BEHAVIOR OF RATS UNDER FIXED CONSECUTIVE NUMBER SCHEDULES: EFFECTS OF DRUGS OF ABUSE

JOURNAL OF THE EXPERIMENTAL ANALYSIS OF BEHAVIOR
1997, 68, 117–132
NUMBER 1 (JULY)
BEHAVIOR OF RATS UNDER FIXED
CONSECUTIVE NUMBER SCHEDULES:
EFFECTS OF DRUGS OF ABUSE
SAMUEL H. SNODGRASS, JANET L. HARDIN, AND D. E. MCMILLAN
UNIVERSIT Y OF ARKANSAS FOR MEDICAL SCIENCES
Four rats responded under a simple ?xed consecutive number schedule in which eight or more
consecutive responses on the run lever, followed by a single response on the reinforcement lever,
produced the food reinforcer. Under this simple schedule, dose–response curves were determined
for diazepam, morphine, pentobarbital, and phencyclidine. The rats were then trained to respond
under a multiple ?xed consecutive number schedule in which a discriminative stimulus signaled
when the response requirement on the run lever had been completed in one of the two ?xed
consecutive number component schedules. Under control conditions, the percentage of reinforced
runs under the multiple-schedule component with the discriminative stimulus added was much high-
er than the percentage of reinforced runs under the multiple-schedule component without the
discriminative stimulus. All of the drugs decreased the percentage of reinforced runs under each of
the ?xed consecutive number schedules by increasing the conditional probability of short run
lengths. This effect was most consistently produced by morphine. The drugs produced few differ-
ences in responding between the multiple ?xed consecutive number components. Responding under
the simple ?xed consecutive number schedule, however, was affected at lower doses of the drugs
than was responding under the same ?xed consecutive number schedule when it was a component
of the multiple schedule. This result may be due to the difference in schedule context or, perhaps,
to the order of the experiments.
Key words: ?xed consecutive number schedule, stimulus control, drug, multiple schedule, lever
press, rats
In 1958, Mechner reported the use of what
stimulus change sets the occasion for the re-
later became known as the ?xed consecutive
sponse on the reinforcement lever. This mod-
number (FCN) procedure (Mechner & La-
i?ed FCN schedule is typically designated as
tranyi, 1963). To earn the reinforcer under
the FCN-SD schedule (Laties, 1972).
this procedure, the subject must ?rst emit a
Under baseline conditions, rates of re-
minimum number of consecutive responses
sponding under the FCN and FCN-SD sched-
on one lever (run lever) and then emit a sin-
ules are typically equivalent (Laties, 1972).
gle response on a second lever (reinforce-
Due to this equivalence in the initial response
ment lever). If a subject switches from the
rates, rate-dependent effects should be simi-
run lever to the reinforcement lever prior to
lar; thus, baseline response rates can be ruled
the completion of the minimum number of
out as the cause of any drug-produced differ-
consecutive responses, the reinforcer is with-
ences in rate of responding under the two
held and the run lever response requirement
schedules (Laties, Wood, & Rees, 1981).
is reset.
Therefore, there is less ambiguity in the de-
The FCN procedure can be modi?ed such
termination of differential drug effects under
that, when the subject emits the required
the FCN and FCN-SD schedules compared to
consecutive number of responses on the run
other stimulus control procedures that lack
lever, a stimulus change occurs. By signaling
the ability to control for rate-dependent ef-
that the run requirement has been met, this
fects (Laties, 1975).
Studies have shown that some drugs affect
responding maintained under the FCN
This work was supported by NIDA National Research
Service Award DA 05332-02 to S. H. Snodgrass and by
schedule at lower doses, or to a greater ex-
NIDA Grant DA 02251-11 to D. E. McMillan. We thank
tent, than responding maintained under the
NIDA for supplying the phencyclidine used in this study.
FCN-SD schedule (Evans, Laties, & Weiss,
Correspondence and requests for reprints should be ad-
1975; Laties, 1972; Laties et al., 1981; Picker,
dressed to D. E. McMillan, Department of Pharmacology
Leibold, Endsley, & Poling, 1986a; Rees,
and Toxicology, University of Arkansas for Medical Sci-
ences, Slot 611, 4301 West Markham Street, Little Rock,
Wood, & Laties, 1985; Wagman & Maxey,
Arkansas 72205 (E-mail: mcmillan@biomed.uams.edu).
1969). These results are consistent with those
117

---

118
SAMUEL H. SNODGRASS et al.
from other studies that have shown that be-
the home cage. A cycle of lights on at 7:00
havior under strong exteroceptive stimulus
a.m. and lights off at 7:00 p.m. was in effect
control is affected less by some drugs than is
in the colony room.
behavior that is without exteroceptive stimu-
lus control (Laties, 1975; Laties & Weiss,
Apparatus
1966; Thompson & Corr, 1974). Thus, it
Sessions were conducted using a Ger-
seems that placing an animal’s behavior un-
brands two-lever operant chamber (Model
der strong external stimulus control may in-
G7322). The left lever was a Gerbrands re-
crease the resistance of that behavior to the
tractable response lever (Model G6311). The
effects of certain drugs (Laties, 1975).
levers were located 8.5 cm above the ?oor of
The present experiments were designed to
the chamber on either side of a rectangular
compare the effects of four drugs of abuse
opening 2 cm above the ?oor which con-
on responding maintained under FCN and
tained a receptacle for the delivery of 97-mg
FCN-SD schedules. The drugs tested were
food pellets (Noyes Formula A). Both levers
pentobarbital, diazepam, morphine, and
could be activated by a force in excess of 0.4
phencyclidine (PCP). These drugs were cho-
N. A bank of four lights (28-V DC lights with
sen because of the importance of understand-
translucent red plastic covers) was mounted
ing how drugs of abuse alter stimulus control
above each lever. Two 28-V DC houselights
of behavior. Also, each of these drugs is a
mounted on the back quarter of the top pan-
member of a different pharmacological class
el served as houselights. The operant cham-
from the others and is a prototype drug of
ber was housed in a sound-attenuating cubi-
the class to which they belong. Therefore, be-
cle equipped with a fan for ventilation and a
cause they are prototypical drugs, their ef-
speaker that provided white masking noise.
fects should be representative, to some de-
Behavioral contingencies were controlled
gree, of the other members in each drug
with, and data collected by, a TRS-80 model
class. However, because each drug does rep-
III microcomputer (Radio Shack) interfaced
resent a different pharmacological class and
with a Microcomputer Interface II (Med As-
because the FCN and FCN-SD schedules pro-
sociates, Inc.). The computer and interface
duce different levels of stimulus control (La-
were located in a room adjacent to that of the
ties, 1972; Laties et al., 1981; Rees et al.,
operant chamber.
1985), the schedules might be useful in dif-
ferentiating among the drugs. For example,
Procedure
under a delayed matching-to-sample sched-
After reduction to 80% of their free-feeding
ule, pentobarbital, diazepam, and phencycli-
body weights, the rats were trained to press the
dine decreased stimulus control, but mor-
right lever for reinforcer delivery. These ses-
phine had little effect (McMillan, 1981,
sions, and all subsequent sessions, began with
1982a). In the present study, we wished to de-
the rats placed in the operant chamber for a
termine whether the different levels of stim-
10-min presession period during which the
ulus control under these schedules can inter-
chamber was darkened and lever presses had
act with the different drugs to produce
no programmed consequences. After the com-
changes in responding that are distinctive for
pletion of this presession period, the house-
each drug.
lights were illuminated and the reinforcement
contingencies were placed in effect.
During the initial phase of training, the
METHOD
run lever (left lever) was retracted, and each
response on the reinforcement lever (right
Subjects
lever) produced a reinforcer. To advance to
Four adult male Sprague Dawley rats, ob-
each new phase of training, the rats had to
tained from Charles River, served as subjects.
meet the criterion of earning 50 reinforcers
They were individually housed in a large col-
within a 30-min session for three consecutive
ony room. The rats were reduced to, and
sessions. For the second phase of training,
maintained at, approximately 80% of their
the run lever was extended, and a single re-
free-feeding weights by supplemental feed-
sponse on this lever resulted in its retraction;
ings and had continuous access to water in
a subsequent response on the reinforcement

---

DRUGS AND STIMULUS CONTROL
119
lever produced the reinforcer. After reinforc-
mum of six sessions intervened prior to the
er delivery, the run lever was reinserted into
testing of a new drug.
the operant chamber. Once the criterion for
Dose–response determinations were con-
advancing to the third phase had been met,
ducted for sodium pentobarbital (Sigma
the number of consecutive run-lever re-
Chemical Co.), diazepam free base (Hoffman
sponses required for each reinforcer delivery
La Roche), phencyclidine (PCP) hydrochlo-
was rapidly increased. Each time the required
ride (National Institute on Drug Abuse), and
number of consecutive responses on the run
morphine sulfate (Mallinckrodt). Pentobar-
lever was emitted, this lever was retracted.
bital, phencyclidine, and morphine were dis-
When the reinforcer was delivered, the run
solved in physiological saline, and the doses
lever was reinserted into the chamber. If a rat
are expressed as the salt. Diazepam was dilut-
responded on the reinforcement lever prior
ed from a stock solution of 5 mg/ml in a ve-
to completing the response requirement on
hicle of 40% propylene glycol, 10% ethanol,
the run lever, the reinforcer was withheld and
and 50% physiological saline. The doses of
the response requirement was reset. Once
diazepam are expressed as the free base. The
the terminal requirement of eight consecu-
control injections for each drug consisted of
tive responses on the run lever was reached,
the appropriate drug vehicle.
the retraction of this lever was eliminated.
After the determination of the dose–re-
Thus, there were no external cues that sig-
sponse curves for the drugs, the schedule of
naled the completion of the run-lever re-
reinforcement was changed to a multiple FCN
sponse requirement of eight consecutive re-
FCN-SD schedule. To avoid confusion, the mul-
sponses.
tiple-schedule components will be referred to
The rats responded under the simple FCN
as the multiple FCN and FCN-SD components,
schedule until they consistently earned 50 re-
and the simple FCN schedule will be referred
inforcers within the 30-min session. For sub-
to by that name. The schedule parameters for
sequent sessions the data were inspected for
the FCN components of the multiple schedule
increasing or decreasing trends in the num-
were the same as those of the simple FCN
ber of runs (the number of times the rats
schedule. The FCN-SD component of the mul-
switched from the run lever to the reinforce-
tiple schedule was initiated by a tone and by
ment lever) and in the percentage of rein-
the illumination of the bank of lights above
forced runs. When no systematic trends in
the run lever. When the rats completed eight
these performance measures were found,
consecutive responses on the run lever during
drug testing was initiated.
Sessions were conducted in the mornings,
the multiple FCN-SD component, the lights
Sunday through Friday, with control sessions
above this lever were turned off. All other
(vehicle injection) conducted on Mondays
schedule conditions were the same as for the
and Thursdays and test sessions (drug injec-
multiple FCN component.
tion) conducted on Tuesdays and Fridays. If
Under the multiple schedule, the session
the subjects did not earn all 50 reinforcers
was terminated after 30 min had elapsed or
within the 30-min sessions on Sundays or
after 50 reinforcers had been earned. Each
Wednesdays, the control and drug sessions
session began with the multiple FCN com-
were discontinued until this criterion had
ponent in effect, with the schedule compo-
been met. Each session lasted for 30 min or
nents alternating after ?ve reinforcers had
until the rats had earned 50 reinforcers.
been earned (?ve reinforced runs) or after 5
Drug and vehicle administrations were by
min had elapsed. After response stability had
intraperitoneal injection. The volume of the
been attained, the rats consistently earned 25
drug and vehicle injections was 1 ml/kg of
reinforcers under each of the multiple-sched-
body weight. After injection, the rats were
ule components. The multiple schedule was
placed in the operant chamber for the pre-
in effect for 30 sessions prior to the analysis
session period. Each drug dose was adminis-
of systematic trends in number of runs and
tered once to each rat in a semirandom order
percentage of correct runs. When no increas-
such that ascending or descending dose or-
ing or decreasing trends in these measures
ders were avoided. After the completion of a
were evident for 10 consecutive sessions, the
dose–response curve for one drug, a mini-
drug testing phase was initiated. All condi-

---

120
SAMUEL H. SNODGRASS et al.
Table 1
tions for drug testing were the same as for
the simple FCN schedule.
Baseline percentage of reinforced runs under the FCN
schedules, presented as means with standard deviations
The dependent variables were the percent-
in parentheses.
age of reinforced runs, the number of re-
sponses emitted in each run (which allowed
Simple
Multiple
Multiple
for the determination of conditional proba-
Subject
FCN
FCN
FCN-SD
bilities), and the session response rate. The
R217
55.0 (7.1)
45.3 (3.0)
95.3 (2.1)
percentage of reinforced runs was de?ned as
R218
61.0 (5.6)
74.0 (4.4)
92.8 (2.4)
the number of runs in which the rats re-
R219
71.3 (4.6)
72.5 (3.9)
93.3 (1.2)
sponded at least eight consecutive times on
R221
41.0 (4.8)
81.3 (3.2)
90.3 (3.3)
M
57.1
68.3
92.9
the run lever, prior to switching to the rein-
forcement lever, divided by the total number
of runs. The session response rate was deter-
mined by dividing the total number of re-
each rat under each schedule. Session re-
sponses emitted on the run lever by the ses-
sponse rates were calculated by ?nding the
sion length in seconds for each rat. The
group mean for the session rates emitted un-
number of runs per run length was deter-
der each drug dose and control condition.
mined for the control as well as for the test
The conditional probability control data were
sessions, and these data were transformed to
the means of the control sessions for each rat
conditional probabilities. The conditional
under each FCN schedule. The conditional
probability measure, which is analogous to
probability control data presented for each
the IRT/Op measure described by Anger
drug condition were obtained by ?nding the
(1963), determines the conditional probabil-
mean conditional probability of the control
ity for switching to the reinforcement lever
sessions for each rat under each FCN sched-
after emitting a particular run length (Laties,
ule that occurred during the determination
1972). This measure adjusts for the differ-
of each drug’s dose–response curve. In cal-
ence in the number of opportunities among
culating conditional probabilities, the de-
the run lengths. For example, the opportu-
nominator is continually being diminished.
nity to switch to the reinforcement lever after
Conditional probabilities were not calculated
a run length of eight occurs only after eight
when fewer than 10 opportunities to respond
or more consecutive responses are emitted,
on the reinforcement key occurred for a run
whereas the opportunity to switch after a run
under the simple FCN schedule or fewer than
length of ?ve occurs for runs of ?ve or more
?ve opportunities remained under the mul-
consecutive responses. Thus, there are more
tiple FCN or multiple FCN-SD components.
opportunities to emit short, compared to
long, run lengths. The conditional probabil-
RESULTS
ity measure takes this discrepancy into ac-
count by dividing the number of times a run
Baseline Performance
length occurred by the number of opportu-
Table 1 shows that, under control condi-
nities the rats had to emit this run length.
tions, the mean percentage of reinforced
The conditional probability measure, there-
runs was higher under the multiple FCN
fore, provides a measure of the probability of
components than under the simple FCN
occurrence of each run length given that a
schedule; however, the effects varied across
response on the reinforcement lever has not
individual rats. The percentage of reinforced
prevented the opportunity for that run
runs for Rat R217 was lower under the mul-
length to occur (cf. Anger, 1963).
tiple FCN component than under the simple
The run length of 16 was the upper re-
FCN schedule, whereas those for Rats R218
corded run-length limit. If the rats emitted
and especially R221 were higher and that of
runs of more than 16 consecutive responses,
R219 was about the same. In contrast, the
they were recorded in the same terminal bin
percentage of reinforced runs under the mul-
as the run length of 16. The baseline per-
tiple FCN-SD component was higher than that
centage of reinforced runs was determined
under both the simple FCN schedule and the
by ?nding the mean percentage of reinforced
multiple FCN component for all rats.
runs emitted during the control sessions by
Figure 1 shows the conditional probability

---

DRUGS AND STIMULUS CONTROL
121
Fig. 1.
The mean conditional probability control data for the simple FCN schedule (open triangle), the multiple
FCN component (closed square), and the multiple FCN-SD component (open circle), for each of the 4 rats. These
data are the means of the control sessions for each rat obtained under each FCN schedule.
of switching from the run lever to the rein-
FCN schedule were slightly to the right of
forcement lever as a function of run length
those emitted under the FCN component of
under all schedules for each rat. For all rats,
the multiple schedule at the longer run
under all schedules, the conditional proba-
lenths.
bility of a switch to the reinforcement lever
generally increased as the length of the run
Dose–Response Cur ves for Percentage
increased. For Rats R217, R218, and R219,
of Reinforcement
the conditional probability of terminating a
Dose–response curves are shown for indi-
run short of the minimum required for re-
vidual rats and for the group mean for all
inforcement (eight consecutive responses)
drugs under all schedules in Figure 2. Only
under the FCN-SD component of the multiple
decreases in the percentage of reinforced
schedule was less than that obtained under
runs were obtained for the group means, al-
the other schedules, while for Rat R221 the
though occasionally individual animals
conditional probability curves were similar
showed small increases in the percentage of
under both components of the multiple
reinforced runs.
schedule. For Rat R221 and to a lesser extent
Responding under the simple FCN sched-
for Rat R218, the conditional probability
ule was affected at the lowest doses of diaze-
curves for the simple FCN schedule were to
pam, with 1 rat no longer responding after
the left of the curves of the multiple-schedule
receiving 1.0 mg/kg diazepam and the other
components, suggesting an increased proba-
2 rats showing a decrease in the percentage
bility of a switch to the reinforcement key af-
of reinforced runs. Under the FCN and FCN-
ter shorter runs under the simple FCN
SD components of the multiple schedule, re-
schedule. For Rats R217 and R219 the con-
sponding was well maintained after receiving
ditional probability curves for the simple
1.0 mg/kg diazepam in all rats and the per-

---

122
SAMUEL H. SNODGRASS et al.
Fig. 2.
The percentage of reinforced runs (% RFD RUNS) for the simple FCN schedule (left column), the
multiple FCN component (middle column), and the multiple FCN-SD component (right column). The effects of the
four drugs are shown with the order of presentation being diazepam (top row), morphine (second row), pentobarbital
(third row), and PCP (bottom row). Each subject is represented by a unique symbol: open circle for R217, ?lled
circle for R218, open square for R219, and ?lled square for R221. The mean dose–response values are represented
by the solid lines. To be consistent, the data of each of the 4 rats are included in the mean values even when a rat
did not emit any reinforced runs (i.e., 0% reinforced runs). The absence of a rat’s symbol indicates that the rat
failed to emit a reinforced run at that dose. The control values are shown on the left side of each graph, with the
vertical lines representing
1 SD and the horizontal line showing the group mean. Because of a lack of behavioral
stability, diazepam was not administered to Rat R219 under the simple FCN schedule.
centage of reinforced runs was only slightly
dose of morphine had greater effects on re-
affected. At 1.7 mg/kg diazepam, responding
sponding maintained under the simple FCN
was eliminated under both multiple-schedule
schedule than under either component of
components in 3 of the 4 rats.
the multiple schedule. The 5.6 mg/kg dose
The effects of morphine were qualitatively
of morphine also produced a greater effect
similar to those of diazepam. The 3.0 mg/kg
under the simple FCN schedule in that re-

---

DRUGS AND STIMULUS CONTROL
123
sponding in all rats was eliminated, whereas
probabilities under the simple FCN schedule
under the multiple-schedule components the
were not striking (Figure 3). For Rat R217,
responding of only 1 rat was eliminated at
the administration of the 0.56 mg/kg dose
this dose.
produced a decrease in the probability of
Unlike the differential effects of diazepam
ending a run too soon and an increased
and morphine, the effects of pentobarbital
probability of switching after 10 consecutive
on responding were similar across all three
responses on the run lever. The conditional
schedule conditions. Rat R221, however, did
probabilities of Rat R218 were little affected
fail to respond after the 5.6 mg/kg dose un-
at doses of diazepam that did not eliminate
der the simple FCN schedule. For PCP, the
its responding under the simple FCN sched-
effects on responding of the 1.7 mg/kg dose
ule. Rat R221 showed somewhat increased
were only slightly greater under the simple
conditional probabilities of run lengths of
FCN schedule than under either component
eight or less at doses of 0.56 mg/kg and 1.0
of the multiple schedule. Although the 3.0
mg/kg diazepam and a decreased probability
mg/kg dose of PCP suppressed the respond-
of run lengths of 10 responses after the 0.3
ing of the other rats under each of the FCN
and 0.56 mg/kg doses under the simple FCN
schedules, the responding of Rat R219 was
schedule. The minimal effect of diazepam on
not affected by PCP.
the conditional probabilities emitted by these
Effects of Drugs on Conditional
rats is largely due to the suppression of re-
Probability Distributions
sponding at the higher doses of this drug.
Under the FCN component of the multiple
The effects of the drugs on the conditional
schedule, one or more doses of diazepam
probability distributions are shown in Figures
produced increases in the conditional prob-
3 through 6. In these ?gures, loss of stimulus
ability of short runs (seven or fewer consec-
control is shown by a shift of the conditional
utive responses) for all rats except R221. The
probability curve to the left or by an increase
conditional probability of long runs (eight or
in the conditional probability of some, or all,
more consecutive responses) was decreased
of Bins 1 through 7. Such effects would be
in all rats, although the effects on the re-
expected to lower reinforcement rates. Loss
sponding of Rat R217 were minimal com-
of stimulus control might also be shown by a
pared to the effects in the other rats. Except
decrease in the conditional probability of run
for increases in the probability of run lengths
lengths of more than seven consecutive re-
of four to eight responses after the 0.56
sponses, although this effect would not nec-
essarily lower reinforcement rate. These two
mg/kg dose of diazepam, the conditional
effects might be combined, and examination
probabilities of Rat R217 were little affected
of Figures 3 through 6 shows that both effects
by diazepam under the multiple FCN-SD com-
were sometimes observed in the same rat at
ponent until responding was eliminated at
the same dose. It is also possible that a drug
the dose of 1.8 mg/kg. For Rats R218 and
might improve stimulus control under an
R219, the doses of 0.56 (Rat R218 only) and
FCN schedule. This is shown by a decrease in
1.0 mg/kg diazepam produced a shift to the
the conditional probabilities of runs of seven
right in the conditional probability function
or fewer consecutive responses or an increase
such that, for these rats, only long run
in the conditional probability of runs of more
lengths were emitted at this dose. For Rat
than seven consecutive responses. These ef-
R221 the 1.0 mg/kg dose increased the con-
fects would generally increase the percentage
ditional probability of short runs and the 0.56
of reinforced responses. Such effects were
mg/kg dose decreased the probability of lon-
rare.
ger runs.
Diazepam. Because the responding of Rat
Morphine. Morphine, at the doses that pro-
R219 was not stable when the simple FCN
duced an effect, typically caused an increase
schedule dose–response curve for diazepam
in the probability of short run lengths (Fig-
was determined, the data for this rat were not
ure 4). Although this was the usual effect of
included in the dose–response determina-
morphine, for Rat R217 under the multiple
tion. At doses that did not eliminate respond-
FCN component and for Rat R218 under the
ing, the effects of diazepam on conditional
simple FCN schedule, the conditional prob-

---

124
SAMUEL H. SNODGRASS et al.
Fig. 3.
The effect of diazepam on the conditional probability of switching from responding on the run lever to
the reinforcement lever. The simple FCN schedule is represented in the left column, and the multiple FCN com-
ponent and the multiple FCN-SD component are represented by the middle and right columns, respectively. Each
row presents the data for 1 rat, with the order being R217, R218, R219, and R221 from the top to the bottom row.
The mean conditional probability control data are represented by the solid line, and the vertical lines show
1 SD.
The ordinate represents the conditional probability, and the abscissa represents the run length. Diazepam was not
administered to Rat R219 under the simple FCN schedule because its behavior was not stable at that time.

---

DRUGS AND STIMULUS CONTROL
125
Fig. 4.
The effect of morphine on the conditional probability of switching from responding on the run lever to
the reinforcement lever. Details are as in Figure 3.
abilities of long run lengths were decreased
the suppression of responding of Rats R219
at some doses.
and R221 at higher doses and the suppres-
The typical effect of morphine to increase
sion of responding of all the rats at the dose
the conditional probability of short run
of 5.6 mg/kg. Under the FCN component of
lengths occurred under each FCN schedule,
the multiple schedule, the increased proba-
although this effect was somewhat less fre-
bility of short runs can be observed as a shift
quent under the simple FCN schedule due to
to the left in the conditional probabilities for

---

126
SAMUEL H. SNODGRASS et al.
Rats R218 and R219, with this shift usually
PCP. PCP produced increases in the con-
occurring over two or more dose levels. A
ditional probability of switching to the rein-
similar shift to the left in the conditional
forcement lever after runs of fewer than eight
probabilities for Rats R217, R218, and R219
responses under the simple FCN schedule in
can be observed under the multiple FCN-SD
each rat after the 1.7 mg/kg dose (Figure 6).
component, with the most consistent effect
This effect was especially pronounced for
occurring at the 5.6 mg/kg dose. Although,
Rats R217 and R218. For Rat R221, the small
for the most part, morphine did not alter the
increase in the conditional probabilities of
conditional probability distributions pro-
short run lengths observed after the 1.7
duced by Rat R221, the responding of this rat
mg/kg dose of PCP may have been due, at
was eliminated by doses of morphine that al-
least partly, to the dominance of short run
tered, but did not eliminate, the responding
lengths in the conditional probability base-
of the other rats.
line. Only Rat R219 showed a reliable de-
Pentobarbital. The effects of pentobarbital,
crease in the conditional probability of runs
as shown in Figure 5, were more like those of
of eight or more responses after PCP admin-
diazepam than those of morphine. There
istration under the simple FCN schedule. In
were few effects of pentobarbital under the
contrast, under the FCN component of the
simple FCN schedule at doses that did not
multiple schedule, decreases in the condi-
severely suppress responding; however, when
tional probability of run lengths of eight or
effects occurred the conditional probabilities
more responses were observed after one or
at run lengths of eight or more responses
more doses of PCP in all rats. Increases in the
were usually decreased in Rats R217 and
conditional probability of runs of fewer than
R218, and Rat R219 showed some increases
eight responses were also observed in most
in the conditional probability of run lengths
rats, but the effects were generally smaller un-
of eight or more responses, especially after
der the FCN component of the multiple
the 1.0 mg/kg dose. Under the multiple-
schedule than under the simple FCN sched-
schedule components, there was an increase
ule. Under the FCN-SD component of the
in the conditional probability of short runs
multiple schedule the responding of Rat
and a decrease in the conditional probability
R217 was little affected, whereas PCP pro-
of long runs for Rats R218 and R219 and an
duced only increases in the probability of
increase in the length of the longest run
shorter runs by Rat R221. The responding of
emitted at the doses of 3.0 and 5.6 mg/kg.
Rat R218 showed both increases in the con-
For R221, there was an increase in the con-
ditional probability of short runs and decreas-
ditional probability of long run lengths at the
es in the conditional probability of long runs.
5.6 mg/kg dose of pentobarbital, although
Rat R219 showed both increases and decreas-
the longest run emitted at this dose was short-
es in the conditional probability of long runs.
er than the longest run emitted under the
control conditions. Few consistent effects oc-
Effects of Drugs on Rate of Responding on
curred for Rat R217. Pentobarbital produced
the Run Lever
few effects on responding under the FCN-SD
Table 2 shows the mean session response
component of the multiple schedule, except
rates under the control conditions and after
for some increases in the conditional proba-
drug administration for each schedule. This
bility of responses in the early bins after re-
table shows that, under the control condi-
ceiving the 5.6 mg/kg dose. Under the FCN
tions, there were no signi?cant differences in
component of the multiple schedule, the 5.6
session response rates within the three sched-
mg/kg dose slightly increased the conditional
ules. It appears that the session response rates
probability of responding in the early bins for
under the morphine control conditions
all rats except R217. The 3.0 mg/kg dose
might have been higher than those of the
(Rats R217 and R218) and the 5.6 mg/kg
other drugs, particularly for the multiple
dose (Rats R218 and R219) decreased the
FCN component. However, because the stan-
conditional probability of responding at or
dard errors overlap, it is unlikely that there is
near the bin associated with the minimum
a true difference in response rates among
number of responses on the run lever re-
these control conditions. The data in Table 2
quired to produce the reinforcer (Bin 8).
also show that doses of the drugs were ad-

---