between auricle and ventricle has been produced by crushing or

6I2,

I72.2

THE

RELATION

BETWEEN

AMPLITUDE

OF

CONTRACTION

AND

RATE

OF

RHYTHM

IN

THE

MAMMALIAN

VENTRICLE.

(INCLUDING

INTERPRE-

TATION

OF

THE

APPARENT

INDIRECT

ACTION

OF

THE

VAGUS

ON

AMPLITUDE

OF

VENTRICULAR

CONTRACTION

.)1

By

ALISON

S.

DALE.

(From

the

Physiological

Laboratory,

Cambridge.)

PART

I.

THE

APPARENT

INDIRECT

ACTION

OF

THE

VAGUS

ON

THE

AMPLITUDE

OF

VENTRICULAR

CONTRACTION.

Introductory.

THE

question

whether

the

ventricle

of

the

mammalian

heart

receives

inhibitory

fibres

from

the

vagus

nerve

has

produced

much

conflicting

evidence.

It

is

a

well-established

fact,

that,

after

complete

dissociation

between

auricle

and

ventricle

has

been

produced

by

crushing

or

cutting

of

the

A.-v.

bundle

[Krehl

and

Romberg,

1892;

Hering,

1905;

Erlanger,

1909],

stimulation

of

the

vagus

nerve

has

no

effect

on

the

rhythm

of

the

ventricle

or

on

the

strength

of

its

contractions.

Injection

of

vago-mimetic

drugs,

however,

still

produces,

in

the

rabbit,

a

slight

slowing,

suggesting

a

weak

negative

chronotropic

action

of

the

vagus

on

the

ventricle

itself

[Krehl

and

Romberg,

1892;

Cullis

and

Tribe,

1913],

though

the

distribution

of

the

fibres

concerned

is

unknown.

With

regard

to

the

inotropic

action

of

the

vagus

on

the

ventricle

the

evidence

is

conflicting.

Cullis

and

Tribe

showed

that,

in

the

per-

fused

intact

heart

in

situ,

stimulation

of

the

vagus

nerve

or

injection

of

vago-mimetic

drugs

slowed

the

heart

rhythm,

and

at

the

same

time

decreased

the

strength

of

contraction

of

both

auricle

and

ventricle.

When

the

bundle

was

cut

the

effects

on

the

ventricle

of

stimulating

the

nerve

and

of

injection

of

drugs

were

abolished,

except

for

the

slight

slowing

1

This

work

was

carried

out

during

the

tenure

of

the

Gilchrist

Studentship

of

Newnham

College,

and

the

Michael

Foster

Studentship

of

the

University

of

Cambridge.

A.

S.

DALE.

produced

by

relatively

large

doses.

More

recent

work

by

Drury

[1923]

and

Rothberger

and

Scherf

[1930]

on

the

whole

animal

(dog)

has

shown

that,

if

the

ventricular

rate

is

maintained

constant

by

rhythmic

shocks

throughout,

vagus

stimulation

does

not

influence

the

strength

of

the

ventricular

contractions.

Cullis

and

Tribe,

finding

that

the

vagus

or

vago-mimetic

drugs

only

affected

the

strength

of

contraction

of

the

ventricle

when

the

latter

was

connected

to

the

auricle

by

the

A.-V.

bundle,

concluded

that

"normally,

the

vagus

exerts

its

effect

on

the

ventricle,

only

indirectly,

through

its

action

on

the

auricular

rhythm."

The

present

experiments

were

carried

out

in

an

attempt

to

determine

what

the

nature

of

this

"indirect"

action

might

be.

Methods.

The

experiments

were

performed,

with

one

exception,

on

the

hearts

of

rabbits.

In

the

one

exception

a

cat's

heart

was

used.

The

animals

were

killed

by

a

blow

on

the

head

and

the

hearts,

after

removal

from

the

body,

were

perfused

with

a

modified

Ringer's

solution

by

means

of

a

cannula

tied

into

the

aorta.

The

cannula

carried

a

thermometer,

and

the

solution

entered

by

a

side

tube.

The

perfusion

pressure,

which

was

about

1

metre

of

saline,

was

maintained

from

a

Mariotte's

bottle,

suspended

at

a

suitable

height.

From

the

Mariotte's

bottle

the

solution

passed

to

a

glass

tower

where

oxygen

was

continuously

bubbled

through

it,

and

thence

to

a

glass

coil

immersed

in

a

water

bath.

The

coil

was

connected

by

rubber

tubing

to

the

side

tube

of

the

cannula.

The

temperature

in

the

water

bath

was

regulated

so

that

the

thermometer

in

the

cannula

registered

36-380

C.

The

composition

of

the

Ringer's

solution

was

as

follows:

NaCl

0-85

p.c.,

Na2HPO4

0-06

p.c.,

KCI

0-042

p.c.,

dextrose

0.1

p.c.,

CaCl2

0x024

p.c.

The

pH

was

adjusted

to

7x5

by

the

addition

of

HCI,

comparison

being

made

with

a

standard

buffer

solution,

using

phenol

red

as

indicator.

The

water

used

in

making

up

the

solution

was

distilled

in

a

porcelain

still.

Great

difficulty

was

experienced

initially

as,

although

the

auricles

beat

quite

vigorously,

the

ventricles

contracted

only

very

feebly

or

not

at

all.

This

was

found

to

be

due

to

the

presence

in

the

distilled

water

of

a

high

concentration

of

CO2.

If

the

water,

after

condensation,

was

thoroughly

boiled

and

allowed

to

cool

before

use

the

hearts

beat

very

well.

It

was

found

later

that

this

precaution

of

using

boiled

water

for

the

solution

was

only

necessary

at

the

beginning

of

an

experiment.

When

the

heart

was

once

beating

vigorously,

solution

containing

C02

could

be

used

without

having

any

deleterious

effect.

The

reason

for

this

effect

of

002

456

HEART

RATE

AND

SIZE

OF

CONTRACTION.

is

not

understood,

but

it

would

seem

to

be

specific

and

not

due

to

its

acid

properties,

as

the

reaction

of

the

solution

was

always

controlled.

The

drug

used

in

most

of

the

experiments

was

arecoline,

though

acetylcholine

was

occasionally

used.

The

latter

is

not

very

satisfactory

in

these

experiments,

however,

as

its

effect

is

too

transient.

The

arecoline

was

obtained

as

the

crystalline

bromide

and

made

up

in

a

0-1

p.c.

solution.

This

was

diluted

with

Ringer's

solution

to

0.001

p.c.

for

each

experiment.

The

usual

dose

was

0-1

c.c.

to

0-5

c.c.

of

this

diluted

solution,

injected

with

a

syringe

into

the

rubber

tubing

leading

to

the

cannula.

The

contractions

of

the

left

auricle

and

left

ventricle

were

recorded

by

means

of

light

straw

levers

carrying

B

ayli

ss

writing

points.

The

levers

were

held

in

brass

bearings,

and

the

vibration

frequency

of

each

was

increased

by

a

small

rubber

band

which

resisted

the

raising

of

the

lever.

The

heart

was

suspended

vertically

from

the

cannula,

and

the

apex

of

the

left

ventricle

was

fixed

by

a

ligature

to

a

metal

rod.

One

lever

was

attached

to

a

point

near

the

base

of

the

ventricle

by

a

silk

thread

sewn

through

the

epicardium,

the

other

to

the

left

auricular

wall

in

the

same

way.

By

this

method

of

attachment

the

use

of

pulleys

for

the

ventricular

record

is

avoided,

and

the

risk

of

distortion

of

the

record

by

movements

(swinging,

etc.)

of

the

whole

heart

is

greatly

diminished.

It

has

the

further

advantage,

of

special

importance

in

these

experiments,

that

the

shortening

of

the

ventricular

muscle

is

recorded,

uncomplicated

by

the

pull

of

the

auricular

contraction.

In

order

to

avoid

distension

of

the

left

ventricle

by

perfusion

fluid

leaking

through

the

aortic

valves,

a

glass

tube

was

passed

into

the

cavity

of

the

ventricle

through

a

small

slit

made

in

its

apex,

and

tied

in

position.

Any

fluid

tending

to

collect

in

the

ventricle

was

thus

drained

away.

Results.

Effects

of

vago-mimetic

drugs.

The

object

of

the

first

experiments

was

to

repeat

the

experiment

of

Cullis

and

Tribe.

The

effects

of

varying

doses

of

arecoline

on

the

heart

with

A.-v.

bundle

intact

were

first

recorded.

The

bundle

was

then

cut

and

the

procedure

repeated.

Fig.

1

gives

a

typical

example

of

the

records

obtained.

It

will

be

noticed

that

in

record

a,

taken

before

cutting

the

bundle,

the

injection

of

0.001

mg.

of

arecoline

is

followed

by

a

slowing

of

the

heart

rhythm,

accompanied

by

a

diminution

in

the

amplitude

of

the

auricular

and

the

ventricular

contractions.

A

curious

point

is

the

rise

of

the

base

line,

which

is

particularly

noticeable

in

the

ventricular

tracing.

The

reason

for

this

rise

is

not

clearly

understood,

but

in

any

case

it

is

457

A.

S.

DALE.

unconnected

with

the

diminution

in

the

amplitude

of

the

contractions,

as

it

occurs

after

the

bundle

is

cut,

when

no

such

diminution

is

seen.

Fig.

1.

Exp.

3.

xii.

29.

Read

from

right

to

left.

In

both

a

and

b

the

upper

tracing

is

the

ventricular

reoord,

the

lower

that

of

auricle

followed

by

ventricle.

a.

Intact

heart.

b.

After

cutting

the

bundle.

Further,

when

another

method

of

registration,

which

will

be

described

later,

is

used,

this

rise

of

the

base

line

does

not

occur,

while

the

diminution

in

the

size

of

the

ventricular

contractions

is

still

observed.

It

will

be

noticed

that

premature

contractions

are

followed

by

greatly

enlarged

ventricular

beats.

The

significance

of

this

will

be

considered

later.

Record

b

shows

the

effect

of

arecoline

after

section

of

the

A.-v.

bundle.

The

dose,

0.01

mg.,

is

sufficient

to

inhibit

the

auricular

contractions

almost

completely,

while

the

ventricular

rate

is

slightly

diminished.

A

close

scrutiny

reveals

a

very

slight

decrease

in

the

height

of

the

ven-

tricular

record,

but

this

decrease

is

very

much

smaller

than

that

observed

before

cutting

the

A.-v.

bundle.

The

results

just

described,

which

are

typical

of

many

obtained,

are

identical

with

those

of

Cullis

and

Tribe,

except

for

one

small

difference.

Cullis

and

Tribe

never

obtained

a

diminution

in

amplitude

of

the

ventricular

contractions

when

vago-

mimetic

drugs

were

injected

after

section

of

the

A.-v.

bundle.

In

the

458

HE"ART

RATE

AND

SIZE

OF

CONTRACTION.

present

experiments

this

diminution

was

only

obtained

with

large

doses

of

arecoline.

Normal

doses

had

no

such

effect.

Cullis

and

Tribe's

conclusion

still

holds,

therefore,

for

normal

doses,

while

for

large

doses

we

may

qualify

it

by

adding

that

arecoline

affects

the

size

of

the

ven-

tricular

contractions

greatly

when

the

A.-V.

bundle

is

intact,

and

slightly

when

the

ventricle

is

no

longer

in

connection

with

the

auricle.

With

regard

to

the

nature

of

this

"indirect"

action,

it

might

be

supposed

that

the

diminution

in

the

size

of

the

ventricular

contractions

was

due

to

a

change

in

the

impulse

passing

by

the

A.-v.

bundle

from

auricle

to

ventricle.

If

that

were

the

case

the

phenomenon

would

present

a

clear

exception

to

the

"All

or

None"

law.

Such

a

supposition,

however,

is

unnecessary,

as

a

much

simpler

explanation

is

at

hand.

Close

inspection

of

the

records

of

the

present

series

of

experiments

shows

that

the

ventricular

contrac-

tions

only

decreased

in

size

after

injection

of

arecoline

when

the

heart

rhythm

was

slowed,

which

suggested

that

the

change

of

rate

is

in

some

way

responsible

for

the

change

in

the

size

of

the

contractions.

This

is

supported

by

the

fact

that

the

idio-ventricular

rhythm

which

obtains

after

section

of

the

A.-V.

bundle

is

only

slightly

influenced

by

arecoline,

and

that

in

the

ventricle

isolated

in

this

manner

the

size

of

the

beat

remained

unchanged

by

any

dose

of

arecoline

which

did

not

affect

the

rate.

It

is

evident

that

if

the

diminution

in

the

amplitude

of

the

ven-

tricular

contractions,

which

follows

injection

of

arecoline

when

the

bundle

is

intact,

is

simply

due

to

the

slowing

of

the

rhythm,

then

a

similar

diminution

should

occur

when

the

rhythm

is

slowed

by

any

other

means.

Effects

of

cooling

the

S.-A.

node.

A

series

of

experiments

was,

therefore,

performed

in

which

the

heart

rate

was

decreased

by

cooling

of

the

S.A.

node.

The

preparation

was

set

up

as

before,

and

the

cooling

was

effected

by

means

of

a

piece

of

narrow

lead

pipe

through

which

iced

water

flowed.

The

pipe

was

bent

sharply

at

one

point,

and

the

outer

surface

of

the

bend

was

brought

into

contact

with

the

region

of

the

S.-A.

node,

thus

producing

localized

cooling.

Fig.

2a

shows

a

typical

example

of

the

result

of

slowing

the

rhythm

by

this

method;

b

and

c

show

the

effect

of

two

different

doses

of

arecoline

for

comparison.

In

c

a

2-1

block

was

produced

by

the

drug,

the

ventricular

rhythm

thus

slowing

to

about

the

same

degree

as

that

produced

by

cooling

in

a,

and

it

will

be

seen

that

the

diminution

in

the

size

of

the

ventricular

contractions

is

prac-

tically

the

same

in

the

two

cases.

Effects

of

drugs

at

constant

rhythms.

Such

results

give

very

strong

support

to

the

hypothesis

that

the

slowing

of

the

rhythm

is

the

deter-

459

mining

factor

in

the

action

of

arecoline

on

the

size

of

the

ventricular

contractions

in

the

intact

heart.

It

now

remained

to

be

determined

Fig.

2.

Exp.

13.

xii.

29.

Read

from

right

to

left.

In

a,

b

and

c

upper

tracing

represents

auricular

followed

by

ventricular

contractions.

Lower

tracing

ventricular

contractions

only.

a.

S.-A.

node

cooled

between

points

indicated

by

arrows.

b.

0002

mg.

arecoline

injected

shortly

before

beginning

of

record.

c.

0005

mg.

arecoline

injected

shortly

before

beginning

of

record.

Arrow

indicates

beginning

of

2-1

block.

whether

the

size

of

ventricular

contractions

would

be

unaffected

by

arecoline

when

the

heart

rate

was

maintained

artificially

constant.

The

heart

was,

therefore,

stimulated

rhythmically

by

electrodes

placed

on

the

right

auricle.

The

rhythmic

shocks

were

produced

by

means

of

a

rotary

contact

breaker

placed

in

the

primary

circuit

of

an

induction

coil.

Fig.

3

shows

the

results

of

such

an

experiment,

and

includes

a

record

of

slowing

by

cooling

for

comparison.

Fig.

3

a

shows

the

usual

result

of

cooling

the

S.-A.

node:

b

shows

the

A.

S.

DAlB.

460

HEART

RATE

AND

SIZE

OF

CONTRACTION.

461

effect

of

injecting

a

small

dose

of

arecoline

while

the

heart

was

being

driven

at

constant

rate.

The

dose

0002

mg.

is

one

which,

with

spontaneous

Fig.

3.

Exp.

11.

ii.

30.

Read

from

left

to

right.

In

a,

b

and

c.

Upper

tracing

represents

auricular

followed

by

ventricular

contractions.

Auricular

contractions

are

smaller

than

ventricular,

but

their

height

can

be

seen

as

the

upper

edge

of

the

denser

portion

of

the

tracing.

Lower

tracing

is

that

of

ventricular

contractions

only.

a.

Cooling

of

S.-A.

node.

Auricular

tracing

disappears

owing

to

substitution

of

nodal

for

sinus

rhythm.

b.

Heart

driven

at

constant

rate.

Injection

of

0'002

mg.

arecoline

at

arrow.

c.

Heart

driven

at

constant

rate.

Injection

of

0

005

mg.

arecoline

at

first

arrow.

Second

arrow

indicates

appearance

of

2-1

block.

rhythm,

would

produce

a

marked

slowing,

with

weakening

of

the

ven-

tricular

contraction.

In

this

case

the

strength

of

contraction

is

unaffected.

Record

c

provides

especially

clear

evidence

on

the

point

under

discussion.

The

dose

was

a

large

one,

0005

mg.,

which

almost

abolished

the

auricular

contractions

at

one

point.

Later,

when

the

auricular

beats

were

re-

PH.

LXX.

30

A.

S.

DALE.

covering,

the

effect

on

the

A.-V.

conducting

system

became

evident,

and

a

2-1

block

was

produced.

The

amplitude

of

the

ventricular

contrac-

tions,

which

till

then

had

remained

unchanged,

immediately

diminished

in

size,

and

this

reduction

of

amplitude

persisted

as

long

as

the

2-1

block

was

maintained.

The

period

of

block

was

followed

by

a

short

phase

in

which

an

occasional

beat

was

dropped.

After

this

the

block

disappeared

completely,

a

ventricular

contraction

following

each

auricular

one,

and

with

this

resumption

of

the

original

rhythm

by

the

ventricle,

the

con-

tractions

returned

to

their

normal

size.

Effects

of

varying

an

artificial

rhythm.

It

seems

clear

from

these

observations

that

arecoline

is

unable

to

exert

its

negative

inotropic

effect

on

the

ventricle

when

the

bundle

is

intact,

unless,

at

the

same

time,

it is

able

to

diminish

the

heart

rate.

It

might

still

be

argued,

however,

that

the

auricle

affects

the

ventricle

in

some

way

other

than

by

change

of

rhythm

when

arecoline

is

injected,

though

this

is

unlikely

in

view

of

the

last

experiments

described

(Fig.

3).

In

order

to

eliminate

this

possibility,

a

few

experiments

were

performed,

in

which

the

bundle

was

cut

and

the

ventricle

driven

at

varying

speeds

without

the

administration

of

any

drug.

Fig.

4,

which

is

an

example

of

the

results

obtained,

is

a

record

taken

from

a

cat's

heart.

The

heart

showed

a

spontaneous

block

and,

as

the

left

ventricle

was

beating

feebly,

the

record

was

taken

from

the

right

ventricle,

which

contracted

vigorously

and

responded

well

to

artificial

stimulation.

The

rotary

contact breaker

mentioned

above

was

used,

and

the

rate

of

stimulation

was

varied

by

changing

the

resistance

in

series

with

the

motor

driving

the

contact

breaker.

I

a

Fig.

4.

Exp.

23.

i.

30.

Read

from

right

to

left.

Ventricle

of

cat

driven

at

various

rates.

a.

Acceleration

from

96

to

120

per

min.

b.

Retardation

from

130

to

120

per

min.

Fig.

4

a

shows

the

result

of

changing

the

rate

of

stimulation

from

96

to

120

per

min.,

namely,

a

definite

increase

in

the

size

of

the

contrac-

tions.

The

change

is

gradual,

as

the

motor

took

a

little

time

to

attain

462

HEART

RATE

AND

SIZE

OF

CONTRACTION.

the

new

speed.

Fig.

4

b

shows

the

change

in

the

reverse

direction,

from

130

to

120

per

min.,

the

result

being

a

diminution

in

the

size

of

the

contractions.

It

will

be

noticed

that

the

size

of

the

contractions

at

130

per

min.

in

record

b

is

smaller

than

that

at

120

per

min.

in

record

a.

This

is

probably

a

fatigue

effect,

since,

when

the

rate

was

increased

from

120

to

130

per

min.,

there

was

an

increase

in

the

size

of

the

contractions

at

first,

but

this

gave

way

to

a

gradual

diminution.

It

seems

probable

that

the

heart

could

not

maintain

such

a

vigorous

contraction

at

this

high

rate

when

depending

only

on

the

oxygen

dissolved

in

the

Ringer's

solution.

When

the

stimulation

rate

was

reduced

again

to

96

per

min.

the

contractions

gradually

regained

their

original

size

corresponding

to

this

rate.

It

is

possible,

therefore,

to

vary

the

size

of

the

ventricular

contraction

by

changing

the

rhythm,

when

this

is

being

produced

artificially,

in

exactly

the

same

way

as

when

the

ventricle

is

receiving

its

stimuli

from

the

auricle

by

way

of

the

A.-v.

bundle.

Experiments

with

perfused

ventricular

strips,

to

be

described

later,

confirm

this.

A

point

which

is

important

for

later

discussion

must

be

mentioned

here.

It

is

evident

in

all

curves

in

which

there

is

an

abrupt

slowing

of

the

heart-rate,

that

the

beat

following

the

first

prolonged

pause

is

en-

larged,

and

that

this

is

then

followed

by

a

descending

staircase

of

beats,

until

the

small

size

of

beat,

typical

of

the

slower

rate,

is

reached.

The

change

from

slow

to

fast

rate

is

usually

more

gradual,

but

where

an

abrupt

change

occurs,

as

in

Fig.

2

c,

where

a

2-1

block

gives

way

to

the

normal

A.-V.

sequence,

it

is

evident

that

the

beat

following

the

first

short

pause

is

diminished,

and

is

followed

by

an

ascending

staircase

of

beats.

Demonstration

of

effect

in

tension

records.

From

the

results

of

the

experiments

so

far

described

it

must

be

concluded

that,

in

the

isolated

perfused

rabbit's

heart

in

which

the

contractions

are

recorded

by

means

of

levers,

the

so-called

indirect

effect

of

vago-mimetic

drugs

on

the

size

of

the

ventricular

contractions

is

due

to

the

slowing

of

the

rhythm

produced

by

the

action

of

the

drug

on

the

pacemaker

of

the

heart.

Experiments

with

levers,

however,

are

always

open

to

criticism

on

the

ground

that

distortioA

of

the

records

may

occur,

owing

to

the

natural

vibration

frequency

of

the

levers

being

too

low.

In

the

case

of

a

rapidly

beating

rabbit's

heart

it

is

essential

that

the

recording

system

have

a

high

natural

frequency,

and

it

might

well

be

argued

that,

in

the

experi-

ments

above

described,

the

excursions

of

the

lever

give

a

true

record

of

the

size

of

the

contractions

only

when

the

rate

is

slow,

the

larger

excursions

at

higher

rates

of

beating

being

due

to

overthrow

of

the

lever.

30-2

463

In

order

to

eliminate

such

an

objection

the

experiments

were

repeated,

using

a

high-frequency

system

for

recording

the

tension

changes

in

the

ventricular

muscle.

Fig.

5.

Diagram

of

apparatus

for

recording

intraventricular

pressure.

a,

brass

tube

carrying

membrane;

b,

rubber

membrane

with

mirror

attached

radially;

c,

connecting

tube

of

lead

piping;

d,

brass

tube

to

which

balloon

e

is

attached;

f,

5

c.c.

pipette

containing

liquid

paraffin

for

filling

balloon;

A

and

B,

3-way

(T)

taps.

Fig.

5

is

a

diagrammatic

representation

of

the

apparatus

used.

It

consisted

essentially

of

a

balloon,

filled

with

fluid

which

was

introduced

into

the

cavity

of

the

left

ventricle,

and

connected

by

tubes

filled

with

fluid

to

a

tightly

stretched

rubber

membrane.

A

mirror

attached

to

the

membrane

reflected

a

beam

of

light

into

a

moving

paper

camera.

Diffi-

culty

was

experienced

at

first

in

the

choice

of

a

suitable

balloon.

Small

thin

rubber

balloons

were

tried

but,

unless

these

were

blown

to

extreme

tightness,

the

contraction

of

the

ventricle

caused

the

top

of

the

balloon

to

bulge

into

the

auricular

cavity.

In

the

process

of

this

bulging

a

large

amount

of

distortion

of

the

record

could

occur.

A

tightly

stretched

rubber

balloon

has

two

disadvantages,

the

first

that

it

bursts

very

easily,

464

A.

S.

DAlE.

HEART

RATE

AND

SIZE

OF

CONTRACTION.

and

the

second

that

the

initial

pressure

is

so

high

that

the

ordinary

glass

taps

cannot

stand

it,

and

leakages

occur.

It

was

decided,

therefore,

to

use

an

inextensible

balloon,

and

for

this

purpose

the

gall

bladder

of

a

small

cat

served

very

well.

The

bladder

was

removed

from

the

animal

and

washed

out

with

saline.

It

was

then

bound

firmly

to

a

small

piece

of

brass

tube,

and

stored

in

saline,

covered

with

a

layer

of

toluene

to

keep

it

sterile.

When

in

use

the

bladder

was

filled

with

liquid

paraffin.

Saline

is

unsuitable

as

it

diffuses

through

the

walls

of

the

bladder

when

this

is

subjected

to

pressure.

The

perfusion

apparatus

was

the

same

as

that

used

in

the

experiments

already

described.

The

bladder

was

intro-

duced

through

a

small

slit

in

the

apex

of

the

left

ventricle,

this

being

done

as

follows.

The

bladder

and

brass

tube

were

filled

with

paraffin

by

means

of

a

fine

glass

pipette,

care

being

taken

to

exclude

all

air

bubbles.

The

5

c.c.

pipette

(Fig.

5,f)

was

also

filled

with

paraffin

and

placed

in

position.

The

taps

A

and

B

were

then

turned

so

that

the

paraffin

flowed

from

the

pipette

(i)

and

filled

the

tubes

and

the

rubber

con-

a

nection

for

the

brass

tube

carrying

the

bladder.

The

brass

tube

was

then

in-

(ij)

serted

into

the

rubber

connection,

care

again

being

taken

to

avoid

the

entrance

of

air

bubbles.

The

paraffin

was

then

sucked

back

into

the

pipette,

and

the

(i)

balloon,

which

was

quite

flexible,

was

thus

inverted

into

the

brass

tube.

The

b

end

of

the

tube

could

now

easily

be

(ii)

inserted

into

the

ventricle,

and

was

tied

firmly

in

place.

Finally,

by

blowing

FiE

from

the

end

of

the

pipette,

the

bladder

was

everted

into

the

ventricular

cavity,

and

filled

to

any

desired

extent.

It

was

then

connected

with

the

membrane

by

turning

the

tap

B.

The

observations

repeated

with

this

method

of

recording

were:

cooling

of

I

the

S.-A.

node,

simple

injection

of

areco-

line,

and

injection

of

arecoline

while

the

heart

was

being

driven

at

constant

rate.

JVyA.J.

/\.A

I.

.

g.6.

Read

from

left

to

right.

Sample

curves

taken

from

two

experiments.

z.

From

Exp.

3.

v.

30.

b.

From

Exp.

6.

v.

30.

a

(i).

On

left,

heart

beating

at

normal

rhythm.

On

right,

luring

cooling

of

S.-A.

node.

a

(ii).

On

left,

heart

beating

at

normal

rhythm.

On

right,

after

injection

of

)0001

mg.

arecoline.

b

(i).

On

left,

ieart

beating

at

normal

rhythm.

)n

right,

after

injection

of

0-002

mg.

Lrecoline.

b

(ii).

On

left,

heart

driven

Lt

constant

rhythm.

On

right,

after

njection

of

0-003

mg.

arecoline.

A

2-1

block

has

appeared

and

ven-

ricular

rate

is

halved.

Fig.

6

shows

a

series

of

records,

taken

from

two

experiments,

illustrating

the

three

methods

of

slowing.

465

A.

S.

DALE.

It

is

evident

from

these

records

that

the

decreased

size

of

ventricular

contraction

produced

by

slowing

of

the heart

rhythm

is

not

an

apparent

one,

due

to

defects

in

the

recording

apparatus,

but

a

real

one,

as

it

still

persists

in

records

taken

with

an

apparatus

in

which

all

possibility

of

overthrow

is

eliminated.

The

rise

of

the

base

line

during

the

action

of

arecoline,

so

evident

in

the

records

taken

with

the

levers,

is

entirely

absent

in

these

records,

and

may

be

taken

to

be

an

artefact

peculiar

to

the

method

of

registration.

Conclusionsfrom

Part

I.

The

conclusions

to

be

drawn

from

the

experiments

described

in

Part

I

of

this

paper

may

be

summarized

as

follows:

The

results

of

other

workers,

which

show

that

vago-mimetic

drugs

such

as

arecoline,

in

small

doses,

only

influence

the

size

of

the

ventricular

contraction

when

the

A.-v.

bundle

is

intact,

have

been

confirmed.

It

has

further

been

shown

that

this

decrease

in

amplitude

is

a

real

one,

and

is

not

produced

by

imperfect

methods

of

registration.

That

the

effect

is

only

observed

when

the

ventricle

is

connected

with

the

auricle

is

due

to

the

fact

that,

in

this

case,

the

heart

rhythm

is

greatly

slowed

by

the

drugs

in

question,

whereas

the

idio-ventricular

rhythm,

which

sets

in

after

bundle

section,

is

only

affected

to

a

very

small

degree.

The

theory

that

it

is

the

slowing

of

the

rhythm

which

determines

the

diminished

size

of

the

ventricular

contractions

is

supported

by

the

facts

that

the

drug

does

not

affect

the

ventricle

when

the

heart

rate

is

kept

constant

by

rhythmic

auricular

stimulation,

and

that

when

the

dose

is

large

enough

to

produce

heart

block,

the

slowing

of

the

ventricular

rhythm

which

results

produces

a

diminution

in

the

size

of

the

contrac-

tions.

Further,

a

slowing

of

the

rhythm

produced

by

cooling

of

the

S.-A.

node

produces

effects

on

the

ventricle

identical

with

those

observed

after

injection

of

arecoline.

That

the

inotropic

effect

is

truly

secondary

to

the

chronotropic

effect,

and

does

not

depend

on

some

obscure

auricular

influence,

is

shown

by

the

fact

that

it

can

be

produced

in

the

isolated

ventricle

when

this

is

stimulated

artificially

and

the

rate

of

stimulation

is

varied.

466

HEART

RATE

AND

SIZE

OF

CONTRACTION.

PART

II.

THE

RELATION

BETWEEN

AMPLITUDE

OF

CONTRACTION

AND

RATE

OF

RHYTHM.

The

relation

between

the

rate

of

beating

and

size

of

contraction

which

has

been

described

in

Part

I

is

not

the

generally

accepted

one.

For

the

mammalian

heart,

at

least,

it

is

usually

stated

that

the

con-

tractions

are

more

vigorous

at

slow

rates

than

at

fast

ones,

the

customary

explanation

being

that

at

the

slow

rates

the

muscle

has

more

time

for

recovery

between

beats.

The

fact

that

the

relation

so

stated

is

the

exact

opposite

of

that

described

in

Part

I

of

this

paper

must

be

due

to

dif-

ferences

in

experimental

method.

There

appear

to

be

two

methods

of

experiment

in

which

increased

amplitude

with

slowing

of

the

rhythm

would

be

observed.

The

first

is

that

in

which

the

whole

animal

is

employed,

and

the

heart

is

left

in

situ.

Slowing

of

the

rhythm,

produced,

for

example,

by

cooling

of

the

S.-A.

node,

will

allow

increased

time

for

filling

of

the

ventricles

during

diastole

with

consequent

greater

length

of

ventricular

fibres

at

the

beginning

of

systole.

The

second

method

uses

the

isolated

perfused

heart,

in

which

slowing

of

the

rhythm

is

produced

by

cooling

of

the

perfusion

fluid.

In

this

way,

not

only

is

the

pacemaker

cooled,

thus

slowing

the

rhythm,

but

the

ventricular

muscle

is

also

cooled.

Cooling

of

the

heart

muscle

is

known

to

produce

larger

con-

tractions,

even

when

the

heart

rate

is

kept

constant.

To

account

for

the

smaller

contractions

at

slower

rates,

observed

in

the

present

series

of

experiments,

two

possibilities

offer

themselves.

A

simple

and

easily

tested

hypothesis

is

based

on

Wiggers'

theory

of

fractionate

contractions.

It

is

quite

conceivable

that,

at

slow

rates,

the

impulse

would

travel

more

slowly,

so

that

fewer

muscular

fractions

would

be

in

action

at

any

one

moment,

and

the

records

would

show

curves

which

were

smaller

in

height,

but

proportionately

greater

in

duration,

than

those

found

at

high

rates.

The

other

possibility

is

that

the

phenomenon

is

a

property

of

the

cardiac

muscle

itself.

(1)

The

former

hypothesis

was

tested

first.

If

it

were

correct,

then

the

duration

of

a

contraction

should

always

show

a

distinct

increase

at

the

slower

rates.

The

measurements

were

made

on

the

optical

records

which

had

been

taken

on

fast

moving

paper

(see

Fig.

6).

In

many

cases

the

decrease

of

amplitudes

with

decrease

of

rate

was

accompanied

by

no

increase

in

duration.

In

a

few

cases

the

duration

showed

a

slight

increase

under

these

conditions;

but

this

was

never

great

enough

to

467

account

for

decrease

of

amplitude

by

slowing

of

transmission,

and

in

many

cases,

notably

those

in

which

arecoline

was

used,

the

slight

increase

of

duration

persisted

after

return

to

the

original

rhythm

and

amplitude.

It

cannot,

therefore,

be

connected

with

the

slowing

of

rhythm

and

loss

of

amplitude,

and

is

probably

a

fatigue

effect

due

to the

filling

of

the

ventricular

cavity

with

the

tense

bladder.

An

unlikely

possibility

still

remained,

namely,

that

at

the

slower

rates

the

impulse

might

travel

by

a

different

path,

and

so

alter

the

shape

and

size

of

the

ventricular

mechanogram.

Rothberger

and

Scherf

[1930]

showed

that

a

change

in

the

point

of

application

of

artificial

stimuli

to

the

ventricle

in

the

whole

animal,

changed

both

the

electro-

cardiogram

and

the

intra-ventricular

pressure

curve.

To

test

this

possi-

bility,

electrical

records

were

taken'.

The

hearts

were

set

up

for

perfusion

in

the

usual

way,

and

simultaneous

records

were

taken

of

the

electro-

cardiogram

and

the

tension

developed

by

the

ventricular

muscle,

the

latter

by

the

balloon

method

already

described.

Slowing

was

produced

by

cooling

of

the

S.-A.

node,

and

by

injection

of

acetylcholine.

Arecoline

is

unsuitable

for

these

observations,

as

it

produces

permanent

changes

in

the

form

of

the

electrocardiogram,

especially

in

the

T-wave,

which

persist

after

the

effect

on

the

mechanogram

has

disappeared.

Measure-

ment

of

the

records

showed

that,

when

slowing

of

the

heart

rate

was

produced

by

cooling

the

S.-A.

node

or

injecting

acetylcholine,

with

a

corresponding

diminution

of

amplitude,

in

many

cases

the

electro-

cardiogram

was

unchanged.

In

some

there

was

a

slight

alteration

in

the

T-wave,

but

in

no

case

was

there

any

change

in

the

initial

QRS

complex,

which

is

determined

by

the

spread

of

the

excitatory

process.

These

experiments

rule

out

the

possibility

that

the

smaller

contractions

at

slow

rates

are

due

to

changes

in

the

rate

or

the

path

of

propagation

of

the

impulse.

(2)

It

seemed

probable,

accordingly,

that

the

phenomenon

was

due

to

a

property

of

the

cardiac

muscle

itself,

and,

to

test

this,

experiments

were

carried

out

on

perfused

ventricular

strips.

Rabbits'

hearts

were

used

as

before.

Strips

of

rabbits'

ventricle

have

the

advantage

that

they

do

not

beat

spontaneously,

so

that

no

difficulty

was

experienced

in

imposing

artificial

rhythms

on

them.

Fig.

7

gives

a

diagrammatic

representation

of

the

apparatus

used.

The

Ringer's

solu-

tion

was

of

the

same

composition

as

that

used

in

the

experiments

on

the

whole

heart,

and

the

arrangements

for

oxygenating

and

warming

it

were

the

same

as

before.

The

cannula,

a

glass

tube

bent

twice

at

right

angles,

1

I

am

much

indebted

to

Dr

A.

N.

Drury

for

his

help

in

the

taking

of

these

records.

A.

S.

DALE.

468

HEART

RATE

AND

SIZE

OF

CONTRACTION.

was

drawn

out

into

a

nozzle,

sufficiently

fine

for

insertion

into

one

of

the

coronary

arteries

through

its

opening

out

of

the

aorta.

A

silk

ligature,

g

coiL

Fig.

7.

Apparatus

for

perfusing

isolated

ventricular

strips.

a,

strip

of

ventricular

muscle;

b,

stimulating

electrodes;

c,

recording

lever

(only

part

shown).

passed

under

the

coronary

artery

with

a

curved

needle,

served

to

tie

the

nozzle

of

the

cannula

firmly

in

position.

Strips

of

right

ventricle

were

tried

at

first,

as

it

was

expected

that

they

would

survive

better;

but

it

was

found

that

the

left

ventricle

was

preferable,

probably

owing

to

the

fact

that

the

left

coronary

artery

is

the

larger,

and

allows

a

bigger

flow

of

perfusion

fluid.

In

any

case,

the

flow

was

never

large

enough

to

maintain

the

temperature

of

the

muscle

much

above

room

temperature.

It

was,

therefore,

necessary

to

immerse

the

strip

in

a

bath

of

the

Ringer's

solution.

This,

in

turn,

was

immersed

in

a

beaker

of

water,

warmed

to

the

desired

temperature

by

a

micro-burner.

The

free

end

of

the

strip

was

attached

by

a

silk

thread

to

a

lever,

similar

to

those

used

in

the

first

experiments

on

whole

hearts.

The

strip

was

stimulated

rhythmically,

one

of

the

electrodes

dipping

in

the

bath

of

Ringer's

solution,

and

the

other

being

attached

to

the

upper

free

end

of

the

strip.

The

latter

electrode

was

made

of

fine

wire

and

the

lead

from

it

was

coiled,

so

that

it

was

freely

movable,

and

did

not

impede

the

movement

of

the

strip.

The

rhythmic

stimuli

were

produced

by

means

of

a

rotary

contact

breaker,

similar

to

that

used

in

previous

experiments,

but

having

two

rotating

arms

connected

in

parallel

in

the

primary

circuit

of

an

induction

coil.

Either

arm

could

thus

complete

the

circuit,

and

they

were

so

469

adjusted

that

the

rhythm

of

shocks,

produced

when

both

were

in

action,

was

halved

by

cutting

one

out.

The

initial

rate

of

stimulation

was

con-

trolled

by

a

variable

resistance

included

in

series

with

the

motor

driving

the

contact

breaker.

Fig.

8.

Exp.

11.

vi.

30.

Read

from

left

to

right.

Record

obtained

with

perfused

strip

of

left

ventricle.

Stimulated

rhythmically

at

a

rate

of

50

per

min.

At

first

arrow

rate

of

stimulation

was

halved.

At

second

arrow

it

returned

to

normal.

Fig.

8

gives

a

typical

example

of

the

results

obtained.

It

will

be

noticed

that,

when

the

rate

of

the

stimulating

shocks

is

halved,

the

beat

after

the

-first

longer

pause

is

enlarged,

and

that

the

following

beats

gradually

decrease

in

size

until

a

minimum

is

reached.

When

the

original

rhythm

is

resumed

the

beat

following

the

first

shorter

interval

is

dis-

tinctly

smaller,

but,

the

following

beats

at

the

more

rapid

rhythm

gradually

increase

in

size

until

they

are

as

large

as

those

initially

pro-

duced

at

that

rate.

This

series

of

events

agrees

in

every

detail

with

what

is

observed

in

the

whole

ventricle

when

the

rhythm

is

slowed.

Another

property,

common

to

the

whole

ventricle

and

the

strips,

is

observed

when

the

slowing

of

the

rhythm

is

long

enough

maintained.

When

the

slowing

persists

for

more

than

ten

or

fifteen

beats,

it

is

noticed

that

the

size

of

the

beats

tends

again

to

increase

gradually

from

the

mnimum,

though

it

does

not

attain

the

amplitude

characteristic

of

the

faster

rate.

This

was

found

to

be

much

more

distinct

in

some

whole

hearts

than

in

others,

and

the

same

was

true

of

the

strips.

After

a

secondary

increase

of

this

kind

the

return

to

the

faster

rhythm

produced

beats

which

were

supernormal

as

compared

with

those

recorded

before

slowing

took

place.

After

a

short

period

the

beats

returned

to

the

original

amplitude.

DiSCUSSION.

The

phenomenon

described

could

be

most

simply

explained

by

as-

suming

that

the

optimal

rhythm

for

rabbit's

ventricular

muscle,

that

is,

the

rhythm

at

which

the

contractions

are

maximal,

is

a

high

one;

in

fact

that

it

is

of

the

same

order

as

the

natural

sinus

rhythm.

A

slowing

of

such

a

rhythm

would

then

cause

the

contractions

to

become

smaller.

470

A.

S.

DALE.

HEART

RATE

AND

SIZE

OF

CONTRACTION.

An

example

of

such

a

high

optimal

rhythm

is

described

by

Mines

for

the

ventricles

of

certain

elasmobranch

hearts

at

temperatures

between

200

C.

and

250

C.

The

natural

rhythm

in

this

case

was

sub-optimal,

as

acceleration

by

artificial

stimulation

augmented

the

contractions.

There

is

a

striking

difference,

however,

between

the

records

obtained

by

Mine

s

and

those

obtained

in

the

present

work

on

rabbits'

hearts.

In

Mine

s'

experiments

the

first

beat

at

the

higher

rhythm

is

augmented,

and

is

followed

by

an

ascending

staircase

of

beats,

until

the

larger

size

charac-

teristic

of

the

faster

rhythm

is

reached.

When

the

artificial

stimulation

is

stopped

and

the

rhythm

returns

to

normal

the

beats

diminish

abruptly

to

the

smaller

size

without

any

intervening

staircase.

In

the

case

of

the

mammalian

ventricle,

as

has

been

already

described,

the

first

beat

at

an

accelerated

rhythm

is

always

diminished,

and

is

then

followed

by

an

ascending

staircase

of

beats

until

the

larger

size

of

contraction

is

attained.

On

the

other

hand,

when

the

rhythm

changes

from

a

high

to

a

low

one,

the

first

beat

at

the

slow

rhythm

is

augmented

and

is

then

followed

by

a

descending

staircase

until

the

smaller

size

of

beat

charac-

teristic

of

the

slow

rhythm

is

reached.

It

seems,

therefore,

that

there

is

some

other

factor

at

work

in

the

mammalian

ventricle

which

is

the

cause

of

these

transition

phenomena.

The

high

rate

of

beating

appears

to

exert

a

favourable

influence

on

the

contractility

of

the

muscle,

and

to

leave

some

"trace

"

behind

it,

which

causes

the

first

beat

at

the

slow

rhythm

to

be

augmented.

The

augmented

beat

occurring

after

a

pre-

mature

contraction

in

the

perfused

heart

could

be

explained

in

the

same

way.

It

is

worth

noting

here

that

a

premature

contraction

occurring

during

a

phase

of

slow

rhythm

with

the

resultant

small

beats

will

cause

the

beat

following

it

to

be

augmented

to

a

size

of

the

same

order

as

that

of

the

first

beat

at

the

slow

rhythm.

This

is

well

seen

in

the

ventricular

record

of

Fig.

1

a.

The

changes

in

amplitude

which

occur

when

the

rhythm

changes

from

a

slow

to

a

fast

one

do

not

fit

in

with

a

simple

explanation

based

on

an

optimal

rhythm

hypothesis.

If

we

were

concerned

with

a

simple

case

of

a

high

optimal

rhythm,

the

first

beat

at

the

higher

rhythm

should

be

enlarged,

as

in

Mines'

experiment.

In

fact

it

is

conspicuously

diminished,

and

is

followed

by

an

ascending

staircase

of

beats.

Here

again

it

seems

that

the

higher

rate

of

beating

exerts

some

favourable

influence

on

the

cardiac

muscle.

It

seems,

therefore,

that

we

cannot

define

these

high

rhythms

as

optimal

for

the

rabbit's

ventricular

muscle

in

the

generally

accepted

sense

of

the

term.

The

large

amplitudes

observed

at

the

high

rhythms

471

A.

S.

DALE.

appear

to

depend

on

a

cumulative

action

of

the

contractions

themselves.

Such

an

action

is

very

similar

to

that

described

by

B

o

wditch

from

the

frog's

ventricle,

and

it

must

be

concluded

that

we

are

dealing

here

with

a

form

of

Bowditch

staircase.

(Woodworth

[1903]

describes

a

similar

beneficial

action

of

rapid

rhythms

on

the

perfused

ventricular

apex

of

the

dog.

He

ascribes

the

enlargement

of

the

first

contraction

at

a

slowed

rhythm

to

the

combined

effects

of

this

cumulative

action

of

the

con-

tractions

and

the

prolonged

pause.)

This

explanation

is

admittedly

inconclusive.

Until

more

experi-

mental

results

are

available

it

is

impossible

to

draw

more

definite

con-

clusions.

It

is

hoped

that

further

experiments

will

make

a

more

exact

interpretation

possible.

So

far

the

relation

between

rate

and

size

of

contraction

has

only

been

discussed

with

regard

to

the

ventricle.

It

will

be

observed,

however,

that

in

the

experiments

in

which

the

S.-A.

node

was

cooled

the

auricular

contractions

diminished

in

size

when

slowing

took

place.

Great

stress

is

not

laid

on

these

results,

as

it

is

very

probable

that

a

certain

amount

of

overthrow

occurred

in

the

recording

of

the

auricular

contractions.

However,

a

recent

paper

by

H.

Loos

[1930]

supports

the

idea

that

the

relation

does

hold

for

the

auricle.

He

finds

that

the

size

of

auricular

contraction

in

the

rabbit's

heart

is

affected

more

profoundly

by

vagus

stimulation

when

the

rhythm

is

allowed

to

slow,

than

when

the

rate

is

maintained

constant

by

artificial

stimulation.

Loos'

experiments

were

performed

on

whole

rabbits,

but

the

phenomenon

was

still

observed

when

the

venae

cavae

were

clamped,

so

that

it

is

not

determined

by

factors

relating

to

the

filling

of

the

auricle.

SUMMARY.

1.

Cullis

and

Tribe's

observation,

that

vago-mimetic

drugs

only

diminish

the

amplitude

of

contraction

of

the

rabbit's

ventricle

when

the

A.-v.

bundle

is

intact,

is

confirmed.

2.

This

action

is

shown

to

depend

on

the

fact

that

the

S.-A.

nodal

rhythm

is

slowed

by

the

action

of

the

drug,

while

the

idio-ventricular

rhythm

which

obtains

after

section

of

the

A.-v.

bundle

is

not

affected,

similar

diminutions

of

amplitude

being

observed

when

the

rhythm

is

slowed

by

other

means.

3.

A

diminution

is

also

obtained

in

the

amplitude

of

contraction

of

perfused

ventricular

strips,

stimulated

artificially,

when

the

rate

of

stimulation

is

diminished.

472

HEART

RATE

AND

SIZE

OF

CONTRACTION.

473

4.

This

relation

between

the

amplitude

of

contraction

and

the

rhythm

is

ascribed

to

a

form

of

Bowditch

staircase.

In

conclusion

I

wish

to

thank

Prof.

Bar

croft

for

permission

to

carry

out

these

experiments

in

his

laboratory,

and

Dr

Anr

e

p

for

his

advice

and

criticism.

REFERENCES.

Cullis

and

Tribe

(1913).

J.

Physiol.

46,

141.

Drury

(1923).

Heart,

10,

405.

Erlanger

(1909).

Pflueger8

Arch.

127,

77.

Hering

(1905).

Pflueger8

Arch.

108,

281.

Krehl

and

Romberg

(1892).

Arch.

exp.

Path.

Pharmak.

30,

49.

Loos

(1930).

Pflueger8

Arch.

225,

1.

Rothberger

and

Scherf

(1930).

Z.

ge8.

exp.

Med.

71,

274.

Woodworth

(1903).

Amer.

J.

Phy8iol.

8,

213.