__________________ ____________________  

Scientific World of LSD



The psychedelic effects of d-Lysergic Acid Diethylamide-25
(LSD) were discovered by Dr. Albert Hoffman by accident in
1938. In the 1950s and 1960s, LSD was used by psychiatrists
for analytic psychotherapy. It was thought that the
administering of LSD could aid the patient in releasing
repressed material. It was also suggested that
psychiatrists themselves might develop more insight into
the pathology of a diseased mind through self
experimentation. 1,2 

During the late 60s, LSD became popular as a recreational
drug. While it has been suggested that recreational use of
the drug has dropped, a recent report on CNN claimed that
4.4% of 8th graders have tried it. LSD is considered to be
one of, if not the, most potent hallucinogenic drug known.
Small doses of LSD (1/2 - 2 ug/kg body weight) result in a
number of system wide effects that could be classified into
somatic, psychological, cognitive, and perceptual
categories. These effects can last between 5 and 14 hours. 

Somatic, psychological, Cognitive, Perceptual mydriasis
hallucinations result in disturbed thought processes,
increased stimulus from environment, hyperglycemia
depersonalization, difficulty expressing thoughts, changes
in shape/color, hyperthermia, reliving of repressed
memories impairment of reasoning, synaesthesia, (running
together of sensory modalities), piloerection mood swings
(related to set and setting), impairment of memory - esp.
integration of short - long term disturbed perception of
time, vomiting, euphoria lachrymation, megalomania
hypotension, schizophrenic-like state.
Respiratory effects are stimulated at low doses and
depressed at higher doses and reult in reduced "defenses"
that are subject to "power of suggestion", brachycardia.
The study of hallucinogens such as LSD is fundamental to
the neurosciences. 

Science thrives on mystery and contradiction; indeed
without these it stagnates. The pronounced effects that
hallucinogens have throughout the nervous system have
served as potent demonstrations of difficulty to explain
behavior. The attempts to unravel the mechanisms of
hallucinogens are closely tied to basic research in the
physiology of neuroreceptors, neurotransmitters, neural
structures, and their relationship to behavior. 

This paper will first examine the relationship between
neural activity and behavior. It will then discuss some of
the neural populations and neurotransmitters that are
believed to be affected by LSD. The paper will conclude
with a more detailed discussion of possible ways that LSD
can affect the neurotransmitter receptors. 

A Brief Foray Into Philosophy and the Cognitive Sciences 

Modern physics is divided by two descriptions of the
universe: the theory of relativity and quantum mechanics.
Many physicists have faith that at some point a "Grand
Unified Theory" will be developed which will provide a
unified description of the universe from subatomic
particles to the movement of the planets. As in physics,
the cognitive sciences can describe the brain at different
levels of abstraction. For example, neurobiologists study
brain function at the level of neurons while psychologists
look for the laws describing behavior and cognitive
mechanisms. Many in these fields believe that it is
possible that one day we will be able to understand
complicated behaviors in terms of neuronal mechanisms.
Others believe that this unification isn't possible even in
theory because there is some metaphysical quality to
consciousness that transcends neural firing patterns. 

Even if consciousness can't be described by a "Grand
Unified Theory" of the cognitive sciences, it is apparent
that many of our cognitive mechanisms and behaviors can.
While research on the level of neurons and psychological
mechanisms is fairly well developed, the area in-between
these is rather murky. Some progress has been made and
cognitive scientists have been able to associate mechanisms
with areas of the brain and have also been able to describe
the effects on these systems by various neurotransmitters.
For example, disruption of hippocampal activity has been
found to result in a deficiency in consolidating short term
to long term memory. 

Cognitive disorders such as Parkinson's disease can be
traced to problems in dopaminergic pathways. Serotonin has
been implicated in the etiology of various CNS disorders
including depression, obsessive-compulsive behavior,
schizophrenia, and nausea. It is also known to effect the
cardiovascular and thermoregulatory systems as well as
cognitive abilities such as learning and memory. The lack
of knowledge in the middle ground between neurobiology and
psychology makes a description of the mechanisms of
hallucinogens necessarily coarse. 

The following section will explore the possible mechanisms
of LSD in a holistic yet coarse manner. Ensuing sections
will concentrate on the more developed studies of the
mechanisms on a neuronal level. 

The Suspects Researchers have attempted to identify the
mechanism of LSD through three different approaches:
comparing the effects of LSD with the behavioral
interactions already identified with neuotransmitters,
chemically determining which neurotransmitters and
receptors LSD interacts with, and identifying regions of
the brain that could be responsible for the wide variety of
effects listed in Table 1. Initial research found that LSD
structurally resembled serotonin (5-HT). As described in
the previous section, 5-HT is implicated in the regulation
of many systems known to be affected by LSD. This evidence
indicates that many of the effects of LSD are through
serotonin mediated pathways. 

Subsequent research revealed that LSD not only has
affinities for 5-HT receptors but also for receptors of
histamine, ACh, dopamine, and the catecholines: epinephrine
and norepinephrine.3 Only a relative handful of neurons
(numbering in the 1000s) are serotonergic (i.e. release
5-HT). Most of these neurons are clustered in the
brainstem. Some parts of the brainstem have the interesting
property of containing relatively few neurons that function
as the predominant provider of a particular
neurotransmitter to most of the brain. For example, while
there are only a few thousand serotonergic cells in the
Raphe Nuclei, they make up the majority of serotonergic
cells in the brain. Their axons innervate almost all areas
of the brain. The possibility for small neuron populations
to have such systemic effects makes the brain stem a likely
site for hallucinogenic mechanisms. 

Two areas of the brainstem that are thought to be involved
in LSD's pathway are the Locus Coeruleus (LC) and the Raphe
Nuclei. The LC is a small cluster of norepinephrine
containing neurons in the pons beneath the 4th ventricle.
The LC is responsible for the majority of norepinephrine
neuronal input in most brain regions.4 It has axons which
extend to a number of sites including the cerebellum,
thalamus, hypothalamus, cerebral cortex, and hippocampus. A
single LC neuron can affect a large target area.
Stimulation of LC neurons results in a number of different
effects depending on the post-synaptic cell. For example,
stimulation of hippocampal pyramidal cells with
norepinephrine results in an increase in post-synaptic

The LC is part of the ascending reticular activating system
which is known to be involved in the regulation of
attention, arousal, and the sleep-wake cycle. Electrical
stimulation of the LC in rats results in hyper-responsive
reactions to stimuli (visual, auditory, tactile, etc.)5 LSD
has been found to enhance the reactivity of the LC to
sensory stimulations; however, LSD was not found to enhance
the sensitivity of LC neurons to acteylcholine, glutamate,
or substance P.6 Furthermore, application of LSD to the LC
does not by itself cause spontaneous neural firing. While
many of the effects of LSD can be described by its effects
on the LC, it is apparent that LSD's effects on the LC are

While norepinephrine activity throughout the brain is
mainly mediated by the LC, the majority of serotonergic
neurons are located in the Raphe Nuclei (RN). The RN is
located in the middle of the brainstem from the midbrain to
the medulla. It innervates the spinal cord where it is
involved in the regulation of pain. Like the LC, the RN
innervates wide areas of the brain. Along with the LC, the
RN is part of the ascending reticular activating system.
5-HT inhibits ascending traffic in the reticular system;
perhaps protecting the brain from sensory overload.
Post-synaptic 5-HT receptors in the visual areas are also
believed to be inhibitory. Thus, it is apparent that an
interruption of 5-HT activity would result in
disinhibition, and therefore excitation, of various sensory

Current thought is that the mechanism of LSD is related to
the regulation of 5-HT activity in the RN; however, the RN
is also influenced by GABAergic, catecholamergic, and
histamergic neurons. LSD has been shown to also have
affinities for many of these receptors. Thus it is possible
that some of its effects may be mediated through other
pathways. Current research however has focused on the
effects of LSD on 5-HT activity. Before specific mechanisms
and theories are discussed, a brief discussion of the
principles of synaptic transmission will be given. 

Overview of Synaptic Transmission 

There are two types of synapses between neurons: chemical
and electrical. Chemical synapses are more common and are
the type discussed in this paper. When an action potential
(AP) travels down a pre-synaptic cell, vesicles containing
neurotransmitters are released into the synapse
(exocytosis) where they affect receptors on the post
synaptic cell. Synaptic activity can be terminated through
re-uptake of the neurotransmitter to the pre-synaptic cell,
the presence of enzymes which inactivate the transmitter
(metabolism), or simple diffusion. 

A pre-synaptic neuron can act on the post-synaptic neuron
through direct or indirect pathways. In a direct pathway,
the post-synaptic receptor is also an ion channel. The
binding of a neurotransmitter to its receptor on the
post-synaptic cell directly modifies the activity of the
channel. Neurotransmitters can have excitatory or
inhibitory effects. If a neurotransmitter is excitatory, it
binds to a ligand activated channel in the post-synaptic
cell resulting in a change in membrane permeability to ions
such as Na+ or K+. This results in a depolarization which
therefore brings the post-synaptic cell closer to

Inhibitory neurotransmitters can work post-synaptically by
modifying the membrane permeability of the post-synaptic
cell to anions such as Cl- which results in
hyperpolarization. Many neurotransmitters that have
system-wide effects such as epinephrine (adrenaline),
norepinephrine (noradrenaline), and 5-HT work by an
indirect pathway. In an indirect pathway, the post-synaptic
receptor acts on an ion channel through indirect means such
as a secondary messenger system. Many indirect receptors
such as muscarinic, Ach, and 5-HT involve the use of G

Indirect mechanisms often will alter the behavior of a
neuron without affecting its resting potential. For
example, norepinephrine blocks slow Ca activated K channels
in the rat hippocampal pyramidal cells. Normally, Ca influx
eventually causes the K channels to open. This causes a
prolonged after hyperpolarization which extends the
refractory period of the neuron. Therefore, by blocking the
K channels, the prolonged after hyperpolarization is
inhibited which results in the neuron firing more APs for a
given excitatory input.5 

Other indirect means of neuromodulation include interfering
with pre-synaptic neurotransmitter synthesis, storage,
release, or reuptake. Inhibiting the reuptake of a
neurotransmitter, for example, can cause an excitatory
response. Stimulation of neurotransmitter receptors can
have a variety of effects on both pre and post-synaptic
cells. Pre-synaptic receptors are sometimes involved in
self regulation while post-synaptic receptors can cause an
increase (excitation) or decrease (inhibition) of AP firing
in a neuron. 

A subtler method of neuromodulation involves molecules that
affect these neuroreceptors. Molecules that excite a
receptor are referred to as agonists while those that
interfere with receptor binding are called antagonists. For
example, 5-HT often acts as an inhibitory neurotransmitter.
A 5-HT receptor antagonist could interfere with the
activation of post-synaptic 5-HT receptors causing them to
be less responsive to inhibition. This disinhibition would
make the post-synaptic cell more responsive to neural
inputs, most likely resulting in an excitatory response. 

Theory: LSD Pre-synaptically Inhibits 5-HT Neurons. 

Raphe Nuclei neurons are autoreactive; that is they exhibit
a regular spontaneous firing rate that is not triggered by
an external AP. Evidence for this comes from the
observation that RN neural firing is relatively unaffected
by transactions isolating it from the forebrain. Removal of
Ca++ ions, which should block synaptic transmission, also
has little effect on the rhythmic firing pattern. This
firing pattern however is susceptible to neuromodulation by
a number of transmitters.7 

In 1968, Aghajanian and colleagues observed that systemic
administration of LSD inhibited spontaneous firing of these
autoreactive serotonergic neurons in the RN. Serotonergic
neurons are known to have a negative feedback pathway
through autoreceptors (receptors on the pre-synaptic cell
that respond to the neurotransmitter released by the cell).
This means that an increase in 5-HT levels causes a
decrease in the activity of serotonergic neurons.
Serotonergic neurons are also known to make synaptic
connections with other RN neurons. This could have the
result of spreading out the effects of negative feedback to
other RN neurons. This led to the theory that LSD causes a
depletion of 5-HT through negative feedback in pre-synaptic
autoreceptors.7 The depletion of 5-HT was thought to be
responsible for the effects on the previously described
systems innervated by the serotonergic neurons. A number of
subsequent observations have called this theory into doubt
Low doses of LSD affect behavior but do not depress firing
in the RN.8
The behavioral effects of LSD outlast the modification of
RNN firing.8
While repeated dosage of LSD results in a decrease of
behavioral modifications (tolerance), its effects on the RN
are unchanged.8
Other hallucinogens such as mescaline and DOM do not effect
R neurons.8
Depletion of 5-HT does not eliminate the effectiveness of
LSD. If LSD worked by inhibiting the 5-HT output of
pre-synaptic 5-HT neurons, it should be ineffaceable if
5-HT is depleted. The opposite result was actually
observed; depletion enhances LSD activity.9
Mianserin, a 5-HT2 receptor antagonist, blocks LSD behavior
but does not block LSD's depression of RN neurons.9 While
LSD does cause a decrease in the autoreactive firing of RN
neurons, this appears to be an effect and not the cause. 

These observations are considered however to be compatible
with a post-synaptic model. Subsequent research found that
LSD and other hallucinogens have a high affinity for
post-synaptic 5-HT1 and 5-HT2 receptors. In fact there is
significant correlation between the affinity of a
hallucinogen for these receptors and its human potency.
While it seems logical that 5-HT activity is modulated at
5-HT receptor sites, it is possible that LSD could be
affecting 5-HT receptor activity indirectly through adrenic
or dopaminic pathways. However, blocking these receptors
caused no change in LSD's activity on the 5-HT receptors,
thus it appears that 5-HT activity is indeed modified by
5-HT receptors.10 Even theough evidence indicates that LSD
is a 5-HT1 agonist, it is debated whether the effects on
5-HT2 receptors is agonistic or antagonistic.11 

Theory: LSD Post-synaptically Antagonizes 5-HT2 Receptors 

Initial post-synaptic theories postulated that LSD was a
5-HT2 agonist. Pierce and Peroutka (P&P), however, argued
that LSD has a number of antagonistic properties and called
into doubt some of the evidence presented as being
compatible with agonist activity. The primary evidence for
agonistic behavior comes from observations that the effects
of LSD are inhibited by 5-HT2 antagonists. 

P&P pointed out that this is not always the case. For
example, some 5-HT2 antagonists such as spiperone do not
block LSD behavior. In addition, radioligand binding
studies have shown that the affinity of 5-HT2 receptor
agonists is pH dependent while the affinity of 5-HT2
receptor antagonists and LSD are pH independent.9 5-HT2
receptors are connected to a phosphatidylinositol (PI)
second messenger system. PI turnover has been found to be
stimulated by 5-HT and antagonized by 5-HT2 antagonists.
P&P found that nM concentrations of LSD do not stimulate PI
turnover. Therefore, LSD does not act as a classic agonist. 

They also found that nM concentrations of LSD inhibited the
stimulatory effect of 10M 5-HT. The ability of LSD to
inhibit a concentration 1000x greater is consistent with it
being a 5-HT2 antagonist P&P also point out that the
excitatory effects of 5-HT on CNS neurons appears to be
caused by a decrease in K+ conductance attributable to
activation of 5-HT2 receptors. P&P found that LSD inhibits
this effect in rat somatosensory pyramidal neurons. This
also is evidence that LSD acts in an antagonistic role.9 

The final line of evidence presented by P&P was from smooth
muscle studies. The guinea pig trachea contracts when M
concentrations of 5-HT are present. The ability of 5-HT
antagonists to inhibit this effect correlates with the
antagonists affinity for the 5-HT2 binding site. Thus it
appears that this muscle contraction is 5-HT2 mediated. It
was found that nM concentrations of LSD did not cause
muscle contraction and inhibited the agonistic effects of M
concentrations of 5-HT. This also is compatible with the
actions of an antagonist. 

Theory: LSD Post-synaptically Partially Agonizes 5-HT

Many of the apparent contradictions in evidence in the
debate over whether LSD acts as a 5-HT2 agonist or
antagonist can be reconciled by the theory that LSD acts as
a partial 5-HT2 agonist. Dr. Glennon presented a number of
arguments for this theory including data from his own
research and from the studies discussed by P&P in the
previous section. One of the primary tools used by Glennon
to determine the effects of various chemicals on the
interactions between LSD and 5-HT was drug discrimination
training in rats. Rats were trained to discriminate
1-(2,5-dimethoxy-4-methylphenyl)-2-aminopropane (DOM) from
saline. Training with DOM stimuli generalized to many
indolealkylamine and phenalkylamine hallucinogens. DOM was
chosen instead of LSD as a training drug because of concern
that LSD had a number of pharmacological effects. It was
thought that if the rat was trained with LSD, it might make
discriminations based on one of the pharmacological effects
of LSD rather than its effects on 5-HT. 

With this tool, Glennon demonstrated that a number of 5-HT2
antagonists inhibited the ability of rats to discriminate
LSD from saline. This indicates that LSD acts as a 5-HT2
agonist. Glennon offered no explanation for P&P's
observation that some antagonists such as spiperone do not
have this effect. However, spiperone and a few other
similar antagonists appear to be only about 40% effective
in inhibiting 5-HT2 sites due to its relative
nonselectivity.13 As discussed in the previous section, PI
turnover has been found to be stimulated by 5-HT and is
antagonized by 5-HT2 antagonists. 

In another study of the effects of LSD on PI turnover, it
was found that LSD acted as a partial agonist (it produces
approximately 25% of the effect caused by 5-HT). The
apparent difference between this second study and P&P's is
that the second study tested the effects from a variety of
doses. The conclusion that was reached was that while LSD
has a higher affinity for 5-HT receptors than 5-HT, it has
a lower efficacy. This is compatible with P&P's observation
that nM concentrations of LSD inhibited the stimulatory
effects of uM 5-HT. If LSD acted as a partial agonist with
low efficacy, it could compete with 5-HT in binding to
5-HT2 receptors. Since 5-HT is a more potent agonist than
the LSD, the effects of LSD would appear antagonistic. 

Glennon argued that the guinea pig trachea may not be a
good example since 5-HT does not work through a PI
mechanism in this case. In the rat aorta, however, 5-HT
does hydrolize PI and the contractile effects of 5-HT are
antagonized by ketanserin (a 5-HT2 antagonist). While LSD
was not tested, another hallucinogen, DOB, was found to
have an agonistic effect that could be antagonized by
ketanserin. This suggests that LSD acts agonistically in
the rat aorta. 

Glennon points out that it may well be the case that in
other cases, the effects may be antagonistic. However,
these effects could be explained if LSD had a low efficacy
for the receptor. Hyperthermia and platelet aggregation are
both affected by 5-HT2 mechanisms. Hallucinogens such as
LSD have been shown to behave agonistically and in the case
of platelets, to be antagonized by 5-HT2 antagonists such
as ketanserin.11 

LSD often has a biphasic response in which low doses have
the opposite effects of higher doses. The head twitch
response in rodents is believed to be 5-HT2 mediated. At
low doses, it has been found that LSD elicits a head-twitch
response while at higher doses it antagonizes the response.
The rat startle reflex is amplified at low dosages of LSD
while decreased at higher doses. This biphasic behavior can
also be explained if LSD behaves as a partial agonist.11 

In summary, this theory claims that: "LSD is a
high-affinity, low efficacy, nonselective 5-HT agonist; in
the absence of another agonist it may function as an
agonist, whereas in the presence of a high efficacy
agonist, it will function as an antagonist." 11 

Theory: LSD Post-synaptically Agonizes 5-HT1 Receptors 

Glennon also gave another possible explanation for the
antagonistic activity of LSD. There is some evidence that
5-HT1 receptors have an antagonistic relationship with
5-HT2 receptors. As discussed in the previous section, head
twitch behavior is believed to be 5-HT2 mediated. DOI acts
as a 5-HT2 agonist and elicits head twitch. 5-OMe DMT also
is a 5-HT agonist but has less efficacy than DOI. If the
subject is pretreated with 5-OMe DMT, the effects of DOI
are attenuated (because many of the receptors are filled
with the lower efficacy 5-OMe DMT molecules.) It has been
found that A 5-HT1 agonist (8-OH DPAT) can also cause DOI

Other studies have also demonstrated that 5-HT1 agonists
can behave functionally as 5-HT2 antagonists.11 Glennon
argued that this theory is lent extra credence from the
observation that 5-HT2 and 5-HT1c have similar
relationships with various hallucinogens. A number of these
hallucinogens have been shown to be 5-HT1c agonists. Like
5-HT2 sites, the affinity of hallucinogens for 5-HT1c sites
correlates with their hallucinogenic potency in humans.
Thus another explanation of the biphasic behavior of LSD is
that increasingly higher doses of LSD cause increased
antagonism of the 5HT2 receptor through agonism of 5HT1

Although, the pre-synaptic theory seems to be fairly well
discredited, it is interesting to note that there is debate
as to whether pre-synaptic serotonin autoreceptors are of
the 5-HT1 type. Whether serotonergic autoreceptors are
5-HT1 or not, it has been demonstrated that there are also
post-synaptic 5HT-1 receptors.12 While the role of these
receptors is not completely known, some researchers have
hypothesized that 5-HT1 receptors may be involved in the
regulation of norepinephrine.13 As discussed previously,
the majority of norepinephrine neurons are located in the
LC which also has system wide innervation. 

Recent research on 5-HT receptors calls the theory that
5-HT1 agonism results in 5-HT2 antagonism into question and
Glennon's paper, the 5-HT1c receptor, has been reclassified
as 5-HT2c. Since the 5-HT2 receptors discussed in this
paper belong to the same family as what was called the
5-HT1c receptor, these have been reclassified as 5-HT2a.14
Additionally,the "5-HT1c" is a member of the 5-HT2 family,
therefore, it is not surprising that the LSD affinities are
similar for the two receptors. While these
reclassifications do not necessarily discount the theory
that one receptor has an antagonistic effect on the other,
it seems likely that the evidence for this may need to be
re-evaluated in terms of recent findings. 


The lack of understanding about the mechanisms of LSD is
indicative of the problems involved in the bridging of the
worlds of psychology and neurobiology. As more is learned
about the roles and interactions of various
neurotransmitters, receptors, and on a larger scale:
portions of the brain, the mystery will be further

With this caveat emptor firmly in mind, it seems that the
best explanation of LSD's effects is that it behaves as a
high affinity partial 5-HT agonist. Depending on the
presence of other molecules and its own concentration, LSD
can have either agonistic or antagonistic effects on
post-synaptic 5-HT2 family receptors. This modulation of
5-HT behavior is probably responsible for many of the
effects attributable to LSD. LSD also has an affinity for
other neurotransmitter receptors that play important roles
in the brain stem such as norepinephrine, dopamine, and

It is also hypothesized that LSD may modulate neural
responses to these transmitters through its activity on
5-HT1 receptors. Both the Locus Coeruleus and the Raphe
Nuclei are part of the ascending reticular activating
system which is implicated in the sensory modalities. The
inhibition of 5-HT in the RN and release of norepinephrine
from LC neurons results in a flood of information from the
sensory system reaching the brain. Some of the cognitive
effects of LSD could be attributed to the effects of brain
stem innervation to areas of the brain such as the cerebral
cortex and the hippocampus.
1.(1995): "FAQ-LSD" From internet newsgroup:
2.Sankar (1975): "LSD: A Total Study"
3.Ashton H (1987): "Brain Systems Disorders and
Psychotropic Drugs"
4.Snyder (1986): "Drugs and the Brain" Sci Am Books Inc.
5.Nicholls J, Martin R, Wallace B (1992): "From Neuron to
Brain: Acellular and Molecular Approach to the Function of
the Nervous System"
6.Aghajanian GK(1980): "Mescaline and LSD Facilitate the
Activation of Locus Coeruleus Neurons by Peripheral
Stimulation" Brain Res 186:492-496
7.Jacobs, B (1985): "An Overview of Brain Serotonergic Unit
Activity and its Relevance to the Neuropharmacology of
Serotonin." From: Green, A: Neuropharmacology of Serotonin
8.Jacobs, B, Trulson M, Heym J, (1981): "Dissociations
Between the Effects of Hallucinogenic Drugs on Behavior and
Raphe Unit Activity in Freely Moving Cats" Brain Res
9.Pierce P, Peroutka S (1990): "Antagonist Properties of
d-LSD at 5-Hydroxytryptamine2 Receptors".
Neuropsychopharmacolgy 3(5-6):509-517
10.Moret C (1985): "Pharmacology of the Serotonin
Autoreceptor" From: Green, A: Neuropharmacology of Serotonin
11.Glennon R (1990): "Do Classical Hallucinogens Act as
5-HT2 Agonists or Antagonists?" Neuropsychopharmacolgy
12.Green R, Heal D (1985): "The Effects of Drugs on
Serotonin Mediated Behavioral Models" From Green, A:
Neuropharmacology of Serotonin
13.Leysen J (1985): "Characterization of serotonin receptor
binding sites" From Green, A: Neuropharmacology of Serotonin
14.Borne R. (1994) "Serotonin: The Neurotransmitter for the
90's" URL:



Quotes: Search by Author