenoxybenzamine placed in the septal region of the S^s in the fornix resection group had a slight positive effect when compared with the fornix resection alone condition. The amygdalectomy phenoxybenzamine group showed a similar depression. The incidence of submissive postures was increased by the application of norepinephrine of the septal region. A lesser, but still strong, effect was obtained from the intraventricular group. Atropine and phenoxybenzamine both decreased the frequency of submissive posturing, as did fornix resection, fornix resection with septal phenoxybenzamine, and amygdalectomy in conjunction with septal phenoxybenzamine . Only atropine, which had no effect, did not depress escape behavior. The effect of norepinephrine was strongest in the intraventricular group, while the effect of phenoxybenzamine was more pronounced in the septal group. Only intraventricular norepinephrine increased fighting. Phenoxybenzamine in the septal group and in the fornix resection group produced a borderline depression of fighting. Fornix resection alone reduced fighting as did the treatment of the amygdalectomy group with phenoxybenzamine . The notable change in the miscellaneous category was the increase in vocalization associated with the application of norepinephrine to the septal region.

PAGE 107

99 The results of Kruskal-Wallis tests for overall comparisons among all the groups for each response category are presented in Appendix B. For every category, the overall test was found to be significant at beyond the .02 level.

PAGE 108

DISCUSSION The saline control group was found to be slightly submissive when compared to their normal opponents (See Fig. 1). The differences were small, however, and the bias produced by them was probably of little significance. The interrelatedness of the response categories makes an interpretation of the outcome of the study difficult. It is always possible that an increase in one category could result in a decrease in another, even though the treatment had no effect on that particular category. It would be wise to consider the extent of the interrelatedness of the response categories. The average number of responses per trial observed in the study did not exceed 60 in any group. Experience has shown that the hamster is capable of generating more than 120 responses per trial. It is not likely, therefore, that the depression in any category can be considered to be a sole function of an increase in responding in another category. For example, Figure 10 indicates the carbachol group explored more than the saline group. It could be argued that the depression in the remaining categories was the result of an increase in exploration; that the animal did not hoard because it was exploring. However, reference to the amygdalectomy-phenoxybenzamine group on the same figure shows that these animals also increased their exploration, but in addition, effected a substantial increase in hoarding behavior. The intraventricular phenoxybenzamine group increased their exploration and hoarding with almost no concomitant reduction in any other category. 100 -

PAGE 109

101 The animals do, therefore, have the freedom to increase or decrease their behavior in any given category without being required to modify their behavior in any other category. Dominant animals are distinguished from submissive animals on the basis of differential characteristics in their response profiles. Dominant animals show more aggressive behavior and submissive animals show more submissive postures and escape behavior. The logic of this can be extended to other categories . In a group where the dominant animals were very aggressive, more fighting would be expected. The social investigatory category might be differentially affected in dominant and submissive animals. Social investigation requires contact with another animal and submissive hamsters are reluctant to initiate contact with their dominant opponents. Attending (I) is also affected since this is an orienting response toward the other animal. Submissive animals spend a good deal of time "tracking" the behavior of their opponents and therefore usually score higher in this category than do dominant animals which have less reason to be alert as to the intentions of their submissive cagemates. Before the completion of the present study, the interrelationships considered above were thought to represent a fairly accurate assessment of the dependencies among the various responses categories . The presence of these dependencies suggested that caution be used in the interpretation of the results of ablation studies. When septal ablation was found to increase the aggressiveness of the naive hamster and these animals also showed a decrease in exploratory activity (Sodetz, 1965) this decrease was not considered to be one of the direct consequences of ablation. Decreased exploration had to be

PAGE 110

102 considered to be an indirect effect produced by the concentration of the activity of the animal in the aggression category. Figure 10 suggests that this interpretation may have been too conservative. The phenoxybenz amine group increased its aggression without decreasing its exploratory behavior. The fornix resection group increased its aggression and its exploratory activity, both with and without phenoxybenzamine treatment. The intraventricular norepinephrine group decreased its aggression without diminishing its exploratory activity and the septal norepinephrine group showed a decrease in both categories. Figure 10 shows that some groups which decreased their aggression did not decrease their social investigatory behavior. This is not a common result for a normal animal, but it is consistent with the response profile of septal hamsters (Sodetz, 1965). More aggressive animals might be expected to show more social investigatory behavior, but this is true of only one of the four groups which were found to be more aggressive as a consequence of chemical treatment or lesion. The combination of phenoxybenzamine and fornix resection produced no change in social investigatory activity while the fornix resection alone and treatment with atropine both produced substantial reductions in this behavior. The question remains as to what categories are invariably related. Vocalization (V) appears to be closely correlated with submissive behavior in the normal animal. As a general rule, it is heard in response to a bite (B) , yet here also there is a contradiction. Norepinephrine when placed in the septal region (see Figure 10) produced an increase in submissive posturing which exceeded 500 per cent over that seen in the saline group. In this group vocalization (V) rose

PAGE 111

103 sharply. These animals, however, did not engage in any more fights than did the saline control S^s and they actually engaged in far fewer fights than did the group which received norepinephrine in the ventricle. These data suggest that it was the treatment with norepinephrine which directly produced an increase in vocalization and that this increase was not simply a concomitant change associated with some other response category. Nowhere in Figure 10 is there any evidence that two categories are negatively or positively correlated across all treatment groups with the exception of aggression and submissive posturing. In each case where aggression was seen to be enhanced by a treatment, submissive posturing was found to decrease. In every case where submissive posturing was observed to increase, aggression was found to be depressed. This applied to all groups except the amygdalectomy-phenoxybenzamine group which was depressed in every social category. This effect is also seen following only amygdalectomy in the naive hamster (Bunnell, et al., 1965). It would seem reasonable to conclude that these two behavior categories, aggression and submission, are negatively correlated and that the effects of treatment may be difficult to dissociate from those produced by their inherent mutual exclusiveness . A possible solution to this problem may be found through a consideration of the differences obtained in the treatments with the stimulating agents and their antagonists. If the logic of chemical stimulation and blocking is correct, the effects of blocking should generally be opposite to those of stimulating. If stimulation with phenoxybenzamine produces an increase in aggression, as it did in this study (see Figure 10) , a decrease in aggression would be expected

PAGE 112

104 following the application of a catecholaminergic agent. While this occurs, there is some suggestion that a decrease in aggression is necessary to an increase in submission, but it not sufficient for such an increase. In this regard, see the amygdalectomy-phenoxybenzamine group and the carbachol group in Figure 10. In these groups, aggression was decreased, but there was either no change, or a reduction in submissive posturing. All of this suggests that treatment with phenoxybenz amine only increases aggression while norepinephrine only increases submissive behavior. This catecholaminergic system may then be characterized as one that controls which class of behaviors will be displayed and it does not appear to simply determine the amount of a single behavior which is to be released. This is to say that blocking does not disinhibit aggressive behavior while stimulation merely suppresses it. What is suggested here is that stimulation releases submissive behavior, while blocking releases aggressive behavior. The nature of this effect can be further evaluated if the lesion groups in Figure 10 are considered. Aggressive behavior is released by transection of the dorsal fornix. A still more marked release is obtained if phenoxybenz amine is introduced into the septal region. The release of aggressive behavior by phenoxybenzamine is blocked by small bilateral lesions of the amygdala. The fact that destruction of the fornix produces an effect which is very similar to that obtained through blocking a catecholaminergic system in the septal region suggests that this system is served by a mechanism which reaches the septal region via the fornix. The only two possibilities then are fibers from the hippocampus and fibers from

PAGE 113

-1Q5 the cingulate gyrus. The hippocampus Is the more likely choice since many of the cingulate fibers would be entering the fornix somewhat rostral to the lesion. An efferent system from the septal region to the hippocampus is precluded by the fact that the application of phenoxybenzamine potentiates the effect in animals in which the outflow from the septal region to the hippocampus has been destroyed by fornix resection. The amygdala cannot be considered to be critical to the motor display of aggressive behavior as experienced hamsters may still display aggressive behavior following amygdalectomy (Bunnell, et al. , 1965) . Since all of the social categories were depressed in the amygdalectomy-phenoxybenzamine group it seems more likely that this structure contributes some sort of more general motivational or perceptual function, the interruption of which, in the socially naive animal, results in a failure to respond to social stimuli. Intraventricular norepinephrine and phenoxybenzamine closely parallel the effects of the septal placement of these agents. While a site of action for intraventricularly administered drugs is impossible to determine without the use of a method designed for this purpose, it is possible to speculate that either the hippocampus or the lateral septal nuclei are being affected in this group. However, a consistent picture is not presented and therefore the possibility of any number of structures being affected cannot be ruled out. The remaining categories may now be considered with respect to a catecholaminergic system. Fornix resection alone produces a large increment in exploration. Stimulation with norepinephrine reduces exploration. These data are again consistent with a hippocampal-septal

PAGE 114

106 catecholaminergic system; however, phenoxybenzamine did not produce an increase in exploration large enough to exceed the 33 per cent cut off imposed on Figure 10, although intraventricular phenoxybenzamine just made the cut off. In this case an additional problem must be considered. Miller (1965) has shown the dose response curves for chemical stimulation of the brain to be inverted "U" shaped functions. This fact points to the need for the use of at least three widely separated dose levels if one is to specify these functions. More importantly, his work suggests that even though the dosage used may exceed the physiological concentrations present in an area, still greater doses may be required to produce a maximal effect and different doses may be required for different agents. The present study used the same dose level for all of the treatments, therefore it is possible that septal phenoxybenzamine did not have a strong effect on exploration because the dose used was of insufficient strength. If this was the case, then the picture is more consistent with a catecholaminergic hippocampal-septal system which exerts inhibitory control over exploratory activity. The amygdala must also exert some suppressive effect on this behavior for, as can be seen in Figure 10, amygdalectomy in combination with phenoxybenzamine has as strong a positive effect on exploration as does transection of the fornix. Hoarding is increased by amygdalectomy in combination with septal phenoxybenzamine. There had been some earlier suggestion of an increase in hoarding with amygdalectomy alone (Bunnell, et al., 1965); however, it had not been possible to dissociate this effect as independent of the depression in social behavior which resulted from the

PAGE 115

-107 lesion. In the present study it has been demonstrated that an increase in hoarding does not systematically follow a decrease in social behavior. It may be reasonable to assume that the effect on hoarding is independent of other behavioral changes. There is some suggestion that hoarding behavior is dependent upon mechanisms which pass through the septal region, but do not terminate there. Septal ablation reduces the hoarding behavior of the hamster by a considerable margin. In cases of near total ablation, this behavior is not seen. In the present study, transection of the fornix alone reduced hoarding to zero, whereas the functional equivalent of septal ablation, the administration of phenoxybenzamine , had a slight positive effect. The administration of phenoxybenzamine had only a very slight beneficial effect on the deficit produced by transection of the fornix. This pattern does not appear to be consistent with the hypothesis that the septal region as treated in the present study has a critical role in the mediation of hoarding behavior. The effect of norepinephrine on hoarding was to bring about a marginal decrease, further suggesting some sort of very subtle inhibitory role and militating against a depressive effect of septal ablation. The subtle facilitatory effect of septal phenoxybenzamine might be reflected in the unpublished observation that hamsters with small anterior septal lesions engage in excessive manipulation of nesting materials to the extent of shredding them into very small pieces , whereas large lesions which invade the posterior septal region, and presumably do more damage to post-commis sural fornix fibers coursing through this region, have been seen to reduce this manipulation of nesting materials and hoarding to zero.

PAGE 116

108 Only transection of the fornix and amygdalectomy were observed to change defensive behavior although norepinephrine produced a borderline Increase In this category and septal phenoxybenzamlne produced a borderline reduction. More fighting was observed In the group treated Intraventrlcularly with norepinephrine. As was stated above, it was not possible to determine the neural locus of this effect. Fornix lesions alone reduced fighting as did the combination of amygdalectomy and phenoxybenzamlne. Septal phenoxybenzamlne and intraventricular phenoxybenzamlne as well as fornix resection in combination with phenoxybenzamlne produced borderline reductions in fighting. These effects are suspect, however, because septal norepinephrine produced no change in fighting behavior. It seems likely (See Figure 10) that no septal catecholamlnergic mechanism is Involved in this behavior. The results of amygdalectomy and of intraventricular ly administered norepinephrine suggest some other site of action. The miscellaneous category consisted of grooming (G) , vocalization (V) , and attending (I) . Two results in this category are of special Interest. Septal norepinephrine produced an Increase in vocalization which exceeded seven per cent of all the responses seen in this group. The figure is higher than any seen either in normal hamsters, hippocampectomlzed hamsters (Bunnell, et al., 1967), septal hamsters (Sodetz, 1965; Sodetz and Bunnell, 1967.a; 1967b), or amygdaloid hamsters (Bunnell, et al., 1965). Although no reduction in vocalization was obtained in the phenoxybenzamlne group this cannot be considered unusual for the Incidence of this response is very low in aggressive hamsters and only slightly higher in submissive

PAGE 117

m animals. Vocalization is most often heard in response to a bite; however, the norepinephrine animals were heard to vocalize intermittently whenever their dominant opponents approached them. This may suggest heightened fear or emotionality on the part of these animals. Attending (I) is the other behavior of interest in this category. Section of the fornix produced a reduction in attending which amounted to five per cent of the response total of that group. This reduction was unaltered by the application of phenoxybenzamine. Behaviorally, attending simply means that the animal orients toward its opponent. While it is not unusual to find that aggressive animals, such as those in this group, are reduced in this category when compared with normal animals , these data indicate that this response almost never occurred. It would not be unreasonable to suspect that this orienting response is functionally equivalent to the orienting response to other stimuli. If Douglas (1967) is correct in his contention that hippocampal theta reflects the inoperation of a sensory gating system, the results related to attending may be significant in explaining changes in the exploratory category. Petsche, et al. (1962) have shown that transection of the fornix eliminates hippocampal theta rhythm as does treatment of the septal region with scopalamine. If this is correct, both the atropine group and the fornix resection group should be lacking hippocampal theta rhythm. If attending is a specific orienting response directed at a particular stimulus, then according to Douglas's (1967) model the loss of hippocampal theta should produce an increase in the incedent of this response. This is exactly the result obtained in the atropine group. Even though these animals were

PAGE 118

110 aggressive, they showed an increase in attending equal to some four per cent of their response total. In the carbachol group in which the predicted incidence of theta would have been higher due to cholinergic stimulation of the septal pacemaker cells , the animal would be expected to be unusually responsive to a wide range of stimuli. That this occurred is suggested by the fact that these animals attended no less than the saline controls but did increase their exploratory behavior by 23 per cent of their response total. These data are represented in Figure 2. In the group in which the fornix had been transected, no hippocampal theta could have occurred; therefore, a reduction in the range of stimuli to which they would have attended should have resulted. This does not seem to have been the case, however, for these animals decreased their attending while increasing their exploration. The lesion data are not consistent with the cholinergic treatment data. It may be that the carbachol and atropine groups are functionally closer to normal than the fornix group and that transection of the fornix has deprived these animals of more than the cholinergic mechanism known to underlie the theta rhythm. What is suggested here is that transection of the fornix produces a narrowing in the organism's sensitivity to stimuli, but unlike atropine blockage, it may produce an additional deficit which might be related to the decision mechanism which determines which aspects of the environment are significant to the animal. Even though the atropine group perseverated in their attention to their submissive opponents, it can be considered significant that they focused their attention on these animals rather than on less relevant stimuli. The carbachol group distributed its attention between cagemates

PAGE 119

Ill and the non-social objects in the environment. The fornix group perseverated in their attention to non-social stimuli, as evidenced by the duration of their exploration which amounted to 75 per cent of all the time they were observed. Yet, these animals did nothing with the objects to which they paid so much attention. For example, they neither picked up food (UO) and nesting materials (UA) nor did they hoard (HO;HA) them. The differences between these groups may reflect the loss, in the fornix group, of some mechanism which selects stimuli relevant to some internal state or to the completion of some task. It is suggested that this system may arise in the hippocampus and, via the fornix, affect some other region of the brain. Further, it is possible that the hoarding deficit observed in the fornix group resulted from damage to this system. The results of the study implicated a cholinergic mechanism in the control of the level of aggression displayed by a given animal. As was noted above, the catecholaminergic system appeared to control the character of the behavior displayed, that is, whether a given animal would behave aggressively or submissively. This conclusion was reached in light of the evidence that a decrease in aggressiveness was not necessarily correlated with an increase in submissive posturing. The carbachol group, and the amygdalectomy-phenoxybenzamine group in Figure 10, both showed evidence of a decline in aggressiveness without an increase in submissive behavior. If these two categories were on a continuum they would be expected to be consistently negatively correlated. The suggestion has been made that the catecholaminergic system determined which of these categories of

PAGE 120

112 behavior would be displayed. It is clear, however, that aggressiveness and submissiveness may vary in their intensity. Some mechanism must control the amount of these behaviors which will be displayed once a catecholaminergic selection of a response category has been made. The data suggest that it is a cholinergic system which has this function. Carbachol serves to reduce aggressive responding while atropine increases only aggressive activity. The escape category (see Figure 10) presents a special problem. The animals of every treatment group except atropine were depressed in this category. No group showed an increase in this behavior. Sodetz (1965) and Sodetz and Bunnell (1967a; 1967b) have found that septal ablation produces an increase in escape behavior. However, this effect is seen only in animals which have had preoperative experience with defeat. All of the animals in the present study were socially naive; therefore, they are not comparable to those which were seen to increase their escape behavior following septal ablation. The results of the present study may now be compared with those of studies which have ablated the septal region (Sodetz, 1965; Sodetz and Bunnell, 1967a; 1967b), amygdala (Bunnell, et al., 1965), and hippocampus (Bunnell, et al., 1967). Amygdalectomy , in combination with phenoxybenz amine, produced an increase in non-social behavior and a decrease in all social behaviors. This effect has been noted before following amygdalectomy alone. Bunnell, et al. (1965) were unable to say with certainty that amygdalectomy produced an increase in non-social behavior. The

PAGE 121

113 present study was not limited in this regard since it demonstrated the absence of a strong negative correlation between non-social and social behavior. A difficulty arises, however, from the fact that the amygdaloid animals in this study were run with the presence of phenoxybenz amine in the septal region. If this chemical had an effect in addition to that contributed by amygdalectomy it served to potentiate the ablation effect. The effect of amygdalectomy alone cannot be determined with certainty from the data of the present study. Septal ablation has been shown to generally decrease non-social behavior and to increase social behavior (Sodetz, 1965). The nonsocial categories are exploration and hoarding. The effect of septal ablation on hoarding can be accounted for by damage to fibers passing through this region and evidence of a chemically produced effect is not required. The chemical data suggest that anticatecholaminergic treatment may slightly increase hoarding behavior and that cholinergic stimulation may depress hoarding. These effects would be masked by a septal lesion which included extensive damage to the post-commissural fornix. The effect of septal ablation on exploration does not find a parallel in the present study. Septal ablation reduces exploration or, at the very least, leaves it unchanged. The blocking agents used in the present study had no effect on exploration. Transection of the fornix produced a marked Increase in exploration which was attenuated somewhat by the addition of septal phenoxybenzamine. These data are contradictory to the lesion data obtained under similar circumstances and suggest that a further evaluation of the relationship of the septal region to exploratory behavior may be necessary.

PAGE 122

114 The effects of septal ablation on aggressiveness in the socially naive and preoperatlvely dominant hamster correlate very well with the results of the present study. Catecholaminergic blocking resulted in the display of aggressive behavior. Catecholaminergic stimulation resulted in the display of submissive behavior. Atropine blockade produced an increase in the level of aggressiveness of the hamster, while cholinergic stimulation produced a decrease in aggressive behavior. These results are consistent with the observation that septal ablation in the naive hamster produces a high level of aggressive responding (Sodetz, 1965). Many of the behavioral changes seen following septal ablation can be observed in animals with hippocampal lesions (Bunnell, et al., 1967). However, rather than becoming aggressive, the hippocampectomized hamster is more likely to become submissive. When the response profiles of submissive hippocampals are compared with those of submissive septals, they are found to be nearly identical. However, these data are difficult to interpret in terms of the present study in which all of the animals were socially naive. The preoperative social history of the septal hamster determines the nature of the ablation effect. Submissive animals become more submissive following surgery, while dominant animals become even more aggressive. No experienced animals were used in the study; therefore, no animal's behavior was characteristic of that of the submissive septal. It is possible to suggest that the catecholaminergic system which mediates the display of submissive behavior does not course through the fornix. Transection of the fibers of the fornix results

PAGE 123

115 in the display of aggressive behavior. Presumably this effect can be mediated via the inhibition of a catcholaminergic system. The naive septal hamster behaves like a fornix lesioned hamster and it is reasonable to assume that blocking, transection of the fornix, and septal ablation are all interfering with the same system. Yet, the aggressive septal hamster, if faced by an equally aggressive septal, can be defeated. When this occurs, the behavior of the animal shifts from aggressive to submissive. If the system responsible for effecting the display of submissive behavior followed the same course as that underlying aggressive behavior, ablation, transection, and blockade, should all "lock" the animal into an aggressive response pattern because the mechanism eliciting submissive behavior would also have been destroyed. The cholinergic mechanism responsible for modifying the intensity of these behaviors would also have been damaged and the data of the present study suggest that the animal would be locked at a high level of whichever of the two patterns it was displaying. This is what happens with septal ablation. Naive and experienced dominant septals exhibit excessive aggression while experienced submissive septals are excessively submissive. Hippocampal hamsters are markedly submissive regardless of their preoperative history. While this is not true of every hippocampal hamster, it can be said that under no condition does a hippocampal animal respond more aggressively following surgery (Bunnell, et al., 1967). Hippocampectomy , therefore, affects the animal defferently than does transection of the fornix. If the entire mechanism underlying the display of aggressive and submissive behavior followed the fornix to the septal region, then no

PAGE 124

116 differences would be expected between hippocampectomy and transection of the fornix. The selective enhancement of submissive behavior seen following hippocampal damage may be related to the results obtained in the present study. The results suggest that a mechanism which inhibits aggressive behavior enters the septal region via the fornix. The hippocampal data are amenable to the hypothesis that the mechanism underlying the suppression of submissive behavior, though hippocampally related, reaches the septal region via some other route. This position is partially supported by the fact that intraventricularly placed norepinephrine produces nearly as powerful an effect on submissive behavior as does septal norepinephrine. Intraventricularly placed phenoxybenz amine has no effect on aggressive behavior. This suggests that the structures mediating these two classes of behavior are not equally accessible from the ventricle. The possibility that these two effects are obtained from fibers following two separate pathways to the septal region could explain the fact that septal hamsters were able to shift from aggressive to submissive behavior and that hippocampectomy and transection of the fornix elicit different behavior patterns which share a common characteristic excessiveness . The data of the present study suggest that at least three systems subserve the social and aggressive behavior of the hamster. One system can be characterized as subserving an attention focusing mechanism. This is a cholinergic system, the operation of which is best illustrated in exploration and attending. The second system is also cholinergic and it appears to govern the level at which aggres-

PAGE 125

117 slve behavior and perhaps submissive behavior will be displayed. The third system is catecholaminergic and appears to exert control over which of two behavioral categories, aggression and submission, will be displayed. Together these three mechanisms can account for much of the ablation data related to the social behavior and aggressiveness of the hamster.

PAGE 126

SUMMARY In earlier studies of the effects of septal ablation on the social behavior and aggressiveness of the hamster, it was found that the changes observed could not be accounted for in terms of a uni— process model. The present study was undertaken in an attempt to identify chemically distinct functional systems within the septal region of the forebrain through the application of cholinergic and catecholaminergic substances and their antagonists to this region of the brain. It was hoped that the identification of these systems would permit a more adequate assessment of the lesion data. All of the 112 adult male hamsters used in the study were raised in social isolation from weaning. They were individually housed in dominance testing cages constructed from double rat cages divided into three compartments of equal size. One S_ lived in each of the end compartments and the single center compartment served as a neutral area. Access to all compartments could be made available to both the experimental and its opponent by raising a guillotine door in each partition. All of the S^s were socially naive at the beginning of the study. In each pair, the experimental S_ had a double-walled cannula directed either into the septal region of the forebrain or into the lateral ventricle. In two groups the experimental S^s had either amygdaloid lesions or fornix lesions in addition to the cannula. Before the beginning of each trial, each experimental S_, depending upon the group to which it was assigned, received an injection of either 118 -

PAGE 127

119 norepinephrine, phenoxybenzamine, carbachol, or atropine sulphate. Following a ten-minute period, the trial was begun. Each trial consisted of five minutes of unrestricted interaction between members of a pair. Two trained observers, one for each animal, recorded the behavior of the S_s using a response inventory of 27 behaviors which can be observed in the interaction of two adult male hamsters. Concurrently, each written response was also entered into the record of a multi-channel event recorder. The output of the recorder was later collated with the written response record to yield data on response frequencies, durations, and sequences. All Ss were run one trial a day for five consecutive days. The results of the study indicated that both cholinergic and catecholaminergic systems were active in the social and aggressive behavior of the hamster. Catcholaminergic stimulation resulted in the display of a response pattern characteristic of submissive animals, while a catecholaminergic antagonist produced aggressive responding in another group of Ss. Cholinergic stimulation reduced the aggressiveness in one group but did not increase submissive responding. Atropine, the anticholinergic agent used in the study, increased the aggressieness of the group which was treated with it. Another cholinergic system was identified which appeared to be related to attention. The data from the fornix resection group suggested that this system was closely related to the hippocampus via the fornix. The present study also found that the hoarding deficit observed following septal ablation in the hamster was probably due to the

PAGE 128

120 interdiction of a pathway running from the hippocampus to some other part of the brain. The hoarding deficit produced by ablation is the result of damage to fibers passing through the septal region. The data of the group which was treated with phenoxybenzamine suggested that septal ablation, if it could be accomplished without damage to the columns of the fornix, might produce a slight facilitatory effect on hoarding. It was found that the effects of the application of cholinergic and catecholaminergic substances and their antagonists to the septal region of the forebrain on the social behavior and aggressiveness of the hamster were consistent with lesion data. The present study supported the view that the septal region has a multiplicity of functions in the control of the species-specific behavior of the hamster.

PAGE 129

REFERENCES Aghajanian, G. K. and Bloom, F. E. Localization of H-3 serotonin in rat brain by electron-microscopic autoradiography. J. pharm. exp . Therap . , 156:23-30, 1967. Akert, K. Diencephalon. In D. E. Sheer (Ed.), Electrical Stimulation of the Brain . Austin: Univ. of Texas Press, 1961, pp. 228-310. Akert, K. Comparative anatomy of frontal and thalamof rontal connections. In J. M. Warren and K. Akert (Eds.), The Frontogranular Cortex and Behavior . New York: McGraw-Hill, 1964. Allen, W. F. Fiber degeneration in Ammon's horn resulting from extirpation of the pyriform and other cortical areas and from transection of the horn at various levels. J. comp. Neurol. . 88:425-435, 1948. Anden, N. E., Dahlstrom, K. , Fuxe, K., Larsson, K. , Olson, L., and Ungerstedt, V. Ascending monoamine neurons to the telencephalon and diencephalon. Acta, physiol. Scand. . 67:313-326, 1966. Anderson, P. and Curtis, D. R. The excitation of thalamic neurons by acetylcholine. Acta, physiol. Scand. . 61:85-99, 1964. Andersson, B. and Larsson, S. Water and food intake and the inhibitory effect of amphetamine on drinking and eating before and after prefrontal lobotomy in dogs. Acta, physiol. Scand. . 38:22-30, 1957. Andy, 0. J. and Stephan, H. Septal nuclei in the soricidae (insectivores) . Cyto-architectonic study. J. comp. Neurol .. 117:251-274, 1961. Bard, P. Central nervous mechanisms for emotional behavior patterns in animals. Res, publ. ass, nerv. ment. Pis. . 19:190-218, 1939. Bard, P. A. and Mount castle, V. B. Some forebrain mechanisms involved in expression of rage with special reference to suppression of angry behavior. Res, publ. ass, nerv. ment. Pis .. 20:551-579. 1948. ~ Bertler, A. and Rosengren, E. Occurrence and distribution of catecholamines in brain. Acta, physiol. Scand. . 47:350-366, 1959. 121 -

PAGE 130

122 Bertler, A. Occurrence and localization of catecholamines in the human brain. Acta, physiol. Scand. . 51:97-107, 1961. Bignami, G. Pharmacologic influences on mating behavior in the male rat. Effects of d-amphet amine, LSD-25, strychnine, nicotine, and various anticholinergic agents. Psychopharm. . 10:44-58, 1964. Bodganski, D. F., Weissbach, H., and Udenfriend, S. The distribution of serotonin, 5 hydroxy tryptophan decarboxylase, and monoamine oxidase in brain. J. Neurochem. . 1:272-278, 1957. Bond, D. 0., Randt, C. T., Bidder, T. G., and Rowland, V. Posterior septal, fornical, and anterior thalamic lesions in the cat; vegetative and behavioral changes with anatomical and physiological correlations. Arch, neurol. Psychiat. . 78:143-162, 1957. Brady, J. V. The paleocortex and behavioral motivation. In H. F. Harlow and C. N. Woolsey (Eds.), Biological and Biochemical Bases of Behavior . Madison: Univ. of Wise. Press, 1958a, pp. 193-235. Brady, J. V. Temporal and emotional factors related to electrical self-stimulation of the brain. In H. H. Jasper (Ed.), Reticular Formation of the Brain . Boston: Little Brown and Co., 1958b, pp. 689-703.' Brady, J. V. Motivational-emotional factors and intra-cranial selfstimulation of the brain. In D. E. Sheer (Ed.), Electrical Stimulation of the Brain . Austin: Univ. of Texas Press, 1961, pp. 413-430. Brady, J. V. and Nauta, W. J. H. Subcortical mechanism in emotional behavior: affective changes following septal forebrain lesions in the albino rat. J. comp. physiol. Psychol. . 46:339-346, 1953. Brady, J. V. and Nauta, W. J. H. Subcortical mechanisms in emotional behavior: the duration of affective changes following septal and habenular lesions in the albino rat. J. comp. physiol. Psychol. . 48:412-420, 1955. Broca, P. Anatomie comparee des circonvolutions cerebrales. Le grand lobe limbique et la scissure limbique dans la serie des mammiferes. Rev. Anthrop. . 1:385-498, 1878. Brody, G. B. and Rosvold, H. E. Influence of prefrontal lobotomy on social interaction in a monkey group. Psychosomat. Med. . 14:406-415, 1952. Brown, D. L. and Slotnick, B. M. Discriminated fear conditioning and post shock hypersensitivity in mice with septal lesions. Paper read at the annual meeting of the Eastern Psychological Association. New York, April, 1966.

PAGE 131

123 Buddington, R. W. , King, F. A., and Roberts, C. Emotionality and conditioned avoidance responding in the squirrel monkey following septal injury. Psychon. Sci. , 8:195-196, 1967. Bunnell, B. N. Amygdaloid lesions and social dominance in the hooded rat. Psychon. Sci. , 6:93-94, 1966. Bunnell, B. N., Bemporad, J., and Flesher, C. F. Septal forebrain lesions and social dominance behavior in the hooded rat. Psychon. Sci. , 6:207-208, 1966. Bunnell, B. N., Burkett, E. E., and Jarrard, L. E. Hippocampal lesions and social behavior in the golden hamster. Paper read at the annual meeting of the Midwestern Psychological Association. Chicago, May, 1967. Bunnell, B. N., Friel, J., and Flesher, C. K. Effects of median cortical lesions on the sexual behavior of the male hamster. J. comp. physiol. Psychol. , 61:492, 1966. Bunnell, B. N. and Smith, M. Septal lesions and aggressiveness in the cotton rat. Psychon. Sci. . 6:207-208, 1966. ^Bunnell, B. N., Sodetz, F. J., and Shalloway, D. M. The effects of septal and amygdaloid lesions on the aggressiveness and social behavior of the hamster. Paper read at the annual meeting of the Psychonomics Society. Chicago, October, 1965. Burkett, E. E. and Bunnell, B. N. Septal lesions and retention of DRL performance in the rat. J. comp. physiol. Psychol. , 62:468, 1966. Chalmers, R. K. and Erickson, C. K. Central cholinergic blockade of the conditioned avoidance response in rats. Psychopharm. , 6:31-41, 1964. Clark, C. V. and Isaacson, R. L. Effect of bilateral hippocampal damage on DRL performance. J. comp. physiol. Psychol. , 59:137-140, 1965. Clemente, C. D., Green, J. D., and deGroot, Jr. Studies on behavior following rhinencephalic lesions in adult cats. Anat . Rec . . 127:279, 1957. Cordeau, J. P. Functional organization of the brainstem reticular formation in relation to sleep and wakefulness . Rev . canad . Biol. . 21:113-125, 1962. Cowan, W. M. , Raisman, G., and Powell, T. P. S. The origin of the mammullary peduncle and other hypothalamic connections of the midbrain. J . Anat . , 98:345-364, 1964.

PAGE 132

124 Cowan, W. M. , Raisman, G. and Powell, T. P. S. The connections of the amygdala. J. neurol. neurosurg. Psychiat. , 28:137-151, 1965. Cragg, B. G. and Hamlyn, L. H. Histologic connections and electrical and autonomic responses evoked by stimulation of the dorsal fornix in the rabbit. Exp. Neurol. , 1:187-213, 1959. Dahlstrom, A., Fuxe, K., Olson, L., and Ungerstedt, V. Ascending systems of catecholamine neurons from the lower brainstem. Acta. Physiol. Scand. . 62:485-486, 1962. Delgado, J. M. R. Free behavior and brain stimulation. Int. rev. Neurobiol. . 6:349-438, 1964. Delgado, J. M. R. Aggressive behavior evoked by radio-stimulation in monkey colonies. Amer. Zool. . 6:669-682, 1966. DeMolina, F. and Hunsperger, R. W. Organization of the subcortical governing defense and flight reactions in the cat. J . Physiol . , 145:251-256, 1962. DeVanzo, J. P., Daughterty, M., Ruckart, R., and Kang, L. Pharmacological and biochemical studies of isolation-induced fighting mice. Psychopharm. . 9:210-219, 1966. Douglas, R. J. The hippocampus and behavior. Psychol. Rev. , 67:416-442, 1967. Douglas, R. J. and Isaacson, R. L. Spontaneous alternation and scopolamine. Psychon. Sci. , 4:283-284, 1966. Douglas, R. J. and Pribram, K. H. Learning and limbic lesions. Neuropsychologia . 4:197-220, 1966. Douglas, R. J. and Raphelson, A. C. Septal lesions and activity. J. comp. physiol. psychol. , 62:465-467, 1966a. Douglas, R. J. and Raphelson, A. C. Spontaneous alternation and septal lesion. J. comp. physiol. Psychol. . 62:468-472, 1966b. Ellen, P. and Powell, E. W. Effects of septal lesions on behavior generated by positive reinforcement. Exp. Neurol. . 6:1-11, 1962a. Ellen, P. and Powell, E. W. Temporal discrimination in rats with rhinencephalic lesions. Exp. Neurol. . 6:538-547, 1962b. Ellen, P., Wilson, A. S., and Powell, E. W. Septal inhibition and timing behavior in the rat. J. comp. Neurol. . 10:120-132, 1964.

PAGE 133

125 Epstein, A. N. Reciprocal changes In feeding behavior produced by intrahypothalamlc chemical injection. Amer. J. Physiol. , 199:969-974, 1960. Falck, B., Hellarp, N. A., Thieme, G., and Torp, A. Fluorescence of catecholamines and related compounds condensed with formaldehyde. J. histochem. Cytochem. . 10:348-354, 1962. Feindel, W. and Gloor, P. Comparison of electrographic effects of stimulation of the amygdala and brainstem reticular formation in cats. Electroenceph . clin. Neurophysiol . , 6:389-402, 1954. Fernandez DeMolina, A. and Garcia-Sanchez , J. C. The properties of stria terminalis fibres. Physiol. Behav. , 2:225-227, 1967. Fisher, A. E. Maternal and sexual behavior induced by intracranial chemical stimulation. Sci. , 124:228-229, 1956. Fisher, A. and Coury, J. W. Cholinergic tracing of a central neural circuit underlying the thirst drive. Sci. , 138:691-692, 1962. French, J. D., Hernandez Peon, R., and Livingston, R. B. Projections from cortex to cephalic brainstem (reticular formation) in monkey. J. Neurophysiol. . 18:44-55, 1955. Fuller, J. C, Rosvold, H. E., and Pribram, K. H. The effects on affective and cognitive behavior in the dog of lesions of the pyrif orm-amygdala-hippocampal complex. J. comp. physiol. Psychol. . 50:89-96, 1957. Fulton, J. F. Frontal Lobotomy and Affective Behavior . New York: W. W. Norton, 1951. Fulton, J. F. and Ingraham, F. D. Emotional disturbances following experimental lesions of the base of the brain (pre-chiasmal) . Amer. J. Physiol. . 90:353, 1929. Fuxe, K. and Gunne, L. M. Depletion of the amine stores in brain catecholamine terminals on amygdaloid stimulation. Acta, physiol. Scand. . 62:493-494, 1964. Gey, K. F. and Fletcher, A. Influence of chlorpromazine and chlorprothixene on the cerebral metabolism of 5-hydroxytryptamine , norepinephrine, and dopamine. J. Pharmacol. . 133:18-24, 1961. Gloor, P. Electrophysiological studies on the connections of the amygdaloid nucleus of the cat. I. The neuronal organization of the amygdaloid projection system. Electroenceph. clin. Neurophysiol. . 7:223-242, 1955. Glowinski, J. and Baldessarini , R. J. Metabolism of norepinephrine in the central nervous system. Pharm. Rev. . 18:1201-1238, 1966.

PAGE 134

126 Goddard, G. V. Functions of the amygdala. Psychol. Bull. . 62:89-109, 1964. Goodman, L. S. and Gilman, A. The Pharmacological Basis of Therapeutics 3rd Ed. New York: The MacMillan Co., 1965. Green, J. D. The rhinencephalon : aspects of its relation to behavior and the reticular activating system. In H. H. Jasper (Ed.), Reticular Formation of the Brain . Boston: Little Brovm and Co., 1958, pp. 607-621. Green, J. D. and Arduini, A. Hippocampal electrical activity in arousal. J. Neurophysiol. , 17:533-547, 1954. Green, P. C. and Kling, A. Effects of amygdalectomy on affective behavior in juvenile and adult macaque monkeys. Proc . amer . psychol. Assn. , 2:93-94, 1966. Grossman, S. P. Eating or drinking elicited by direct adrenergic or cholinergic stimulation of the hypothalamus. Sci. , 132:301-302, 1960. Grossman, S. P. and Mountford, H. Effects of chemical stimulation of dorsal hippocampus on learning and performance. Am. J. Physiol. . 207:1387-1393, 1964. Grossman, S. P. Behavioral effects of chemical stimulation of the amygdala. J. comp. physiol. Psychol. . 57:29-36, 1964a. Grossman, S. P. Effect of chemical stimulation of the septal area on motivation. J. comp. physiol. Psychol. , 58:194-200, 1964b. Grossman, S. P. and Peters, R. H. Acquisition of appetitive and avoidance habits following atropineinduced blocking of the thalamic reticular formation. J. comp. physiol. Psychol. . 61:325-332, 1966. Grossman, S. P., Peters, R. H., Friedman, P. E., and Wilier, H. I. Behavioral effects of cholinergic stimulation of the thalamic reticular formation. J. comp. physiol. Psychol. . 59:57-65, 1965. Gunne, L. M. and Lewander, T. Monoamines in brain and adrenal glands of cats after electrically-induced defense reactions. Acta, physiol. Scand. . 67:405-410, 1966. Gurowitz, E. M. and Lubar, J. F. Changes in activity, food ingestion and passive avoidance behavior following limbic-septal ablation in the cat. Proc. amer. psychol. Assn. . 2:107-108, 1966. Harrison, J. M. and Lyon, M. The role of septal nuclei and components of the fornix in the behavior of the rat. J . comp . Neur ol . . 108:121-137, 1957.

PAGE 135

127 Harvey, J. A., Heller, A., and Moore, R. Y. The effect of unilateral and bilateral medial forebrain bundle lesions on brain serotonin. J. pharm. exp. Therap. , 140:103-110, 1963. Harvey, J. A. and Hunt, H. F. Effect of septal lesions on thirst in the rat as indicated by water consumption and operant responding for water reward. J. comp. physiol. Psychol. , 59:49-56, 1965. Hatton, G. I. Retention of discrimination and avoidance habits following lesions in the interpeduncular nucleus. J. comp. physiol. Psychol. . 59:331-334, 1965. Hearst, E. Drug effects on stimulus generalization in the monkey. Psychopharm. . 6:57-70, 1964. Heimstra, N. W. and Sailer, S. T. Effects of early drug treatment on adult dominance behavior in rats. Psychopharm. , 8:235-240, 1967. Heller, A., Harvey, J. A., and Moore, R. Y. A demonstration of a fall in brain serotonin following central nervous system lesions in the rat. Biochem. Pharmacol. , 11:859, 1962. Heller, A. and Moore, R. Y. Effect of central nervous system lesions on brain monoamines in the rat. J. pharm. exp. Therap. , 150:2-9, 1965. Heller, A., Seiden, L. S., Porcher, W., and Moore, R. Y. Regional effects of lateral hypothalamus lesions on brain norepinephrine in the cat. Int. J. Neuropharm. , 5:91-107, 1966. Hernandez Peon, R., Ibarra, G. C., Morgane, P. J., and Timo-Iaria, C. Limbic cholinergic pathways involved in sleep and emotional behavior. J . comp . Neurol . , 8:93-111, 1963. Hess, W. R. The Functional Organization of the Diencephalon . New York: Greene & Stratton, 1957. Hillarp, N. A., Fuxe, K., and Dahlstrom, A. Demonstration and mapping of central neurons containing dopamine, noradrenaline, and 5hydoxytrytamine and their reactions to psychopharmacology . Pharm. Rev. . 18:727-742, 1966. Hilton, S. M. and Zybrozya, A. W. Amygdaloid region for defense reactions and its efferent pathway to the brainstem. J. Physiol. . 165:160-173, 1963. Horowitz, Z. P., Furginelle, A. R., Brannick, C. J., Burke, B., and Carver, B. N. A new chemical structure with specific depressant effects on the amygdala and on the hyper-irritability of the "septal rat". Nature (Lond.), 200:369-370, 1963.

PAGE 136

128 Horowitz, Z. P., Ragozzian, P. W. , and Leof, R. C. Selective block of mouse-killing by anti-depressants. Lif. Sci. , 4:1909-1912, 1964. Horwath, F. E. Effects of basolateral amygdalectomy on three types of avoidance behavior in cats . J. comp. physiol. Psychol. , 56:380-389, 1963. Hunt, H. F. Some effects of meprobamate on conditioned fear and emotional behavior. Am. N. Y. acad. Sci. , 67:712-722, 1957. Isaacson, R. L., Douglas, R. J., and Moore, R. Y. The effect of radical hypocampal ablation on acquisition of avoidance response. J. comp. physiol. Psychol. , 54:625-628, 1961. Kaada, B. R. Somato-motor , autonomic and electrocorticographic responses to electrical stimulation of "rhinencephalic" and other structures in primates, cat and dog. Acta, physiol. Scand. , 24:1-258, 1951. Kaada, B. R., Rasmussen, E. W., and Kveim, D. Effects of hippocampal lesions on maze learning and retention in rats. Exp . Neurol . . 3:333-355, 1962. Karli, P. Septum, hypothalamus posterieur et aggresivite interspecifique Rat-Souris. J. Physiol. (Paris), 52:135-140, 1960. Kennard, M. A. Effect of bilateral cingulate ablation on behavior of cats. J. Neurophysiol. , 18:159-169, 1955. Kenyon, J. and Krieckhaus, E. E. Enhanced avoidance behavior following septal lesion in the rat as a function of lesion size and spontaneous activity. J« comp. physiol. Psychol. , 59:466-468, 1965. King, F. A. Effects of septal and amygdaloid lesions on emotional behavior and conditioned avoidance responses in the rat. J. nerv. ment. Pis. , 126:57-63, 1958. King, F. A. Relationship of the septal syndrome to genetic differences in emotionality in the rat. Psychol. Reports , 5:11-17, 1959. King, F. A. and Meyer, P. M. Effects of amygdaloid lesions upon septal hyperemotionality in the rat. Sci. , 128:655-656, 1958. Kling, A. Amygdalectomy in the kitten. Sci. , 137:429-430, 1962. Kluver, H. and Bucy, P. C. Preliminary analysis of functions of the temporal lobes in monkeys. Arch, neurol. Psychiat. , 42:979-1000, 1939.

PAGE 137

129 Koelle, G. B. The histochemical localization of cholinesterases in the central nervous system of the rat. J. comp. Neurol. , 100:211-235, 1954. Krieckhaus, E. E., Simmons, H. J., Thomas, G. D., and Kenyon, J. Septal lesions enhance shock avoidance behavior in the rat. Exp. Neurol. . 9:107-113, 1964. Lubar, T. F. Effects of medial cortical lesions on the avoidance behavior of the cat. J. comp. physiol. Psychol. , 58:38-46, 1964. Maclean, P. D. Chemical and electrical stimulation of the hippocampus in unrestrained animals. Part II. Behavioral finding. Arch, neurol. Psychiat. , 78:128-142, 1957. Maclean, P. D. The limbic system with respect to self-preservation and the preservation of the species . J. nerv. ment. Pis. , 127:1-11, 1958. Maclean, P. D. New findings relevant to the evolution of psychosexual functions of the brain. J. nerv. ment. Pis. . 135:289-301, 1962. Matalka, E. S. Septal Forebrain Lesions, Emotionality, and Avoidance Behavior in the Hamster . Unpublished Master's Thesis. Univ. of Fla., 1964. Mayer, C. and Stumpf, C. Bestimmung des optimalen reizparameter bei reizung des aczendierenden retikularen systems. Experientia. . 14:143, 1958. McCleary, R. A. Response specificity in the behavioral effects of limbic system lesions in the cat. J. comp. physiol. Psychol. . 54:605-613, 1961. McMullen, M. F. and Slotnick, B. M. Peficits in passive avoidance in mice with limbic system lesions. Paper read at the annual meeting of the Eastern Psychological Association. Boston, April, 1967. Meyers, B. Some effects of scopalamine on a passive avoidance response in rats. Psychopharm. . 8:111-119, 1965. Meyers, B. and Pomino, E. F. The effect of cholinergic blocking drugs on spontaneous alternation of rats. Arch, int. Pharmacodyn. . 150:525-529, 1964. Miller, N. E. Chemical coding in the brain. Sci. , 148:328-338, 1965.

PAGE 138

130 Mirsky, A. F., Rosvold, H. E., and Pribram, K. H. Effect of cingulectomy on social behavior of monkeys. J. Neurophysiol. . 20:588-601, 1957. Moore, R. Y. Effects of some rhinencephalic lesions on retention of conditioned avoidance behavior in cats. J. comp. physiol. Psychol. . 57:65-71, 1964. Moore, R. Y. and Heller, A. Monoamine levels and neuronal degeneration in rat brain following lateral hypothalamic lesions. J. pharm. exp. Therap. . 156:12-22, 1967. Moore, R. Y., Wong, S. L. R., and Heller, A. Regional effects of hypothalamic lesions on brain serotonin. Arch. Neurol. . 13:346-354, 1965. Nauta, W. J. H. Some projections of the medial wall of the hemisphere in the rat's brain (cortical areas 32 and 25, 24 and 29). Anat. Rec . 115:352, 1953. Nauta, W. J. H. An experimental study of the fornix system in the rat. J . comp . Neurol . , 104:247-271, 1956. Nauta, W. J. H. Hippocampal projections and related neural pathways to the midbrain in the cat. Brain, 81:319-340, 1958. Nauta, W. J. H. Fibre degeneration following lesions of the amygdaloid complex in the monkey. J . Anat . . 95:515-531, 1961. Nauta, W. J. H. Anatomical relationships between the amygdaloid complex, the dorsomedial thalamic nucleus, and the orbitofrontal cortex in the monkey. Anat . Rec . . 136:251, 1962. Nauta, W. J. H. Some connections of the prefrontal cortex in the monkey. In J. M. Warren and K. Akert (Eds.), The Frontogranular Cortex and Behavior . New York: McGraw-Hill, 1964. Nauta, W. J. H. and Kuypers, H. G. J. M. Some ascending pathways in the brainstem reticular formation. In H. H. Jasper (Ed.), Reticular Formation of the Brain . Boston: Little Brown and Co., 1958. Norton, S. Behavioral patterns as a technique for studying psychotropic drugs. In S. Garatinni and V. Ghetti (Eds.), Psychotropic Drugs . London: Elsevier Pub. Co., 1957. Olds, J. A. Preliminary mapping of electrical reinforcing effects in the brain. J. comp. physiol. Psychol. . 49:281-285, 1956. Olds, J. Adaptive functions of paleocortical and related structures. In H. F. Harlow and C. N. Woolsey (Eds.), Biological and Biochemical Bases of Behavior . Madison: Univ. of Wise Press 1958, pp. 237-262.

PAGE 139

131 Paasonen, M. K., Maclean, P. D., Giarman, N. J. 5-HT content of structures in the limbic system. J . Neurochem. , 1:326-333, 1957. Papez, J. W. Visceral brain and its component parts and their connections. J. nerv. ment. Pis. , 126:40-57, 1958. Pellegrino, A. S. The effects of septal and amygdaloid lesions on dominance in the hooded rat. Paper read at the annual meeting of the Eastern Psychological Association. Boston, April, 1967. Peretz, E. The effect of lesions of the anterior cingulate cortex on the behavior of the rat . J. comp. physiol. Psychol. , 53:540-548, 1960. Pechtel, C, McAvoy, T., Levitt, M. , Kling, A., and Masserman, J. H. The cingulates and behavior. J . nerv . ment . Pis . . 126:148-152, 1958. Petsche, H., Stumpf, C, and Gogolak, G. The significance of the rabbit's septum as a relay station between midbrain and the hippocampus. Electroenceph . clin. Neurol. . 14:202-211, 1962. Pirch, J. H. and Norton, S. Beta-Phenylisopropylhydrazine (JB-516) on septal-hyperirritability and brain amine levels in the rat. .. Psychopharm. . 8:181-190, 1965. Plrch, J. H. and Norton, S. "Rebound" increase in fixed interval lever pressing following a single does of chlorpromazine in septal rats. J. Psychopharmac . , 1:16-26, 1966. Plotnick, R. J. Changes in Social Behavior in Squirrel Monkeys after Temporal Lobectomy and an Interspecies Comparison of Social Pominance . Unpublished Poctoral Pissertation. Univ. of Fla. , 1966. Powell, E. W. Septal efferents revealed by axonal degeneration in the rat. Exp. Neurol. . 8:406-422, 1963. Powell, E. W. Corticolimbic interrelations revealed by evoked potential and degeneration techniques. J . exp . Neurol . , 10:463-474, 1964. Powell, T. P. S., Cowan, W. M. , and Raisman, G. The central olfactory connexions. J . Anat . . 99:791-814, 1965. Pribram, K. H. A review of theory in physiological psychology. Ann, rev. Psychol. . 11:1-40, 1960. Pribram, K. H. Limbic system. In D. E. Sheer (Ed,), Electrical Stimulation of the Brain . Austin: Univ. of Texas Press, 1961.

PAGE 140

132 Pribram, K. and Kruger, L. Functions of the "olfactory brain". Ann. N. Y. acad. Sci. . 58:109-138, 1954. Quarterman, D. and Miller, N. E. Sensory feedback in time response of drinking elicited by carbachol in the preoptic area of rat. J. comp. physlol. Psychol. , 62:350-353, 1966. Quay, W. B. and Wilhoft, D. C. Comparative and regional differences in serotonin content of reptilian brains . J . Neurochem. . 11:805-811, 1964. Ralsman, G. The connections of the septum. Brain , 89:317-348, 1966. Raisman, G., Cowan, W. M. , and Powell, T. P. S. An experimental analysis of the efferent projections of the hippocampus. Brain, 89:83-108, 1966. Raitt, J. R., Nelson, J. W. , and Tye, A. Effect of chlorpromazlne on septal hyperactivity in the rat. Brit. J. Pharm. . 17:473-478, 1961. ... Randall, L. 0. Pharmacology of methamlnodiazepoxide . Pis, nerv. Syst. . 21:7-10, 1961. Reivitch, M. and Glowinski, J. An autoradiographic study of the distribution of C^^ norepinephrine in the brain of the cat. Brain . In press, 1966. Rich, I. and Thompson, P. Role of the hlppocampal septal system, thalamus, and hypothalamus in avoidance conditioning. J . comp . physlol. Psychol. . 59:66-72, 1965. Roberts, W. W., Dember, W. N., and Brodwlck, M. Alternation and exploration in rats with hlppocampal lesions. J. comp. physlol. Psychol. . 55:695-700, 1962. Rosvold, H. E. and Delgado, J. M. R. The effect on delayed alternation test performance of stimulating or destroying electrically structures within the frontal lobes of the monkey's brain. J. comp. physlol. Psychol. , 49:365-372, 1956. Rosvold, H. E., Mirsky, A. F. , and Pribram, K. H. Influences of amygdalectomy on social behavior in monkeys. J. comp. physlol. Psychol. . 47:173-178, 1954. Rothfield, L. and Harman, P. J. On the relation of the hlppocampal fornix system to the control of rage responses in cats. J. comp. Neurol. . 101:265, 1954.

PAGE 141

133 Routtenberg, A. Discussion: Limbic-midbrain system and behavior. Paper read at the annual meeting of the American Psychological Association. New York, September, 1966. Schmaltz, L. W. and Isaacson, R. L. Retention of a DRL-20 schedule by hippocampectomized and partially neodecorticate cats. J. comp. phvsiol. Psvchol. . 62:128-132, 1966. ~ Schreiner, L. and Kling, A. Behavioral changes following rhinencephalic injury in cat. J. Neurophysiol . . 16:643-659, 1953. Schreiner, L. and Kling, A. Rhinencephalon and behavior. Amer. J Phvsiol. . 184:486-490, 1956. Schwartzbaum, J. S. Changes in the reinforcing properties of stimuli following ablation of the amygdaloid complex in monkeys. J. comp. phvsiol. Psychol. . 53:388-395, 1960. Schwartzbaum, J. S., Green, R. H., Beatty, W. W., and Thompson, J. B. Acquisition of avoidance behavior following septal lesions in the rat. J. comp. physiol. Psychol. . 63:95-104, 1967. Silverman, A. P. Barbituates, lysergic acid diethylamide, and the social behavior of laboratory rats. Psychopharm. . 10:155-171 1966. ' Sodetz, F. J. Social Dominance and Emotionality in the Golden Hamster Following Septal Ablation . Unpublished Master's Thesis Univ. of Fla. , 1965. Sodetz, F. J. and Bunnell, B. N. Interactive effects of septal lesions and social experience in the hamster. Paper read at the annual meeting of the Eastern Psychological Association. Boston, April, 1967a. Sodetz, F. J. and Bunnell, B. N. Septal lesions and the social behavior of the hamster. Paper read at the annual meeting of the Midwestern Psychological Association. Chicago, May, 1967b. Sodetz, F. J., Matalka, E. S., and Bunnell. B. N. Septal ablation ?o«Q fof ^TL^^^^''^°'' hamster. Psychon. Sci. . Spiegel, E. A., Miller, H. R., and Oppenheimer, M. J. Forebrain and rage reactions. J. Neurophysiol. . 3:538-548, 1940. Spiegel, E. A. and Wycis, H. T. Physiological and psychological results of thalamontomy . Proc. rov. soc. Med. . Suppl. 42:1-9 Stamm, J. S. Control of hoarding in rats by the median cortex J. comp. physiol. Psychol. . 47:21-27, 1954.

PAGE 142

134 Stainin, J. S. The function of the median cerebral cortex in maternal behavior of rats. J. comp. physiol. Psychol. . 48:347-356, 1955. Sterman, M. B. and Clemente, C. D. Forebrain inhibitory mechanisms: cortical synchronization induced by basal forebrain stimulation. Exp. Neurol. , 6:103, 1962. Stumpf, C. Drug action on the electrical activity of the hippocampus. Int. rev. Neurobiol. , 8:77-138, 1965. Sudack, H. S. and Maas , J. W. Behavioral-neurochemical correlations in reactive and non-reactive strains of rats. Sci. , 146:418-420, 1964. Thomas, G. J., Moore, R. Y., Harvey, J. A., and Hunt, H. Relations between the behavioral syndrome produced by lesions in the septal region of the forebrain and maze learning of the rat. J. comp. physiol. Psychol. . 52:527-532, 1959. Thomas G. J. and Slotnick, B. M. Effects of lesions of the cingulum on maze learning and avoidance conditioning in the rat. J . comp . physiol. Psychol. . 55:1085-1091, 1962. Thomas, G. J. and Slotnick, B. M. Impairment of avoidance responding by lesions in cingulate cortex in rats depends on food drive. J. comp. physiol. Psychol. . 56:959-964, 1963. Thompson, R. and Langer, S. K. Deficits in position reversal learning following lesions of the limbic system. J. comp. physiol. Psychol. . 56:982-995, 1963. Valenstein, E. S. and Nauta, W. J. H. A comparison of the distribution of the fornix system in the rat, guinea pig, cat, and monkey. J. comp. physiol. Psychol. . 113:337-361, 1959. Vellute, R. and Hernandez Peon, R. Atropine blockage of a cholinergic hypnogenic circuit. Exp . Neurol . . 8:20-29, 1963. Vilar, A. N., Gentil, C. G., and Covian, M. R. Alterations in sodium chloride and water intake after septal lesions in the rat. Physiol. Behav. . 2:167-170, 1967. Vogel, J. R., Hughes, H. A., and Carlton, R. L. Scopolamine, atropine, and conditioned fear. Psychopharm. . 10:409-416, 1967. Votaw, C. L. Study of septal stimulation and ablation in the macaque monkey. Neurol. , 10:202-209, 1960. Weiskrantz, L. Behavioral changes associated with ablation of the amygdaloid complex in monkeys. J. comp. physiol. Psychol . 49:381-391, 1956.

PAGE 143

135 White, E. E. A morphologic concept of the limbic lobe. Int. rev. Neurobiol. . 8:1-30, 1965. Whitehouse, J. M. , Lloyd, A. J., and Fifer, S. A. Comparative effects of atropine and methylatropine on maze acquisition and eating. J. comp. physiol. Psychol. , 58:475-476, 1964. Wycis, H. T. and Spiegel, E. A. Physiological and psychological results of thalamontomy . Proc. roy. soc. Med. . Suppl. 42:1-9, 1951. Yutzey, D. A., Meyer, P. M. and Meyer, D. R. Emotionality changes following septal and neocortical ablations in rats. J . comp . physiol. Psychol. . 58:463-465, 1964. Zucker, I. Effects of lesions of the septal-limbic area on the behavior of cats. J. comp. physiol. Psychol. . 60:344-352, 1965. Zybrozya, A. W. Defense reactions from the amygdala and stria terminalis. J. Physiol. . 153:27-28, 1960.

PAGE 144

APPENDIX A DISCUSSION OF S

PAGE 145

137 Subject 9N was part of the group which received treatment with norepinephrine. Each of the other six S^s in the group was submissive. Subject 9N was very aggressive and dominated its normal opponent. The cannula in this S^ was found to be located caudal to the septal region in an area where it may have permitted stimulation of structures adjoining the third ventricle. There is also a strong possibility that the stria terminalis could have been stimulated. This S^ showed 25% exploration, 2% hoarding, 10% social investigatory behavior, and 26% aggression. Its behavior, though characterized by aggression, was unlike that of any of the groups in the study. Perhaps, the behavior of this S^ could best be characterized as being somewhat the opposite of that seen in amygdaloid hamsters. That is, this S^ showed excessive social behavior and little non-social behavior.

PAGE 146

4' APPENDIX B THE RESULTS OF KRUSKAL-WALLIS TESTS APPLIED TO THE DATA.

PAGE 147

139 Kruskal-Wallis tests were applied to the data. Overall tests for significance were performed for each behavior category. The results of these tests are shown below. Category H df Level of Si Exploration 19 .69 9 .02 Hoarding 20 .42 9 .02 Social Investigatory 44 .81 9 .001 Aggression 19 .74 9 .02 Defense 32 .00 9 .001 Fighting 23 .99 9 .01 Submission 34 .66 9 .001 Escape 28 .33 9 .001 Miscellaneous 27 .99 9 .001 In none of the tests was any correction made for ties, therefore, the levels of significance reached were not as extreme as could have been obtained had this been done. The accepted level of significance was .05. Since all of the tests surpassed this level, a correction for ties was thought to be unnecessary.

PAGE 148

BIOGRAPHICAL SKETCH Frank Jack Sodetz, Jr. was born on June 10, 1941 in Chicago, Illinois. He attended public schools in Chicago and was graduated from Christian Fenger High School in 1959. He attended Knox College in Galesburg, Illinois from 1959 until 1963 when he received his B.A. degree with a major in Experimental Psychology. He began graduate work at the University of Florida in 1963 and received his M.S. degree with a major in Physiological Psychology in 1965. He continued his graduate work until August, 1967 when he received his Ph.D. with a major in Physiological Psychology. He minored in the medical sciences under a program administered by the Center for Neurobiological Sciences of the University of Florida. 140 -

PAGE 149

This dissertation was prepared under the direction of the chairman of the candidate's supervisory committee and has been approved by all members of that committee. It was submitted to the Dean of the College of Arts and Sciences and to the Graduate Council, and was approved as partial fulfillment of the requirements of the degree of Doctor of Philosophy. August 12, 1967 Dean of the College of Arts and Sciences Dean, Graduate School Supervisory Committee: