L-ent he e

ee re
eS a enn
wg ale “a = -
ld eo

= Gee ee bane ee
: - Se me tee

OE ie es
oS on ot

ams

eos _

o Bee we eS

: Se ee ae
see

bie ha

i " I _

¥ H aye ‘
A

W

Renee |

THE AMERICAN JOURNAL

OF

PHYSIOLOGY.

EDITED FOR

Che American ph psiological Society

BY
H. P. BOWDITCH, M.D., BOSTON FREDERIC S. LEE, Pu.D., NEW YORK
R- H. CHITTENDEN, Px.D., NEW HAVEN W. P. LOMBARD, M.D., ANN ARBOR
W. H. HOWELL, M.D., BALTIMORE G. N. STEWART, M.D., CHICAGO

W. T. PORTER, M.D., BOSTON

a ey

THE

AMERICAN JOURNAL

PriYyolQLOGY

VOLUME XX.

7 O
i A of
G b
| 0
BOSTON, U.S.A. ~|
GINN AND COMPANY
1907-1908.

a 4’ 4 ere : ee a ae
ar? a ae ak
AX : | :
\
A

Copyright, 1907-1908
By GINN AND COMPANY.

Guiversity IBress

Joun Witson AND Son, CAMBRIDGE, U.S.A.

CONTENTS:

No. I, Ocroper 1, 1907.

PAGE
THE MECHANICAL EFFECTS PRODUCED BY THE CONTRACTION OF INDI-

VIDUAL MUSCLES OF THE THIGH OF THE FroG. By Warren P.
nn ORE. A AEE og ce) os Ra ae ee ee 8 I

FURTHER OBSERVATIONS ON THE RESUSCITATION OF THE RESPIRATORY
NERVOUS MECHANISM. By G. NW. Stewart and F. H. Pike. . . O61

THE ACUTE EFFECTS OF GASTRIC AND PERITONEAL CAUTERIZATION
AND IRRITATION OF THE BLOOD PRESSURE AND RESPIRATION. Sy
Torald Sollmann, E. D. Brown, and W. W. Williams . . . . . 74

CHEMICAL STUDIES ON GROWTH.—I. THE INVERTING ENZYMES OF
THE ALIMENTARY TRACT, ESPECIALLY IN THE EMBRYO. Sy Lafay-
tite B. Mendel and Philip H. Mitchell. . . «© «© 2. «© w@ «© + « BI

CHEMICAL STUDIES ON GROWTH. —II. THE ENZYMES INVOLVED IN
PuURIN METABOLISM IN THE EmBryo. Sy Lazayette B. Mendel and
SESE sg ms 8 Se ie ee ay ee en 197

CHEMICAL STUDIES ON GROWTH. — III. THE OCCURRENCE OF GLYCO-
GEN IN THE Emsryo Pic. By Lafayette B. Mendel and Charles S.
EPROP TAM Are. waht, ee et) ne ee fe ee eS ee TE

THE INFLUENCE OF ELECTROLYTES AND OF CERTAIN OTHER CONDI-
TIONS ON THE OSMOTIC PRESSURE OF COLLOIDAL SOLUTIONS. By
RETR 2) EP Wa ae i ee, iene stiin tes ew ee Sal gins eR eT

THE ACTION OF NORMAL FATIGUE SUBSTANCES ON MusSCLE. By
CPE POON er ators be te We cite rn el a ck, 241 ea we Oe gO

v1 Contents.

THE RELATION BETWEEN THE BLOOD SUPPLY TO THE SUBMAXILLARY
GLAND AND THE CHARACTER OF THE CHORDA AND THE SYMPA-
THETIC SALIVA IN THE DOG AND THE Cat. By A. F. Carlson,

Fe ht. Greer, and FC. Beh cs ee OS ee
THE INNERVATION OF THE CEREBRAL VESSELS AS INDICATED BY THE
ACTION OF DruGs. By Carl 7. Wiggers . . «© «© «© «+ «© « =

THE NUTRITIVE VALUE OF GELATIN. —II. SIGNIFICANCE OF GLYCO-
COLL AND CARBOHYDRATE IN SPARING THE Bopy’s PROTEID. By
SONNAMNIUEIER «85 ep Ea 5 eM eon ae, MPS be eiaenies is: eee

PERISTALTIC RUSH. By Sicf) Meélizer and F. Auer 3. << <3 «ee

No. II, NOVEMBER 1, 1907.

THE ACID CONTROL OF THE PyLorus. By W. B. Cannon.

A FuRTHER STUDY OF THE ACTION OF MAGNESIUM SULPHATE ON THE
HEART. By Wm. deB. Macnider and S. A. Matthews.

ON THE SWELLING OF FisRiIN. By Martin H. Fischer and Gertrude
BOB E Nn oe Se a hee Pe PR ee, & oe ey on

ON THE USE OF BONE ASH WITH THE DIET, IN METABOLISM EXPERI-
MENTS ON Docs. By Matthew Steel and William F. Gites

A SIMPLE ELECTRICAL ANNUNCIATOR FOR USE IN METABOLISM Ex-
PERIMENTS, AND IN CONNECTION WITH FILTRATION, DISTILLATION,
AND SIMILAR OPERATIONS. By William H. Welker. . . . . .

AN IMPROVED ANIMAL HOLDER. By Gustave M. Meyer. . . « .

THE ELIMINATION OF RADIUM FROM NORMAL AND NEPHRECTOMIZED
ANIMALS. By William Salant and Gustave M. Meyer .

ON THE CHEMICAL NATURE OF PARANUCLEOPROTAGON, A NEW PRODUCT
FROM BRAIN. By Matthew Steel and William F. Gies. . . . .

THE EFFECT OF UNIFORM AFFERENT IMPULSES UPON THE BLOOD -

PRESSURE AT DIFFERENT LEVELS. By W. T. Porter

No. III, DECEMBER 2, 1907.

SOME OBSERVATIONS ON THE BEHAVIOR OF THE AUTOMATIC RESPIRA-
TORY AND CARDIAC MECHANISMS AFTER COMPLETE AND PARTIAL
ISOLATION FROM Extrinsic NERVE ImpuLsEs. By G. WN. Stewart

XANTHIN AS A CAUSE OF FEVER AND ITS NEUTRALIZATION BY SALI-
CYLATES: (By Arthur Rh. Mandel -i 7 ee ee

THE RELATION OF AFFERENT IMPULSES TO FATIGUE OF THE VASO-
MOTOR CENTRE. By W. T. Porter, H. K. Marks, and J. B. Swift, Jr.

PAGE

180

206

234
259

283

323

339

343

358

362

366

378

399

407

439

444

Contents.

No. IV, January 1, 1908.

FURTHER OBSERVATIONS ON THE RELATION BETWEEN BLOOD PRES-
SURE AND RESPIRATORY MOVEMENTS. By C. C. Guthrie and F. H.
Pike.

FURTHER STUDIES ON THE RELATION OF THE OXYGEN SUPPLY OF THE
SALIVARY GLANDS TO THE COMPOSITION OF THE SALIVA. Sy A. /.
Carlson and F. C. McLean : :

HYDROLYSIS OF AMANDIN FROM THE ALMOND. Ay Thomas B. Osborne
and S. H. Clapp

HYDROLYSIS OF THE PROTEINS OF MAIzE, ZEA Mays. By Thomas B.
Osborne and S. H. Clapp

THE HyDROLYsIS OF GLIADIN FROM Rye. By Thomas B. Osborne and
S. H. Clapp

PAGE

457

470

477

494

FURTHER DATA REGARDING THE CONDITION OF THE VASOMOTOR NEURONS -

IN “SHocK.” By W. 7. Porter and W. C. Quinby

INDEX

500

507

THE

A merican Journal of Physiology.

VOE. XX. OCTOBER 1, 1907. NO. I.

THE MECHANICAL EFFECTS PRODUCED BY THE CON-
TRACTION OF INDIVIDUAL MUSCLES OF THE THIGH
OF THE FROG.

By WARREN P. LOMBARD anp F. M. ABBOTT.
[from the Physiological Laboratory of the University of Michigan.]

INTRODUCTION.

ANY years ago one of the writers of this paper attempted,
under the guidance of Professor Carl Ludwig, to determine the
method of spread of reflex processes in the spinal cord of the frog.!
In the course of this work he was compelled to recognize that many
of the movements of the hind limb of the frog which have the appear-
ance of being the result of finely adjusted nervous co-ordinations, are
really due to the mechanical conditions under which the muscles act
on the bones. The lesson was forced home, that before we can arrive
at reliable conclusions as to the method of action of the central ner-
vous mechanisms concerned in reflex actions, we must obtain a clear
picture of the mechanics of the limb itself.

In a recent paper? it has been shown that not only the effective-
ness of the contraction of a given muscle alters with the gain and
loss of leverage which comes as the bones have their position changed,
but that the character of the movement which a given muscle may
produce may be greatly altered, indeed in some cases reversed, by a
change in the relative positions of the bones on which the muscle in

1 Die raiimliche und zeitliche Aufeinanderfolge reflectorisch contrahirter
muskeln. Archiv fiir Anatomie und Physiologie, 1885, p. 408.

2 “ The tendon action and leverage of two-joint muscles of the hind leg of the
frog with special reference to the spring movement.” Contributions to Medical
Research, dedicated to Victor Clarence Vaughan, University of Michigan, June

1903, p. 280. :

2 Warren P. Lombard and F. M. Abbote.

question acts. Thusa muscle which in one position of a bone may act
as a flexor, in another position of the bone may act as an extensor,
and a muscle which in one position of a bone may carry it dorsally,
in another position may carry it ventrally. In the same paper the
remarkable action of the two joint muscles was pointed out, and it
was shown how muscles which act as antagonists in certain positions
of the bones may in other positions act as synergists. It is evident
that we can form no estimate ‘of the part played by the central ner-
vous system in co-ordinated movements of locomotion, for example,
until we have ascertained in how far these movements are determined *
by the purely mechanical conditions existing in the limbs.

Just as it is unsafe to draw any conclusions concerning central co-
ordination before the mechanics. of the limbs have been exhaustively
studied, so it is unsafe to draw conclusions concerning the combined
action of the muscles until the method of action of the separate
muscles has been ascertained. Nor can even this be done profitably
until it has been found in how far the character of the movements of
the bones is determined by the structure of the joint surfaces and
ligaments, for these largely decide not only the limits but the direc-
tions in which the bones shall move under the influence of muscular
contractions.

In the course of the experiments reported in this paper, the action
of the structures of the joints to determine the direction and to limit
the extent of the movements of the bones was examined with con-
siderable care, but was not made the subject of exhaustive investiga-
tion. It is believed, however, that sufficient work was done in this
direction to permit us to recognize how far the motion which results
from the contraction of a muscle is caused by the action of the muscle
itself, as distinguished from the passive influence exerted by the
structures of the joint.

This research does not deal with the action of muscles under physi-
ological conditions, but with the mechanical effects of the shortening
of individual muscles.

All of the infinite number of positions which the bones entering
into the hip and knee joints can take with respect to each other can
be reached by appropriate muscular action. It would be, however,
impossible to determine experimentally the exact method of action of
each of the muscles which takes part in placing the leg in each of
these innumerable positions, or to ascertain what would be the action
of each of the muscles when the leg occupied each of these positions.

Individual Muscles of the Thigh of the Frog. 3

In this research the attempt was made to find out what kinds of
movements could be caused by contraction of the muscles of the
thigh when the bones entering into the hip and knee joints occupied
certain definite positions. In the case of the muscles which act on
the hip, for example, when the pelvis was fixed and the thigh was free
to move, under each of the following conditions, namely, —

a. When pelvis and femur were placed horizontally, and the femur

was without rotation and was flexed, half extended, and extended.
6. When the pelvis was horizontal, and the femur had been carried
either ventrally or dorsally out of the horizontal plane a desired num-
ber of degrees, without rotation, and was flexed, half extended, and
extended. . '

c. When the pelvis had been tilted dorsally or ventrally a desired
number of degrees, and the femur which was horizontal and without
rotation, was flexed, half extended, and extended.

_ d. The effect of rotation of the femur about its long axis was also
noted in the above positions of the bones.

The effect on the pelvis when the femur was fixed and the pelvis
was free to move was also studied.

A similar plan of work was carried out in studying the action of the
‘muscles of the thigh on the knee joint.

When the ability of a given muscle to flex and extend the thigh or
lower leg, to carry it dorsally and ventrally, to rotate it dorsally and ven-
trally, and to rotate the pelvis about the transverse axis, has been ascer-
tained, when the bones occupy the above positions, a fairly accurate
picture of the mechanical action of the muscle, when acting alone, is
obtained. Under normal conditions muscles never act independently,
all ordinary movements being the resultant of the synchronous action
of many muscles. In some few cases,the effect of the combined action
of two or more muscles was studied, but in general this research was
limited to studying the action of individual muscles when acting alone.

Terms used to describe the muscles and the movements which they
produce. — One of the greatest difficulties which we encountered at the
beginning of our work was to settle upon suitable terms for describing
the direction, character, and extent of the movements produced by the
muscles, and before we enter upon a description of our experiments it
will be necessary for us to define the terms which we employ.
Gaupp states, in his admirable revision of Ecker’s and Wieder-
_ heim’s ‘‘ Anatomie des Frosches,” ; his reason for the choice of the

1 GaupP: Braunschweig, 1896, pp. viii and ix.

4 Warren P. Lombard and F. M. Abbott.

names which he has applied to the muscles. His point of view is well
chosen, and his nomenclature will be followed in this paper. Short
names and names the same as those given to homologous human
muscles are much more convenient to the physiologist than long com-
pounds giving the origin and insertion, or purporting to describe the
function, of the muscles. Indeed the function of the muscles is so
imperfectly understood that the last method is very undesirable.
Although Gaupp will be followed in the naming of the muscles, we
shall have to depart somewhat from the terminology which he
employs for the description of the movements which they produce.
It is unfortunate that we cannot merely adopt the terms made use of
by the human anatomist. The body of the frog is propelled with its
long axis approximately horizontal, and that of man with its long
axis nearly vertical. The body of the frog is supported by the hind
limbs only when the animal is squatting and during the early part

of the leap, z. e,, when the legs are considerably flexed; the body |

of man has to be supported by the legs not only in the above posi-
tions, but when the legs are extended, by standing and ordinary
movements of locomotion. This difference in the direction of the
movement of the body and the way it is supported, is associated with
a difference in the position of the limbs with respect to the trunk.
The structure of the hip joint and the attachment of the muscles
surrounding it, in the case of the frog, is such that the thigh and
consequently the whole leg when in the extended position, compared
with the human leg, is rotated outward through about a quarter of a
circle. As Gaupp states,’ ‘By extended leg, z. ¢., put as far back
as possible in the direction of the prolongation of the long axis of
the body, the surface of the lower leg which corresponds with the
front surface by man faces laterally; the back of the lower leg in
the median direction; the lateral side dorsally; and the median side
ventrally.”

As a result of this difference in the position of the leg with respect
to the trunk, movements of the leg, which, from the consideration of
the muscles involved, may be classed as the same for frog and man,
take place in different directions. For example, flexion of the hind
leg both of the frog and man causes adjacent segments of the limb to
approach, and extension to separate, but occurs in general in the case
of the frog in a horizontal plane, and in the case of man in a verti-
cal plane. Abduction and adduction carry the human leg in the lat-

1 GaupP: Braunschweig, 1896, p. 189.

—_—

Individual Muscles of the Thigh of the Frog. 5

eral and the median directions, but corresponding movements carry
the leg of the frog dorsally and ventrally. Rotation of the leg of the
frog, in the sense which is termed external rotation by man, turns the
leg dorsally, and internal rotation turns it ventrally.

It is not easy to choose satisfactory terms to describe the move-
ments of the hind leg of the frog, and the terms employed in this
paper were decided upon only after the character of the movements
had been carefully studied, and many attempts had been made to
describe them with accuracy. In giving these definitions, unless it
be otherwise stated, the body of the frog will be considered to be
placed horizontally, with the dorsum upwards.

The initial position.—In order to state clearly the position which
the bones hold at a given time, and the movements which they
undergo as a result of the action of the muscles, it is necessary to
establish for each of the bones three definite axes and planes. This
must be done arbitrarily. The femur, for example, has a double S-
shaped curvature, and the peculiar relation of the heads of the bone
to the shaft is such that there are no definite points on the bone
which can be taken to fix these axes. It was decided, therefore, to
place the bones in some special position, and then establish a long,
a vertical, and a horizontal axis for each of them. The position
first chosen was the resting, semi-extended position produced by
the elastic tension of the soft parts. This position, however, was
found to vary somewhat in different frogs and under different con-
ditions, and it was finally found best to place the leg in a position
which differs but little from this, and to which we have given the
name “initial position.” In this position the pelvis-leg preparation
lies, dorsum upwards, with the iliac wings, the femur, os cruris, and
tarsus horizontal, and the femur forming a right angle with the long
axis of the pelvis, the os cruris a right angle with the femur, and the
tarsus a right angle with the os cruris.

In the initial position the following axes may be established for
the pelvis and the bones of the leg. These axes are not to be con-
sidered as axes about which the bones rotate, but as axes which can
be referred to in describing the movements of the bones.

The axes of the pelvis. — The long axis passes through a point mid-
way between the centres of the sacro-iliac joints, and a point midway
between the centres of the acetabula. In all of our measurements,
however, the long axis of the iliac wing was used. This is represented
by a line passing through the centre of the proximal end and the

6 Warren P. Lombard and F. M. Abbott.

centre of the point of origin of the glutzeus magnus on the superior
process of the wing ilium. The ¢vansverse axis passes through the
centres of the acetabula, cuts the long axis at a right angle, and is
parallel to a line connecting the centres of the sacro-iliac joints.
The sagittal axis passes through the point of intersection of the
other two axes, and is perpendicular to them.

The axes of the femur. — The long axis of the femur passes through
the centre of the proximal end, and the centre of the distal end of
the shaft of the bone. The ¢vansverse axis passes through the centre
of the shaft of the femur at its distal end, lies in the horizontal plane,
and forms a right angle with the long axis. The sagzttal axis passes
through the point of intersection of the other two axes, and forms a
right angle with each of them. It has been found possible, by means
of a special arrangement described under “ Method of Work,” to
bore a hole through the shaft of the femur in the direction of the
sagittal axis and to insert a slender needle in this hole. The needle
then represents the sagittal axis of the femur, and is of the greatest
assistance in the determination of the amount that the femur is
rotated about its long axis.

The axes of the os cruris. — These axes are established in the same
manner as those of the femur.

The axes of the tarsus. — The long axis passes through the centres
of the extremities of the bone. The sagittal axis passes through the
distal extremity of the bone in the direction of the long diameter of
the end of the bone, and makes a right angle with the long axis. In
the initial position it is not vertical, but it is inclined about 30°; that
is, the bone can be considered to be rotated ventrally 30°, clockwise,
as seen from the distal end of the bone. The ¢vansverse axis passes
through the centre of the distal extremity of the bone, and forms a
right angle with the other two axes.

The frontal, transverse, and sagittal planes of the pelvis, femur, os
cruris, and tarsus. — The frontal plane passes through the long and
transverse axes of the bone in question; the transverse plane passes
through the transverse and sagittal axes; and the sagittal plane passes
through the sagittal and longitudinal axes.

Flexion and extension.— Gaupp uses the terms flexion and ex-
tension when speaking of the ordinary bending and straightening of
the limb at the knee, but describes similar movements at the hip as
abduction (for flexion) and adduction (for extension).

When the leg is thrust out or drawn up to the body in swimming,

:
i

Individual Muscles of the Thigh of the Frog. 7

the thigh and lower leg move caudad and cephalad in practically the
same plane. This being so, it does not seem advisable to describe
the movements of the thigh as adduction and abduction, and move-
ments of the crus of a similar type as extension and flexion. More-
over, the term adduction does not serve to describe the whole course
of the thigh as it is thrust away from the trunk; it first leaves, and
then approaches, the long axis of the body, is first abducted and later
adducted. The terms abduction and adduction should not be used
for any of the movements of the thigh of the frog; they are not
necessary and are apt to cause confusion.

Gad says:! “ The drawing up of the thigh to the position ready
for the spring, because the thigh is rotated outwards so far that the
knee moves in the transverse plane, resembles abduction in man, but
is to be regarded as flexion because carried out by muscles homolo-
gous to the flexors in man.” The point is well taken, but the move-
ment is more complex than Gad ‘describes it. In taking the squatting
position, the limb does not simply move in what he describes as the
transverse plane; during the last part of the movement it is carried
ventrally, so as to bring the thigh somewhat under as well as side of
the belly. Gaupp describes this oblique movement, in which the knee
is carried cephalad and ventrally at the same time, as ventral-flexion,
and the reverse movement, which straightens the limb, as extension.
This usage is found to be very inconvenient when one attempts to
give a detailed statement of the action of individual muscles.

We have found it best to describe movements of the thigh which
cause it to approach the pelvis, and movements of the crus. which
cause it to approach the thigh, as flexion, and movements which cause
these parts to separate as extension. We use these terms regardless
of the plane in which the movements occur. If, in the course of flex-
ion or extension at the hip, the thigh is carried out of the horizontal
plane, this fact can be indicated by stating that it is carried ventrally
or dorsally at the same time that it is flexed or extended. In general
it is more convenient to speak of the more distal portion of the limb
as being flexed or extended with respect to the more proximal, ¢. ¢.,
to speak of the thigh as being flexed on the pelvis. As a matter
of fact the same effect is produced in a joint regardless of which of the
two parts is fixed and which moves with respect to the other.

Over-flexion. — In the case of man flexion of the knee is limited by

1 Gap: Verhandlungen der physikalischen medicinischen Gesellschaft in
Wiirzburg, N. F., 1884, xviii, p. 171. ,

8 Warren P. Lombard and F. M. Abbott.

the lower leg coming in contact with the thigh or sooner. In the
case of the frog this limitation does not exist, because the lower leg
can be carried dorsally and rotated ventrally sufficiently to enable it
to pass by the dorsal side of the thigh. This is made possible by the
curvature of the lower end of the femur, the form of the joint surfaces,
the method of attachment of the ligaments, and the way the muscles
are arranged on the femur and os cruris near the knee. Flexion
movements continued until the lower leg passes by and begins to
separate from the thigh, according to the definition just given, might
be termed extension; as, however, this position is reached by a con-
tinuation of flexion, we shall speak of that part of the flexion move-
ment which carries the lower leg past the thigh as over-flexion.

Movements of the foot which bend and straighten the limb at the
ankle are described by Gaupp as dorsal flexion and plantar flexion.
It is not very satisfactory to speak of a straightening of the limb at
the knee as extension and a straightening of the limb at the ankle
as flexion, nor is it consistent to use the same term, flexion, for two
antagonistic types of movement. Since the terms extension and flex-
ion are to be used in the case of the hip and knee, it seems best to
employ these same terms for similar movements at the ankle.

In the case of the movements of the toes, however, one has to speak
of plantar and dorsal flexion, since the toes can pass from the straight-
ened, extended position to one which is more bent by moving in the
direction of either the dorsum or the plantar surface of the foot.

To be carried dorsally or ventrally. — If rotation occurs in a joint of
a type to carry the free end of the bone out of the horizontal plane in
the dorsal direction, the bone is said to be carried dorsally, and if the
free end is carried out of the horizontal plane in the ventral direction,
the bone is said to be carried ventrally. If the pelvis is fixed and the
thigh is free to move, the thigh may be carried dorsally by rotating
about an axis which passes horizontally through the head of the femur
and which is perpendicular to its long axis. Similarly the pelvis
might be carried dorsally or ventrally when the femur is fixed, by
rotating about the horizontal transverse axis of the pelvis. We shall
have frequent occasion to refer to this latter type of movement, and
shall speak of the pelvis as being tilted dorsally and ventrally, so as
to become inclined to the horizontal plane.

When the femur is fixed with its frontal plane horizontal, the os
cruris, in flexing, may be carried past the femur and become over-
flexed. In making this movement the os cruris is carried dorsally.

zy

Individual Muscles of the Thigh of the Frog. 9

Under normal conditions this same movement is made when the thigh
itself has been carried dorsally out of the horizontal plane, and conse-
quently the movement of the crus does not take place in the dorsal
direction; nevertheless, inasmuch as the os cruris makes the same
movement with respect to the frontal plane of the femur, the same
term will be used to describe it.

Ventral and dorsal rotation. — These terms imply rotation about the
long axis of the bone. Ventral rotation is employed in the same
sense as inward rotation by man, z. ¢., clockwise, as seen from the dis-
tal end of the bone; and dorsal rotation is used in the same sense as
outward rotation by man, z. e., contra-clockwise.

Names of the surfaces of the limbs. — Quain states that, from the
standpoint of the morphologist, the dorsum of the foot and the ante-
rior surface of the lower leg of man belong to the extensor surface of
the limb, and the sole of the foot and the back of the lower leg to the
flexor surface, so that the muscles on the front of the lower leg should
be called extensors and the muscles on the back of the lower leg
flexors. Nevertheless it is customary to term the raising of the
foot on the lower leg flexion, and its depression, extension.’ In the
very next sentence he proceeds to say: “ The gastrocnemius acts both
as a flexor of the knee and extensor of the ankle joint,” 2. e., although
on the flexor surface it is both flexor and extensor. As a matter of
fact, each of the many two-joint muscles of the hind leg, with one
exception, is a flexor with respect to one, and an extensor with respect
to the other of the joints upon which it acts. Not all of the one-
joint muscles, even, have a constant flexing and extending action.

When the hind leg of the frog is drawn up to the sitting position,
it is folded upon itself, each successive link in the chain of bones, as
far as the toes, moving in the opposite direction to the preceding
one. Physiologically the leg has no flexor or extensor surface; it
does not coil up, but folds together. We shall follow Gaupp in refer-
ring to the surfaces of the leg as the dorsal, ventral, lateral, and median
surfaces, by which are meant those which look in these directions when
the frog lies on its belly with extended legs. This is convenient in
spite of the fact that when the leg has been drawn forward to the
sitting position, the rotation taking place at the hip, knee, and ankle,
causes the directions which these surfaces face to be completely
changed; ¢. g., the lateral surfaces of the thigh and tarsus and the
median surface of the lower leg come to look nearly dorsally. We

1 QuaIN: Anatomy, 1892, ii, p. 273.

IO Warren P. Lombard and F. M. Abbott.

shall avoid the terms flexor and extensor surface of the leg. It is
often convenient, however, to refer to the fervor and extensor sides of
the Aip, knee, and ankle joints, as Hering does, meaning the sides over
which the tendons of the muscles which bend or straighten these
joints pass; and we shall adopt this usage.

The body of a frog usually holds a more or less horizontal, and that
of a man a more or less vertical, position. There are, consequently,
a number of terms which are ordinarily employed to describe the
movements of the body as a whole or of its members, which, unless
defined, are apt to cause confusion.

Forwards and backwards refer to movements towards or away from
the direction in which the animal usually progresses. In the case of
the frog a forward movement is cephalad, and a backward movement
is caudad.

Up and down, away from and towards the earth, as applied to the
movements of the frog, mean in the ventral and dorsal directions
respectively.

Inwards and outwards, median and lateral, signify respectively
towards and away from the sagittal plane of the trunk.

Proximal and distal mean towards and away from the trunk.

A METHOD OF PICTURING THE CHARACTER AND EXTENT OF THE
MOVEMENTS OF THE PELVIS, FEMuR, AND Os CrurRIS.

It is so difficult to form a clear mental picture of the movements of
the hind limb of the frog that, although we have just defined the
terms which we use, a little repetition may perhaps be pardoned.

The positions of the bones entering into a joint can be readily de-
scribed by the use of the following three pairs of terms: flexion
and extension, carried dorsally and carried ventrally, rotated dorsally
and rotated ventrally. The simplest picture of the way in which we
use these terms can be obtained as follows. Think of a book partly
opened and standing on its edge, and of the two bones entering into
a given joint, as lying, dorsal surface upwards, across the middle of the
opposite pages, that is, as horizontal, and the joint in question at the
point where the piges meet. Opening the book can be said to ex-
tend, and closing the book to flex, the joint. If one page is fixed and
the other is moved, the bone on the moving page can be said to flex
with respect to the other when the book is closed, and extend with
respect to the other when the book is opened. The amount of flexion

Individual Muscles of the Thigh of the Frog. II

and extension can be stated in degrees by stating the angle made by
the pages. The book could be said either to be opened 60° or closed
120°. We have found it best to describe the angle in the sense of
the amount that the book is opened, z. e., that the bones are ex-
tended, and for our purposes it has been sufficient to state this angle
as 10, 20°, 30, etc.

Instead’ of the two bones lying directly across the middle of the
pages, one of them might lie diagonally across the page, with its free
end pointing either towards the top or towards the bottom of the
page. If a bone is placed so that its free end points upwards, we
speak of it as having been carried dorsally out of the horizontal plane,
and if its free end points downwards, as having been carried ventrally
out of the horizontal plane. The terms dorsal and ventral are used
because in practice we work with the dorsum of the frog looking up-
wards. The amount that the bone has been carried dorsally could be
measured by the angle which it would make with the horizontal plane,
and could be stated in degrees, and the amount it had been carried
ventrally could be similarly described.

Evidently a bone which lies diagonally across the page, pointing
towards the upper corner, for example, that is carried dorsally, can be
extended or flexed with respect to the other bone by opening or clos-
ing the book, and its position with respect to the other bone can be
readily described by stating the two angles, e. g., carried dorsally 45°
and extended 110°. Both of the bones might have their positions
changed with respect to the horizontal plane; one might point up
the page and the other down the page, for example, and their relative
positions could be stated: by describing the angles which they make
with the horizontal plane, and the amount the book was opened, 7. ¢.,
the amount the joint was extended. In practice, however, it is more
convenient to consider the more cephalad of the two bones as hori-
zontal, and to speak of the other as carried dorsally or ventrally with
respect to it.

To get a mental picture of the meaning of the terms “ rotated dor-
sally and ventrally,” think first of the bones as lying, with the dorsal
side upwards, horizontally across the middle of the two pages, the
book being more or less open, and the distal end of the bone to be
rotated as projecting slightly beyond the edge of the page. Then
think of a needle put into the end of the bone in such a position that
it shall be parallel with the edge of the page, 7. ¢., vertical. We will

12 Warren P. Lombard and F. M. Abbott.

say that in this position the bone is without rotation. Now imagine
the bone to be turned round, while still in the same position on
the page, so that the needle shall make an angle with the page. The
size of the angle determines the amount of the rotation, and if the
rotation, has carried the needle in the direction that the hands of a
clock move, when the observer is looking towards the edge of the
page, the end of the bone, we can speak of the rotation as ventral,
and if the needle has moved in the opposite direction, z. e., contra-
clockwise, we can speak of the rotation as dorsal. In case the bone,
instead of lying across the middle of the page had pointed towards
the upper corner the needle would be no longer vertical, but the
amount of rotation could be estimated the same as before, by sight-
ing along the bone and noting the angle which the needle made with
the plane of the page.

In the above description we have thought of the book as standing
vertically on its lower edge, and the dorsum of the bone as directed
upwards. Of course the positions of the two bones would be the
same with respect to each other, in case the book were lifted up and
given some other position, for the surfaces and long axes of the bones
would hold the same relation to each other, regardless of the positions
which they might hold in space. If the position of the book were
altered, a bone lying across the middle of the page might be no
longer horizontal, but to avoid confusion we should still have to
describe its position as before, and when it was carried dorsally, for
example, still have to speak of it as being carried dorsally out of the
horizontal plane.

In brief, the amount that a bone is extended with respect to an-
other can be expressed by stating the angle made by the projection
of the long axis of the free bone on the horizontal plane with the
sagittal plane of the bone which is regarded fixed.

The amount that a bone has been carried dorsally or ventrally can
be expressed by stating the angle which the projection of the long
axis of the bone on the vertical plane passing through that axis makes
with the horizontal plane.

The amount that a bone is rotated about its long axis can be
expressed by stating the angle which the projection of the sagittal
axis of the bone on the transverse plane makes with the vertical
plane passing through the long axis of the bone.

Individual Muscles of the Thigh of the Frog. 13

METHOD OF WoRK AND APPARATUS.

Our experiments were made chiefly with the leopard frog, but the
dissections shown in Plates I and II were made on the bull frog.
These two forms do not show any marked differences in the structure
of the knee joint and in the method of attachment of the muscles.

A freshly prepared pelvis-limb, pelvis-thigh, thigh-leg, or leg-foot
preparation was used in our experiments. A frog was rapidly decap-
itated; the pelvis and hind limbs were separated from the trunk by
cutting away the abdominal muscles and viscera, and severing the
spinal column just above the last vertebra; finally, the skin was
removed. The pelvis-limb preparation thus obtained was used in
certain experiments, but in most of them all superfluous parts were
cut away, for it was soon found that the weight of distant parts inter-
fered with the action of the more delicate muscles, or through the
action of gravity introduced movements which did not belong to
them. For example, when only the one-joint muscles of the hip were
to be studied, the lower leg was cut away at the knee, and the large
two-joint muscles of the thigh were removed.

Under normal conditions a muscle when acting shortens by con-
tracting, and thus causes its points of attachment to approach. The
ideal method would be to cause contraction of the muscle by electri-
cal excitation of the muscle itself, or, better, its nerve. The study of
the action of a muscle requires a long time, however, for it must be
made to act repeatedly and under a great variety of conditions.
Since the physiological contraction power of the isolated muscle is
lost at the end of a few tetani, it was found necessary in the early
experiments to secure a shortening of the muscle, or rather an
approach of its points of attachment, in another way. This was done,
when the muscles acting on the hip were to be studied, by fastening a
thread to the central end of a cut tendon or muscle, passing this thread
through a fine hole, drilled with the aid of a dental engine in the
bone of the pelvis at the middle of the point of attachment of the
tendon or the muscle fibres, and pulling on this thread by hand or by
a rubber band or spring. In case the attachment of the muscle was
a broad one, it was found necessary to split the muscle longitudinally
for a short distance, to tie a thread to each of the halves, and to pass
these threads through two holes in the pelvis, bored at points corre-
sponding with the outer borders of the attachment of the muscle. By
pulling on these threads, first separately and then together, an

14 Warren P. Lombard and F. M. Abbote.

approximately correct picture of the action of the muscle was
obtained.

In all cases the results thus observed were verified by repeatedly
observing the effect of direct electrical excitation of the muscle, and
in some cases the effect of indirect excitation through its nerve.
When the muscle was electrically excited, the feeblest effective
tetanizing induction current was used, the key being closed only long
enough to reveal the effect of the contraction. The electrodes were
of fine, flexible copper wire, so that the movement should not be
checked. When electrical excitation was employed, great care was
taken to avoid the effects of spread of current. In some cases the
neighboring muscles were insulated by strips of rubber dam.

In most cases all the muscles which might influence the result were
dissected off, unless it was thought that their removal would alter
’ the direction of the strain of the muscle being studied, and those
which were left were cut in halves, so as to do away with the effect
of their contraction. The removal of superfluous masses of muscle
had the advantage that it made the parts to be moved much lighter,
did away with the elastic tension of antagonists, and made it possible
to use much feebler exciting currents than would otherwise be re-
quired. Many of the muscles which cross the hip joint help to keep
the head of the femur in the socket, and care was taken when they
were cut away to see that the head of the bone kept its proper position
when the muscle which was being studied contracted.

The suspension method.— To successfully study the mechanical
action of individual muscles of the leg of the frog, three conditions
must be provided: one of the bones entering the joint must be fixed,
the other must move with the utmost freedom, and the effects of
gravity must be eliminated. These conditions were obtained by the
following method. In studying the action of muscles on the hip, for
example, the pelvis was fastened firmly in a clamp with the iliac
wings either parallel with the earth or tilted dorsally any desired
number of degrees. The thigh, leg, and foot were suspended hori-
zontally from hooks which passed through the fascia over the knee,
ankle, and one of the toes, the hooks being fastened to flexible
threads which passed over freely moving pulleys to weights which
just counterbalanced the weights of the parts supported. To insure
freedom of movement of the parts, the threads to the pulleys were
about 50 cm. long, and the pulleys themselves were supported by
threads about a metre long. The counterbalancing weights were

Individual Muscles of the Thigh of the Frog. 15

little cornucopias of paper filled to the required amount with shot. By
this arrangement the effects of gravity were largely overcome, and a
given muscle was permitted to show not only its power to flex and
extend, but to move the thigh ventrally or dorsally, and to rotate it
about its long axis.

Method of inserting the needle used to study rotation. — In order to
determine the amount of rotation of the femur, for example, about its
long axis, it was found best to insert a needle into the distal end of
the shaft, in such a position that the needle should be vertical when
the femur was horizontal and was without rotation, z. ¢., in the
initial position, and that the needle should not interfere with the
action of the muscles. To do this the bone had to be fastened
securely, and a hole had to be drilled in it with a fine drill. A special
apparatus was needed for this purpose. This apparatus consisted
of three pieces,—a horizontal shelf on which the preparation was
placed, a standard carrying a small clamp to hold the femur, and a
drill press. The device for clamping the femur was as follows.
A standard with a worm gear attachment, which permitted it to be
rotated in a vertical plane, carried a clamp with a worm gear which
allowed a rod which it supported to be rotated in a horizontal plane.
A bone clamp projected downwards at a right angle from this hori-
zontal rod, and in practice with the assistance of a plumb line was
made vertical. This bone clamp consisted of two parts,—a rod
which was clamped to the horizontal rod and which had at its lower
extremity a socket and set screw, and the bone clamp proper on the
end of a rod which fitted the socket, and which could be raised or
lowered or rotated without changing its position with respect to the
vertical plane.

The drill press consisted of a Basilischer Universalstativ, a stand
which had surrounding its vertical rod a metal sleeve which could be
raised or lowered by means of a worm screw, and which had attached
to it at aright angle a two-joint arm, at the end of which the drill-
holder of a dental engine was clamped vertically. This improvised
drill press permitted the drill to be raised or lowered with accuracy,
and to be given any desired vertical position. ;

When the hole was to be bored, the preparation was placed on the
horizontal shelf and given the “ initial position.” In this position the
femur formed a right angle with the long axis of the pelvis, the os
cruris a right angle with the femur, and the tarsus a right angle with
the os cruris, and the bones were levelled up, so that they should be

16 Warren P. Lombard and F. M. Abbott.

horizontal, by means of little pieces of cork placed beneath them.
It was assumed that in this position the femur was without rotation.
The reason for this assumption is that ,this is about the position
which is given to the bones by the elastic tension of the soft parts
when the muscles are at rest, and the frog is floating with the hind
limb in its normal semi-extended position. The standard carrying
the bone clamp was then moved to the preparation; the clamp was
carefully adjusted to the shaft of the bone, and made fast. The
clamped bone was then brought beneath the drill by moving the
standard supporting it, and the hole was bored. Finally, a delicate
needle was inserted into the hole to represent the sagittal axis of the
bone.

THE RELATION WHICH CERTAIN MOVEMENTS OF THE FEMUR BEAR
TO CERTAIN MOVEMENTS OF THE PELVIS, AND THEIR EFFECT ON
THE ACTION OF MUSCLES.

One is likely to misunderstand the effect of various positions of
the femur and pelvis on the movements produced by the contraction
of muscles, unless he first observes the relation which certain move-
ments of the femur bear to certain movements of the pelvis, and the
relative positions of the bones which result.

I. The simplest case is when pelvis (iliac wings) and femur are
horizontal and the pelvis is fixed. Ifthe femur rotates about an axis
which passes vertically through the head of the bone, it is flexed if
the femur approaches the pelvis and extended if the femur leaves the
pelvis; if the femur rotates about an axis which passes horizontally
through the head of the bone and at a right angle to its long axis,
the femur is carried either ventrally or dorsally out of the horizontal
plane; if the femur rotates about the long axis, it rotates ventrally if
it rotates clockwise and dorsally if it rotates contra-clockwise as seen
from the distal end.

A more complicated case is when the femur is rotated about an
axis which passes obliquely through the head of the bone. It then
makes movements of two different types at the same time: it moves
either dorsally or ventrally out of the horizontal plane and at the same
time either flexes or extends.

A still more complicated case is when the femur rotates about an
axis passing obliquely through the head of the bone, and at the same
time rotates about its long axis, in which case all three types of move-
ment occur simultaneously. This last case is a common one, and may

PE Te Te =

Lndividual Muscles of the Thigh of the Frog. 7

be still further complicated by the direction of the oblique axis being
changed while the movement is in progress.

II. Not only is it possible for the femur to move and change its
position with respect to the pelvis, but the pelvis may move and
change its position with respect to the femur.

The simplest case is when pelvis (iliac wings) and femur are hori-
zontal, and the femur is fixed. If the pelvis rotates about an axis
which passes vertically through the head of the femur, or if the pelvis
rotates about an axis which passes horizontally through the head of
the femur and perpendicular to its long axis, the same effects are
produced in the hip joint and on the relative positions of the points of
origin and insertion of the muscles, as when the pelvis is fixed and
the femur is free to move and rotates about these axes.

Rotation of the pelvis about its transverse axis may have the same
effect or a very different effect from rotating the femur about its long
axis, the effect depending on the position which the femur holds to
the pelvis at the time the rotation occurs.

A. If the femur and pelvis are horizontal, and the femur is ex-
tended 90°, so that the sagittal plane of the femur and the transverse
plane of the pelvis are in the same vertical plane, the femur is to be
considered to be without rotation and the pelvis without inclination.
In this case rotation of the femur about its long axis and rotation of
the pelvis about its transverse axis produce like rotation effects in the
hip joint, for these two axes coincide, the one being a prolongation of
the other. Tilting the pelvis dorsally is the same as rotating the
femur ventrally, and tilting the pelvis ventrally is the same as rotating
the femur dorsally. Although this correspondence exists, the effects
on the action of the muscles surrounding the hip joint are quite dif-
ferent. The diameter of the femur is so smal] that rotation of the
bone about its long axis, although it may alter the tension of the rota-
tion muscles, has little effect to alter the direction of the strain of the
muscles. On the other hand, rotation of the pelvis on its transverse
axis, z. ¢., tilting it dorsally or ventrally out of the horizontal plane,
may so change the relation of the points of origin of muscles on the
pelvis to the points of insertion on the femur as to entirely alter the
action of the muscles; for example, a muscle which in one position of
the pelvis is a flexor of the femur, in another position may become an
extensor, or one which in one position carries the femur ventrally, in
another position may carry it dorsally.

B. Tf the femur and pelvis were horizontal, and the femur could be

,

18 Warren P. Lombard and F. M. Abbott.

flexed or extended until the sagittal planes of femur and pelvis were
parallel, the femur would be without rotation and the pelvis without
inclination. In this case rotation of the femur about its long axis and
rotation of the pelvis about its transverse axis would not produce like
rotation effects in the hip joint, for these two axes would not coin-
cide, but the long axis of the femur would be perpendicular to the
transverse axis of the pelvis.

(a) Tilting the pelvis dorsally when the femur was flexed, would
be the same as carrying the flexed femur ventrally; and tilting the
pelvis ventrally, the same as carrying the femur dorsally.

(6) Tilting the pelvis dorsally when the femur was extended,
would be the same as carrying the extended femur dorsally; and tilt-
ing the pelvis ventrally, the same as carrying the femur ventrally.

C. If the femur and pelvis are horizontal, and the femur is flexed
or extended incompletely, so that the sagittal plane of the femur
makes an angle of less or of more than go° with the sagittal plane of
the pelvis, and the sagittal plane of the femur is vertical, the femur
is without rotation and the pelvis without inclination. In this case
rotation of the femur about its long axis and rotation of the pelvis
about its transverse axis do not produce like rotation effects in the
hip joint, because the axes of rotation are different.

(a) Tilting the pelvis dorsally when the femur is flexed, is the
same as carrying the flexed femur ventrally and rotating it ventrally;
and tilting the pelvis ventrally, the same as carrying the flexed femur
dorsally and rotating it dorsally.

(6) Tilting the pelvis dorsally when the femur is extended, is the
same as carrying the extended femur dorsally and rotating it ven-
‘trally; and tilting the pelvis ventrally, the same as carrying the
extended femur ventrally and rotating it dorsally.

It is evident, from what has been said, that rotation of the pelvis
about its transverse axis not only must influence the movement of
the frog through altering the position of the trunk with respect to the
earth, and consequently the direction in which it will be propelled by
the extension of the thighs, but will have a marked effect on the
action of the muscles, the character of this influence depending
largely on the amount of extension of the thighs with respect to the
pelvis at the time that the tilting of the pelvis takes place. Some of
these effects on the action of the muscles will be brought out later in
the description of the action of the individual muscles.

Individual Muscles of the Thigh of the Frog. 19

ACTION OF MUSCLES OF THE THIGH.

All the muscles acting on the hip joint arise from the pelvis.
From the standpoint of their action they can be divided into two
classes: (a) those inserted on the femur, which are one-joint muscles
and act only on the hip joint; (4) those inserted on the os cruris,
which are two-joint muscles and act on both the hip and knee joints.
The long two-joint muscles form the superficial layer surrounding
the thigh on all sides. The short one-joint muscles, most of which
are covered by the two-joint muscles, are divided by Gaupp into
surface, middle, and deep layers. The surface layer lies on the
lateral and dorsal surfaces of the hip; the middle layer surrounds
the whole inner surface of the joint from the anterior almost to the
posterior spine of the pelvis: the deep layer consists of only one
muscle, which lies close to the capsule of the joint, everywhere
concentric with the middle layer, only extending somewhat further
dorsally. The following is the list, as given by Gaupp:

I. Long muscles of the thigh.
a. Muscles of the lateral (forward) surface.
Triceps. — Caput anticum, the cruralis.
Caput medium, the tensor fascia late.
Caput posticum, the glutaeus magnus.
b. Muscles of the medio-ventral surface.
Sartorius. -
Adductor longus.
Adductor magnus.
Gracilis major.
Gracilis minor.
c. Muscles on the medio-dorsal surface.
Ilio-fibularis.
Semimembranosus.
Semitendinosus.
II. Short muscles of the thigh.
a. Surface layer.
lliacus internus.
Iliacus externus.
Iliv-femoralis.
Pyriformis.
b. Middle layer.
Pectineus.

20 Warren P. Lombard and F. M. Abbott.

Obturator externus.
Quadrator femoris.
Gemellus.

c. Deep layer.
Obturator internus.

The positions of the points of origin of these muscles with respect
to the acetabulum are shown in Fig. 1; the points of insertion of
the long muscles at the knee are shown in the dissections given in
Plates I and II.

How MoveEMENTS OF THE Hip JOINT ARE INFLUENCED BY THE
PLACE OF ORIGIN AND OF INSERTION OF THE MUSCLES.

The position of the point of origin of a muscle on the pelvis, with
respect to the acetabulum and the head of the femur, goes far to
determine the action of the muscle, although sometimes the place of
insertion on the femur modifies the movement, this being especially
the case with the rotators. If one thinks of the pelvis, seen from
the lateral side, as projected on the sagittal plane, one can plot a
map of the positions of the points of origin of the muscles with
respect to the centre of the acetabulum. Fig. 1 shows such a map.
The cephalo-caudad axis, CC, has been drawn through the centre of
the acetabulum, parallel to the long axis of the wing of the ilium
(that is, a line drawn through the centre of its proximal end and
through the centre of the point of origin of the gluteus magnus on
the superior process). The dorso-ventral axis, DV, has been drawn
through the centre of the acetabulum at right angles to the cephalo-
caudad axis. Lines have been drawn also from the centre of the
acetabulum, through the centre of the place of origin of each of the
muscles; and these indicate in a general way the direction in which
the femur could be supposed to move, cephalad, caudad, dorsad, and
ventrad, when any one of the muscles in question shortened by
contracting. <A circle having its centre in the centre of the acetabu-
lum has been divided in degrees, and this permits one to state in
degrees how far dorsal or ventral to the cephalo-caudad axis and
how far cephalad or caudad to the dorso-ventral axis the point of
origin of any given muscle lies. For example, the origin of the
glutaeus magnus lies 72° dorsal and 18° cephalad; the ventral head
of the semitendinosus lies 30° ventral and 60° caudad.

The principal action of the muscles which have their origin

FiGuRE 1.— The points of origin of the muscles of the thigh with
respect to the acetabulum.

!
'
:
eo

ii
.

f
:

7
.

:
;

—_—=

Individual Muscles of the Thigh of the Frog. 21

cephalad to the acetabulum is to flex the femur, and of those the
origin of which is caudad to the acetabulum is to extend the femur,
The chief action of muscles the points of origin of which are ventral
and dorsal, is respectively to carry the femur ventrally and dorsally.
Such muscles may also act to tilt the pelvis dorsally or ventrally.
Certain muscles may rotate the femur dorsally or ventrally, these
effects being largely determined by the method of attachment on the
femur.

AcTION OF MUSCLES TO FLEX AND EXTEND THE FEMUR.

Pelvis and femur horizontal.— Muscles the points of origin of
which lie on the pelvis 30° or more cephalad to the dorso-ventral
axis of the map, act to flex the femur. Such muscles are the
cruralis, adductor longus, sartorius, pectineus, iliacus externus,
tensor fasciz latz, iliacus internus, and gluteus magnus.

Muscles the points of origin of which lie on the pelvis 30° or
more caudad to the dorso-ventral axis, except the ventral head of the
adductor magnus, extend the femur. Such muscles are the semiten-
dinosus (both ventral and dorsal heads), quadratus femoris, adductor
magnus (dorsal head), gemellus, semimembranosus, and gracilis
minor.

A muscle the point of origin of which lies directly dorsal to the
acetabulum acts to carry the femur dorsally when the long axis of
the femur forms a right angle with the long axis of the pelvis; but if
the femur be flexed, the line of strain of the contracting muscle is
carried to the flexor side of the axis for flexion and extension, and
the muscle acquires the power to flex as well as to carry dorsally ;
and if the femur is extended, the muscle for a similar reason ac-
quires the power to extend the femur as welk as to carry dorsally.
Gad! has already attracted attention to this reversal of the action
of the vastus internus (gluteus magnus) and the ‘adductors”
(muscles the points of origin of which lie ventral to the acetabulum).
In speaking of the ilio-psoas Gad says: “It distinguishes itself in
this respect from other muscles which, like the vastus internus, flex
the already flexed thigh further, but from a given extension on act
to extend. I think that the adductors also show such a reversal of
the method of action. The explanation is that these muscles lie

? Gap: Verhandlungen der physikalischen medicinischen Gesellschaft in
Wiirzburg, N. F., 1884, xviii, p. 172.

EXPLANATION OF PLATE ONE.

PLATE I.— Dissection of right knee of bull frog. A. Z., adductor longus; Z. B., ex-
tensor brevis; Z. 4’., tendon of extensor brevis; Z. Z. Z., external lateral ligament ;
G., gastrocnemius (plantaris longus); G’, extra tendon of gastrocnemius; G. JZ,
gracilis magnus; G. A/z., gracilis minor; 7. F%., ilio-fibularis (biceps); 7 Z. Z., inter-
nal lateral ligament; ?., peroneus; 7’., tendon of peroneus; S., sartorius ; S7z., semi-
membranosus ; S¢., semitendinosus; 7., triceps; 7. A., tibialis anticus longus; 7: 4’.,
tendon of tibialis anticus longus.

FicureE 1.— Map of points of origin of muscles of right thigh of leopard frog, on the
lateral side of pelvis. C. C., cephalo-caudad axis; D. V., dorso-ventral axis; A. L.,
adductor longus; Cv., cruralis; S.,sartorius; P¢, pectineus; O. 7. (V), obturator in-
ternus, ventral head; O. Z., obturator externus; 4. JZ. (V), adductor magnus, ventral
head ; S¥. (V.), semitendinosus, ventral head; Q. #, quadrator femoris; 4. AZ. (D.),
adductor magnus, dorsal head; G. J/., gracilis magnus ; S¢. (D.), semitendinosus, dorsal
head ; O. 7. (D.), obturator internus, dorsal head; G., gemellus ; Svz., semimembranosus ;
G. Min., gracilis minor; 7. F%., ilio-fibularis ; 7. Fy., ilio-femoralis; G/. 47, glutzeus
magnus; /. £., iliacus externus; 7. #& Z., tensor fascie late; /. /, iliacus internus.

~~ f

VOLUME XX. OCTOBER 1, 1907. NO. 1.

ERICAN JOURNAL OF PHYSIOLOGY

j

——

oie

Es

LOMBARD AND ABBOTT.

RIGHT LEG OF FROG.

VENTRAL SIDE.

Lndiwidual Muscles of the Thigh of the Frog. 22

outside of the plane of rotation of the os femoris, and are so inserted
above that the result of their pull comes to lie now on the flexor,
now on the extensor, side of the sagittally directed axis of rotation.”
Such muscles are the following:

Adductor magnus (ventral head) —
flexes femur if extended less than 45°.
extends femur if extended more than 45°.
Obturator externus —
flexes femur if extended less than 70°.
extends femur if extended more than 70°.
Tlio-femoralis —
flexes femur if extended less than 80°.
° extends femur if extended more than 80°.
Tlio-fibularis —
flexes femur if extended less than go°.
extends femur if extended more than 90°.

The obturator internus also shows a reversal of its action, but be-
haves differently from the rest of these muscles, because the relation
of the tendon of insertion to the head of the femur so changes, as the
thigh is flexed and extended, that it flexes when the femur is extended
more than 120°, and extends when it is extended less than 120°.

Pelvis tilted dorsally 40°, the femur being horizontal. — Not only may
the action of certain muscles be reversed, from the standpoint of flex-
ion and extension during the progress of extension of the femur, when
the pelvis and femur are horizontal, but a further change in their action
may take place when, the femur remaining horizontal, the pelvis is
tilted dorsally, as occurs when the frog takes the sitting position.
The dead point for the obturator externus is changed from 70° to 110°
extension, and for the ventral head of the adductor magnus, from 45°
to 90° extension. The ilio-fibularis and ilio-femoralis, which acted to
flex when the pelvis was horizontal and the femur extended not more
than 80°, become pure extensors when the pelvis is tilted dorsally.
Finally, the gluteus magnus, which is a pure flexor when the pelvis
is horizontal, becomes a pure extensor when the pelvis is tilted dorsally
40°, as in the sitting position.

The explanation offered by Gad —“ that the muscles lie outside of
the plane of rotation of the os femoris, and are so inserted above that
the result of their pull comes to lie now on the flexor, and now on the
extensor side of the sagittally directed axis of rotation’ — is correct in

24 Warren P. Lombard and F. M. Abbote.

part, but examination of the map shows that the character of the re-
versal of the action of certain muscles is not wholly explained by the
position of the points of origin with respect to the axis of rotation.
The method of insertion on the femur and os cruris also plays its part
in causing the line of strain to be transferred from the one to the
other side of the axis of rotation. The fact that the ilio-fibularis,
ilio-femoralis, and obturator externus are inserted to the median
side of the long axis of the femur, and the glutaus magnus some-
what to the dorsal side, helps to explain the divergence in the action
of these muscles.

ACTION OF MUSCLES TO CARRY THE FEMUR DORSALLY AND
VENTRALLY.

Muscles the points of origin of which lie on the pelvis 20° or more
dorsal to the cephalo-caudad axis of the map, act to carry the femur
dorsally. Such muscles are semitendinosus (dorsal head), gemellus,
semimembranosus, ilio-fibularis, ilio-femoralis, pyriformis, gluteeus
magnus, iliacus internus, tensor fasciz lata.

Muscles the points of origin of which lie on the pelvis 20° or more
ventral to the cephalo-caudad axis of the map, act to carry the femur
ventrally. Such muscles are: adductor longus, sartorius, pectineus,
obturator externus, adductor magnus (ventral head), semitendinosus
(ventral head).

Muscles the points of origin of which lie on the pelvis less than
20° dorsal or ventral to the cephalo-caudad axis, show a reversal of
their action. Ifthe femur has been carried out of the horizontal plane
ventrally, they tend to carry it dorsally; or if it has been carried out
of the horizontal plane dorsally, they tend to carry it ventrally. Such
muscles are: quadratus femoris, adductor magnus (dorsal head),
gracilis major, and iliacus internus. The cruralis, being attached over
the capsule, shows little effect to carry the femur either dorsally or
ventrally, and the obturator internus, because of its extensive origin
is omitted from the list. The adductor magnus (dorsal head), the ori-
gin of which lies caudad and to both sides of the cephalo-caudad axis,
shows a peculiar action; it carries the femur dorsally in case it has
been flexed and carried dorsally, or extended and carried ventrally,
and it carries it ventrally if it has been flexed and carried ventrally,

or extended and carried dorsally. The quadratus femoris would prob-
ably do the same.

Individual Muscles of the Thigh of the Frog. 25

Just as in the case of muscles which are flexors and extensors, so
with those that carry the femur dorsaily and ventrally, the action
may be markedly changed by tilting the pelvis dorsally.

ACTION OF MUSCLES TO ROTATE THE FEMUR.

Of course the method of insertion of the muscle on the femur is of
importance in determining its action. A muscle may run parallel
with the femur from origin to insertion, in which case the femur will
move towards the side on which the muscle lies. When the femur
rotates about its long axis, the radius of movement is so short that
the displacement is small, and the power to flex, extend, carry ven-
trally, and carry dorsally are not altered much. On the other hand,
if a muscle in passing from its point of origin on the pelvis to its inser-
tion on the femur winds round the shaft of the femur, it acquires the
power to rotate the femur about its long axis. Muscles which rotate
the femur ventrally, z. ¢., clockwise to one looking at the distal end of
the bone, wind round the bone from the point of origin contra-clock-
wise; and those that rotate the femur dorsally, z. ¢., contra-clockwise,
wind round it clockwise. The most important ventral rotator is the
iliacus externus, and the most important dorsal rotator is the obtura-
tor internus. Other muscles which rotate ventrally are: sartorius,
ilio-fibularis, ilio-femoralis, gracilis magnus, semimembranosus (when
the femur is rotated dorsally), and iliacus internus ; and one which ro-
tates dorsally is the dorsal head of the adductor magnus.

The power of a muscle to rotate the femur may be increased or de-
creased by tilting the pelvis, because a muscle which winds from its
point of* origin round the shaft of the bone will be made to do so
either more or less by changing the position of the point of origin with
respect to the shaft of the bone.

AcTION oF MuscLEs To TILT THE PELVIS DORSALLY OR VENTRALLY.

Muscles having their points of origin on the pelvis dorsal to the
cephalo-caudad axis tilt the pelvis dorsally if the femur be extended,
and ventrally if it be flexed, and the effect is the stronger the more
dorsal the origin of the muscle. Muscles which have their points of
origin on the pelvis ventral to the cephalo-caudad axis, tilt the pelvis
dorsally if the femur be flexed and ventrally if it be extended, and the
effect is the stronger the more ventral the point of origin.

26 Warren P. Lombard and F. M. A bbote.

When the femur is flexed, the muscles which tilt the pelvis dorsally
are: sartorius, pectineus, obturator internus, obturator externus,
adductor magnus (ventral head), semitendinosus (ventral head).
Those that tilt the pelvis ventrally are: semitendinosus (dorsal
head), gemellus, semimembranosus, ilio-fibularis, ilio-femoralis, glu-
tzus magnus, iliacus externus, tensor fasciz late, and iliacus inter-
nus. These two lists would have to be reversed when the femur is
extended. The quadratus femoris, adductor magnus (dorsal head),
gracilis major, cruralis, and adductor longus have their points of ori--
gin so.near the cephalo-caudad axis that they have but little action to
tilt the pelvis.

EFFECT OF STRUCTURE OF KNEE JOINT ON THE ACTION OF MUSCLES.

The effect of the muscles on the knee joint is largely determined
by the character of the joint surfaces in apposition at the time the
contraction occurs, and on the action of the ligaments to limit and
guide the movements. The amount that the knee is extended at the
time that the muscles contract has therefore a great influence on the
resulting movement. The peculiar character of the knee joint of the
frog deserves a careful description, but only a few points concerning
it can be brought out here (see Plate II, Fig. 1). It is only the more
median portions of the condyles which form the rotation surfaces, the
lateral part of the head of the femur serving for the attachment of the
tendons of the peroneus, tibialis anticus longus, and extensor brevis
muscles, and the lateral part of the head of the os cruris carrying the
grooves in which the tendons of these muscles play. The middle por-
tion of the joint is occupied by the strong ligaments which bind the
bones together. In addition to the internal ligaments, there is a
strong ligament on the dorsal and another on the ventral side of the
joint which binds together the condyles. These ligaments (“the
lateral ligaments”) slide over the smooth surfaces of the condyles
as the joint is flexed and extended, and the condylar surfaces are
so formed that the ligaments are always kept tense, and keep the
rotation surfaces in close apposition.

The structure of the joint is such that it can be considered to have
two distinct rotation surfaces, the one formed by the extremities, the
other by the median surfaces of the heads of the bones. The change
from the one to the other of these surfaces comes when the joint has
been flexed, to about 110° extension. Up to this point the move-

Individual Muscles of the Thigh of the Frog. 7)

ment is one of simple flexion, and beyond this point, in the direc-
tion of flexion, a carrying of the os cruris dorsally with respect to
the femur and a ventral rotation of the os cruris become possible in
addition to the flexion. The proximal portion of the os cruris may
be described as consisting of a dorsal (lateral, Gaupp) and a ven-
tral (median, Gaupp) condyle, and between these a central portion
which serves for the attachment of the internal ligaments of the
joint. Looked at from the dorsal side, the dorsal condyle has the
appearance of a disk, with rounded border, growing out of the shaft of
the bone. By extended knee the part of the disk most proximal to
the shaft of the bone is opposed to the rotation surface of the corre-
sponding condyle of the femur, and when the knee is overflexed, the
part of the disk nearest the median surface of the.shaft of the os cruris
is in contact with the median surface of the dorsal condyle of the fe-
mur. Indeed, the overflexion is checked by the ligaments only when
the median surface of the shaft of the os cruris almost touches the
median surface of the shaft of the femur. In the process of flexion
the border of the disk of the dorsal condyle slides on the surface
of the femur in such a way as to give the effect of rolling upon it, but
the motion is really more of a sliding than of a rolling character. As
the crus passes from flexion to overflexion, the os cruris rotates ven-
trally and moves dorsally, a twisting movement being seen at the
place of contact of the rotation surfaces of the dorsal condyles.

The lateral part of the ventral condyle has the same disklike form
as the dorsal condyle, and the rounded border is continued as a gentle
curve on to the most proximal part of the disk. Beyond this, toward
the median side, the curvature becomes much more rapid, and on the
inner side of this part there is a distinctly flattened surface sloping
inward to form the ventral side of the intercondylar fossa. When the
joint is extended, the proximal surface of the ventral condyle is in
contact with the most distal part of the ventral portion of the head of
the femur, and slides and rolls on this as the joint flexes to about 110°,
but beyond this point new surfaces come in contact. The middle
portion of the head of the femur projects somewhat beyond the rest,
fitting into the intercondylar fossa of the os cruris, and the flattened
inner surface of the ventral condyle of the os cruris slides on this
portion of the head of the femur, which is here slightly flattened to
correspond. It is the flattening of these two surfaces and their direc-
tion which makes it possible for the os cruris to be carried dorsally and
rotated ventrally as it passes into overflexion. The movement of these

EXPLANATION OF PLATE TWO.

PLATE II, FicurE 1. — Dissection of ventral side of right knee of bull frog. #, femur;
C., os cruris ; S7z., semimembranosus muscle. The knee is almost completely flexed,
and the rotation surfaces of the condyles of the os cruris are in contact with the
median surface of the head of the femur. If the os cruris had been flexed 20°
further, the median surface of the head of the femur would have come to bear in the
intercondylar fossa, on the median side of the head of the os cruris. The dissection
shows the relation of the tendon of the semimembranosus to the joint, and explains
the action of the muscle to over-flex the os cruris on the femur.

FicurRE 2.— Dissection of ventral side of right knee of bull frog, showing the way the
tendons of the muscles of the thigh enter the joint.

TOBER 1, 1907. NO. 1.
AMERICAN JOURNAL OF PHYSIOLOGY VOLUME XX. OCTOBE

Ventral side of right leg.

SII,

STUVULIWIOL48DB

LOMBARD AND ABBOTT.
VENTRAL SIDE. RIGHT LEG OF FROG.

Lndiwidual Muscles of the Thigh of the Frog. 29

two surfaces on each other is distinctly a sliding motion. The os
cruris when completely flexed can be carried dorsally about 45°. The
ventral rotation does not commence until overflexion has commenced,
and is the result of the form of the surfaces on the median sides of
the two bones, and the action of the lateral ligaments of the joint.
The dorsal condyle of the crus acts as if pivoted on the dorsal condyle
of the femur, and as the os cruris is carried dorsally and is rotated
ventrally, its ventral condyle describes a circle about its dorsal condyle.

ACTION OF INDIVIDUAL MUSCLES.

In the following description of the action of the muscles of the
thighs they have been arranged, for convenience of reference, alpha-
betically. The position of the points of origin of the muscles of the
thigh on the pelvis is shown in Fig. 1, and the relation of the tendons.
of the long muscles to the knee joint is shown in Plate I and Plate II.

Adductor longus. — This muscle resembles in form the sartorius,
by which all of it but the lateral border is covered. It arises from
the outer surface of the illum just caudad to the anterior spine of the
pelvis, and beneath the line of insertion of the sartorius. The line
of origin lies cephalad and slightly ventral to the acetabulum and the
head of the femur, and from:this point the muscle runs along the
ventral side of the thigh, crossing from the lateral to the median
side, to be inserted by a tendon common to it and the ventral head
of the adductor magnus on the epicondylus medialis. This is a
feeble muscle which acts to flex the thigh and to a less extent to
carry it ventrally.

Adductor magnus, — This is a powerful muscle lying on the median
and ventral sides of the thigh. It is seen in a triangular space
between the sartorius and gracilis muscles in the proximal part of the
thigh, and is covered by them in the distal portion. It arises from
the ischium and the adjacent part of the remanens cartilage by three
heads, and is inserted on the distal end of the femur, where it almost
completely surrounds the bone. The muscle as a whole acts to
extend the thigh on the pelvis, to carry the thigh ventrally, and to
rotate it dorsally.

The caput ventrale arises from the border of the remanens cartilage
and the neighboring part of the ischium close to the symphysis
pubis, the line of origin being more ventral than caudad to the
acetabulum and the head of the femur. The muscle fibres extend to

30 Warren P. Lombard and F. M. Abbote.

the distal end of the femur, where they are inserted just proximal to
the epiphysis, in part on the median and to a lesser extent on the
ventral surface of the shaft of the bone, and in part along with the
fibres of the adductor longus on a tendon which is attached to
the epicondilus medialis. In the distal part of its course the fibres
of this head are closely related to those of the accessory and dorsal
heads, and where they are inserted on the median surface of the bone
they form a common muscle. The ventral head lies for the most
part along the medio-ventral surface of the femur, but the fibres shift
their place with respect to the shaft of the femur as it is carried from
the flexed to the extended position, lying when it is flexed somewhat
more to the latero-ventral, and when it is extended to the medio-
ventral side of the bone. When the pelvis is tilted dorsally, the
origin of this head is carried cephalad, and the proximal part of the
muscle lies still more to the latero-ventral side of the flexed femur.

Electrical excitation shows that the ventral head of the muscle,
when pelvis and femur are in the horizontal plane, has above all
the power to carry the thigh ventrally. In addition to this it can
extend the thigh if it is already extended 45°; but if the amount
of extension is less tham 45°, this head acts to flex the thigh. When
the pelvis is tilted dorsally, the thigh being in the horizontal plane,
the origin -of the ventral head is carried somewhat more cephalad
with respect to the head,of the femur, and this head acts as a flexor
when the thigh is extended 90° or less; when the extension is more
than this, the thigh is extended. The power of this head of the
muscle to carry the thigh ventrally is so great, in all positions of the
limb, that its extensor and especially its flexor action are well seen
only when the ventral movement is prevented, for example, in the
case of a suspended preparation, by holding the thread supporting
the knee, or by holding a smooth needle against the limb. In case
the flexed femur is carried ventrally before the ventral head of the
muscle is excited, the femur is seen to rotate dorsally, and to describe
a segment of a cone as it passes from the flexed and ventral to the
extended horizontal position. These effects are not due directly to
the action of the muscle, but to the passive action of joint surfaces
and the ligamentum ventrale.

The caput dorsale arises in sequence to the ventral head from the
border of the ischium, and lies directly caudad to the acetabulum and
the head of the femur. The fibres which have the more ventral
origin are inserted near the distal end on the median side of the

Lndividual Muscles of the Thigh of the Frog. 21

re)

shaft of the bone, and those having the more dorsal origin wind
round the end of the shaft and are inserted on the dorsal, dorso-
lateral, and lateral surfaces, the insertion on the lateral surface
including the lower third of the bone. Thus the ventral part of the
head of the muscle lies along the median, and the dorsal part along
the median and dorsal surfaces of the shaft of the femur, some of the
distal portion winding over the dorso-lateral surface of the shaft well
on to the lateral side.

When both pelvis and femur are in the horizontal plane, this head
of the muscle acts as a powerful extensor. In addition to this it
rotates the thigh dorsally, this action being most marked when the
thigh is flexed and the pelvis tilted dorsally. This latter action is
important, because by it the proximal end of the os cruris is kept
beneath the distal end of the femur, and the trunk is thrown forwards
and upwards by the extending thighs, instead of the legs being
driven out sideways. This head of the muscle shows little tendency
to carry the thigh either dorsally or ventrally when the pelvis and
thigh are horizontal, but if in the flexed position the femur is carried
slightly dorsally, this head will carry it dorsally, and if the femur be
carried slightly ventrally, it will carry it ventrally; if in the extended
position the femur has been carried dorsally, this head will carry it
ventrally as far as the horizontal plane, and if the femur has been
carried ventrally, this head will carry it dorsally as far as the
horizontal plane.

A word with respect to the effect of tilting the pelvis dorsally. As
has been explained elsewhere, when the thigh is horizontal, tilting
the pelvis dorsally when the thigh is in the flexed position has the
same effect on the action of the muscles as carrying the femur
ventrally; therefore, when the frog is at rest in the sitting position,
tilting the pelvis dorsally enables the dorsal head of the adductor
magnus muscle, when it contracts, to help the ventral head to carry
the thigh ventrally, —in other words, to raise the body from the
ground at the same time that it is thrown forward by the extending
thigh. When the thigh is extended, tilting the pelvis dorsally has
the same effect on the action of the muscle as carrying the femur
dorsally, and this head will carry the femur ventrally. Thus through-
out the jumping movement this muscle would help not only to extend
the thigh, but to propel the body upwards. If the femur is at 90°
extension, tilting the pelvis dorsally acts in the same way as rotating
the femur ventrally, and increases the power of the muscle to rotate

32 Warren P. Lombard and F. M. Abbott.

the femur dorsally. Thus, when the thigh is half extended, the
muscle helps to keep the distal end of the femur dorsal to the
proximal end of the crus, so that the extremity of the femur shall be
properly supported, and at the same time it favors a backward rather
than a lateral movement of the extending limb.

The caput accessorium arises from the slender tendon of the ventral
head of the semitendinosus. It is composed of a few bundles of
fibres lying median to the shaft of the femur between the two other
heads of the adductor magnus, with which it is closely connected.
It acts to extend the femur.

Cruralis. — The cruralis (vastus internus of Ecker), which forms
the largest and strongest head of the triceps group, lies to the lateral
and ventro-lateral sides of the femur, the ventral border of the muscle
being covered by the adductor longus and sartorius muscles.

The cruralis arises by two heads. The outer or more lateral head
arises by a very short tendon which is attached to the capsular liga-
ment and partly to the cephalad border of the cartilage which helps
to form the acetabulum. Some of the fibres of the inner head take
origin from the capsular ligament by a short tendon, while the re-
mainder spring from a Y-shaped ligament which arises from the cap-
sule and is prolonged as a slender cord which runs along the median
side of the muscle and is inserted into the common tendon of the
triceps. The muscle runs parallel with the femur beneath the tensor
fascize lates, and is inserted into the common tendon of the triceps
group.

The cruralis is a two-joint muscle which acts on the knee and the
hip joints. It extends the os cruris on the femur, and this is perhaps
its most important action. When the pelvis and the thigh are in
the horizontal plane, and the leg in the resting semi-extended position,
it flexes the femur until it forms an angle of about 90° with the long
axis of the pelvis. If the crus be flexed on the femur and be kept
from extending, the muscle will cause the femur to be completely
flexed. This action is not, however, strong, because the origin of
the muscle on the capsular ligament brings the strain close to the
axis of rotation. On the other hand, the muscle tends to bind the
head of the bone in the acetabulum from the cephalad side of
the joint, just as the obturator internus and iliacus externus do
from the caudad side. This action of the muscle is very important.

With the pelvis horizontal or tilted dorsally 40° and the thigh
in the horizontal plane, or carried ventrally or dorsally out of the

Individual Muscles of the Thigh of the Frog. 33

horizontal plane, the action of the muscle is the same as above
described.

Gemellus.— The gemellus lies somewhat dorsally and somewhat
caudad to the hip joint. It is an extensor of the thigh in all
positions of the limb, its power lessening rapidly as the thigh
nears complete extension because of the lessening tension of the
muscle. If the extensor action is prevented, the muscle will be
seen to carry the thigh dorsally. In the sitting position the extensor
action of the muscle is increased, at the expense of the power to
carry dorsally. It tilts the pelvis ventrally if the femur is flexed, and
tilts it dorsally if the femur is extended.

The gluteus magnus (vastus externus of Ecker) is a strong two-
joint muscle lying on the dorsal and dorso-lateral surfaces of the
thigh. It arises from the outer surface of the posterior process of
the wing of the ilium, the point of origin lying cephalad, dorsal,
and ‘lateral to the acetabulum. The fibres take a course in general
parallel to the femur, to be inserted into the common tendon of the
triceps group. The muscle vigorously extends the os cruris on the
femur. It acts strongly to carry the femur dorsally, in all positions
of the femur and pelvis. Gad! says that the vastus externus flexes the
already flexed thigh further, but extends it from a certain degree of
extension. He goes on to say that he thinks that the adductors may
have such a reversal of their action. ‘“ This lies in the fact that
these muscles lie outside of the plane of rotation of the os femoris,
and are so inserted above that the resultant of their pull comes to lie
now on the flexor, now on the extensor side of the sagittally directed
axis of rotation.” We cannot altogether agree with this, as far as
the gluteus magnus is concerned. When the pelvis and thigh are
horizontal, the muscle causes the femur to be flexed in all positions
of the leg. If the pelvis is tilted dorsally more than 20° and the
thigh is in the horizontal plane, the point of origin of the muscle
having been carried caudad with respect to the axis of rotation for
flexion and extension of the hip joint, the flexor action of the muscle
on the hip is changed to an extensor action, and it will extend the
femur in all positions of the thigh, from complete flexion to complete
extension. If the femur is fixed, the glutzus magnus will tilt the
pelvis dorsally if the femur is extended, and carry the pelvis ventrally
if the femur is flexed.

1 Gab: Verhandlungen der physikalischen medicinischen Gesellschaft, in
Wiirzburg, N. F., 1884, xviii.

34 Warren P. Lombard and F. M. Abdbbote.

In the sitting position, with the pelvis tilted dorsally and the thigh
horizontal, the muscle acts not only as a powerful extensor of the
knee, but as an extensor of the hip, and so aids in the leap. Aside
from the extensor action on the knee, the most important action of
the muscle is to carry the thigh dorsally. As the thigh extends
and the trunk is raised from the ground, this muscle, by carrying
the thigh dorsally, brings it more in line with the pelvis, and so
favors the forward movement of the trunk. The carrying of the
thigh dorsally also helps to raise the knee from the ground. As
the femur comes into the plane of the pelvis, the muscle changes
its action from being an extensor to being a flexor of the hip, and
probably at the end of the leap helps to flex the thigh. Inasmuch as
the muscle is a strong extensor of the knee, it is doubtful whether it
can do more than help initiate the flexion movement.

Gracilis major.— This is a strong two-joint muscle, the proximal
part of which lies on the median, and the distal part towards the médio-
ventral side of the thigh. The ventral border of the muscle partly
covers the adductor magnus, and its distal extremity is partly covered
by the sartorius. Dorsally it is in contact with the semimembranosus,
It arises caudad to the acetabulum from the most caudad part of the
pelvic disk, in the region of the ischium, by a thin, ribbon-like tendon.
It covers the ventral head of the semitendinosus, the tendon of which
can be seen to emerge from beneath its dorsal border at the knee.
The tendon of insertion of the gracilis major passes under the ex-
pansion of the sartorius tendon, over the ligamentum mediale of the
knee joint, to be inserted on the ventral (median, Gaupp) surface of
the median condyle of the os cruris. The muscle, as it descends the
thigh, therefore crosses from the median to the medio-ventral side of
the femur, and so acquires the power to carry the femur ventrally
and rotate it ventrally. Two tendinous slips are given off from the
main tendon, one of which passes across the median surface of the
proximal end of the shaft of the os cruris, where it is inserted well
towards the dorsal side, just distal to the point from which the lateral
condyle springs; the other slip connects with the tendon of the semi-
membranosus (see Fig. 4).

The most important action of this muscle on the hip joint is to ex-
tend the femur on the pelvis; as has been said, it also has a slight
action to carry it ventrally, and it rotates it ventrally. These effects
are produced when both pelvis and femur are in the horizontal plane,
and when either of them has been carried either dorsally or ventrally

]

Individual Muscles of the Thigh of the Frog. 35

out of this plane. In the sitting position the femur is rotated dorsally,
which increases the power of the gracilis magnus to rotate it ventrally at
the same time that it extends it. In this position it would also tend
to tilt the pelvis dorsally, which would help to raise the trunk from the
ground during the leap. When the pelvis is tilted dorsally, the origin
of the muscle is carried more ventrally with respect to the head of the
femur, which would increase the power of the muscle to carry the femur
ventrally.

By its distal end the muscle either flexes or extends the os cruris
on the femur, according to the position of the knee at the time that
the muscle contracts. (a) If the degree of the extension of the knee
is less than 110° when the muscle contracts, it will cause flexion, and
this flexion movement, if continued, will result in an over-flexion,
(4) If the degree of extension of the knee at the time that the muscle
contracts is between 120° and 140°, the muscle causes extension,
(c) If the degree of extension is more than 140°, the muscle will cause
flexion. This latter effect is, however, only produced when the muscle
is put under strong tension. There is a dead point between the po-
sitions a and 4, and another between 4 and c. The dead point between
a and 6 is about at the position the knee takes when in the resting
semi-extended position, z. ¢., extended 110°-120°.

When the action of the muscle is being tested by electrical excitation,
a marked ventral rotation of the thigh occurs; and if this be prevented
by holding the femur with forceps, it will be found that as the crus is
fiexed it is also rotated ventrally and carried dorsally, the rotation in-
creasing as the flexion progresses; and in extreme over-flexion the
crus may form an angle of 45° with the femur on its lateral side.
The carrying dorsally begins to show itself when the crus is flexed
about 110°. This action of the crus is partly due to the muscle
being inserted on the ventral side of the median condyle, but more to
the character of the joint surfaces and ligaments (see p. 27). These
movements produced by the gracilis major resemble those caused by
the semimembranosus; but its extension action on the knee is less,
and its rotation effect on the knee more, than that of the semi-
membranosus. The flexion action is seen regardless of the extension
at the hip, except in so far as flexion of the knee is interfered with by the
muscles on the lateral surface of the thigh, which are put under tension
by the extension of the hip and flexion of the knee. If the extension of
the hip is not more than 150°, the flexion of the knee caused by the
contraction of the gracilis magnus may be sufficient to carry the crus

36 Warren P. Lombard and F. M. Abbote.

past the femur into the position of extreme overflexion. This muscle,
by extending the thigh and overflexing the lower leg, will bring the
foot into the position which will enable it upon the contraction of the
gastrocnemius to wipe a piece of paper off the back. Of course these
two muscles would not be the only ones engaged in such a movement.

For the same reason that complete flexion of the crus cannot take
place when the thigh is extended more than 150°, over-flexion of the
crus prevents complete extension of the femur. During life, contrac-
tion of the muscles on the lateral side of the thigh, by shortening them,
would still further limit the extension of the thigh with respect to the
pelvis when the knee was in over-flexion, and would likewise limit
the flexion of the crus on the thigh when it was extended.

The power of the gracilis magnus to flex the knee is important, be-
cause it opposes the extension of the knee at the beginning of the
leap and consequently prevents the lower leg and foot from being
driven out laterally with respect to the trunk during the early part of
the stroke. The gracilis magnus is an extensor of the hip as well as
a flexor of the knee. Moreover, it has greater leverage upon the hip
than the knee, and helps to extend the hip at the expense of its flexor
action on the knee. As the hip becomes more and more extended,
the tendon action of the two-joint muscles which flex the thigh comes
into play, and this fact, together with the direct extensor action of
these muscles on the knee, overcomes the flexor effect of the gracilis
magnus muscle on the knee. When the knee has been extended 110°—
120°, the dead point between the sliding and rotation surfaces of the
knee is passed, and the action of the gracilis magnus is suddenly
changed from flexion to extension, and it assists in thrusting the leg
backwards.

Gracilis minor.— The gracilis minor lies on the median surface of the
thigh. It forms a thin narrow ribbon, and lies directly beneath the
skin. The fascia covering the cephalad two thirds of the muscle is
connected with the skin at frequent intervals by connective tissue
bridges.

There is a very thin sheet of fibrous tissue attached in the median
line to the cartilagio-marginalis, which covers the border of the pelvic
disk. This delicate sheet of tissue extends ventrally as far as the ori-
gin of the rectus abdominis and dorsally to the insertion of the more
caudad fibres of the sphincter ani muscle. The gracilis minor arises
with its fellow of the opposite side from the middle of this sheet of
tissue. The muscle is inserted at the knee to the tendon of the
gracilis major.

Individual Muscles of the Thigh of the Frog. a7

During contraction, the muscle, by virtue of the connective tissue
bridges which bind the cephalad two thirds to the skin, causes the
skin to be drawn into minute folds. These are different from the
large folds seen when the thigh is made to extend after the connec-
tions between the skin and the muscle have been severed. The con-
traction of the muscle puts the skin on the median side of the thigh
under tension and holds it close to the thigh. The muscle tends to
cause the same movements of the thigh and lower leg as the gracilis
major, but is very much weaker.

Iliacus externus.— This muscle arises from the lateral surface of
the wing of the ilium, and its fibres pass caudad and somewhat ven-
trally to be inserted on a little projection (trochanter) near the me-
dian and distal border of the dorsal flattened surface of the head of the
femur, close to where the head joins the shaft. The tendon winds
over the joint, and slides over the capsule freely as the bone is flexed
and extended.

If the pelvis and femur are horizontal, the muscle can flex the femur
to 30° extension, can rotate the flexed femur ventrally 30°, and if the
flexed femur has been carried ventrally, can bring it back to the horizon-
tal plane. The power to carry dorsally is increased, and the power to
flex and rotate is lessened, as the leg is extended. If the pelvis is
tilted dorsally, the power of the muscle to carry the femur dorsally is
increased, but its power to flex and to rotate is lessened. When the
thigh is fixed and flexed, the flexor action of the muscle would turn
the body of the frog to the corresponding side, and the two muscles
acting together would tilt the pelvis ventrally. When the thigh is
extended, the muscle would tend to tilt the pelvis dorsally.

Although in certain positions the muscle is a strong flexor, the
most important function of the muscle is to rotate the femur ventrally,
the muscle being the antagonist of the obturator internus. The power
of the iliacus to produce ventral rotation is greatest when the pel-
vis and the femur are in the usual sitting position, z. ¢., when the pel-
vis is tilted dorsally 30°-40°, and the femur is horizontal, flexed to
30° extension, and rotated dorsally 60°-70°. In this position the
iliacus, because of the dorsal rotation of the femur and the dorsal tilt-
ing of the pelvis, is under considerable tension, and its power corre-
spondingly increased. This fact indicates that it is made use of when
the leg is extended during the leap.

The muscle helps to keep the head of the femur in the acetabulum.
The tendons of the iliacus externus and of the caudad and ventral

“s

38 Warren P. Lombard and F. M. Abbott.

fibres of the obturator internus wind round the femur in opposite
directions, that of the iliacus coming from cephalad and above, and
that of the obturator from caudad and below the joint. The iliacus
opposes ventral and caudad displacement, the obturator dorsal and
caudad displacement, and both of the muscles act together, especially
in the squatting position, to bind the head of the bone in the joint,
and prevent it from being dislocated by the leap.

If the flexed femur be rotated dorsally, as in the sitting position,
the tendon of the iliacus is wound round the head of the bone close
to the shaft and prevents extension of the thigh in the horizontal plane
unless ventral rotation occurs (Hering). The thigh can be extended
if it be carried ventrally, however, for then the tendon, by sliding
over the flattened surface of the head of the bone, acquires a new po-
sition with respect to it, the strain being more in the direction of the
long axis of the shaft. The femur could be extended, therefore, if the
surface on which the foot rests when the frog is squatting is removed,
or if the body is raised away from the support. In jumping the lat-
ter occurs, the body being raised from the ground at the same time
that it is thrown forward by the extension of the femurs.

If the femur be extended in the horizontal plane by any means, the
head of the bone moving beneath the tendon brings a strain on the
iliacus externus which causes ventral rotation, even though the muscle
itself does not contract. Any muscle, therefore, which extends the fe-
mur in the horizontal plane, when the femur is flexed and rotated dor-
sally, acts through the iliacus externus to cause ventral rotation.
The amount of rotation produced in this way is only enough to bring
the femur into the position which it holds in the resting extended posi-
tion. This passive action of the iliacus is important, because in
swimming the femur must be rotated ventrally as it is thrust away
from the trunk, so that the force exerted by the extending muscles
may drive the foot backwards, instead of downwards.

It is because of the considerable amount of ventral rotation produced
in the hip joint by this muscle, together with the fact that the lower
leg can be overflexed at the knee, that the frog is able to brush his
foot across his back.

Iliacus internus (Ilio-psoas ). — This muscle arises dorsal and cepha-
lad to the head of the femur, from the rim and cavity of the pelvis
and from the cephalad part of the capsule of the hip, towards the dor-
sal side. The origin from the median surface of the root of the wing
of the ilium within the cavity of the pelvis considerably increases the

Ludiwidual Muscles of the Thigh of the Frog. 39

length of the fibres. The muscle runs across the dorsal part of the
capsule of the hip joint, then along the dorsal surface of the shaft of
the femur, and is inserted on the dorso-median and median surface.

This is a very strong one-joint muscle, which acts as flexor of the
hip in all positions of the thigh. Gad? writes: ‘To bring the extended
leg to the position ready for the spring, the frog needs, at least with
respect to the movements of the hip and knee joints, to innervate only
one muscle, namely, the ilio-psoas.” “ This strong one-joint muscle is
by the frog the flexor of the thigh yar’ é€oynv. There can be no
doubt of its flexor action by any position of the thigh.”

The effect of the iliacus internus to flex the knee joint is produced
indirectly, through the passive tendon action of the semitendinosus
The flexion of the knee ordinarily is quite complete, but overflexion
is not produced in the case of the dead frog. The flexion of the knee
causes a partial flexion of the ankle to about 90° extension, through
the tendon action of the tibialis anticus longus.

When pelvis and femur are horizontal, if the femur has been ex-
tended, the muscle tends to produce flexion of the femur in the horizon-
tal plane. If the femur has been partially flexed and carried dorsally,
this muscle will carry it ventrally as far as the horizontal-plane, and if
the femur has been carried ventrally, the muscle will carry it dorsally
as far as the horizontal plane. The power of the muscle to bring the
limb into the horizontal plane is best seen when the flexor action is
prevented by a smooth instrument placed cephalad to the shaft of
the femur.

If the pelvis has been tilted dorsally, the origin of the muscle is car-
ried dorsally, and if the femur is in the horizontal plane, the muscle
flexes the femur and carries it dorsally. The power of the muscle to
carry the femur dorsally when the pelvis has been tilted dorsally is
important, as it helps to raise the knee off of the ground when the leg
is being drawn into the sitting position.

When the femur is extended, the muscle acts to tilt the pelvis dor-
sally, but when the femur is flexed, it tilts the pelvis ventrally.

Since the insertion of the muscle extends round on to the median
surface of the shaft of the femur, it gives to the muscle the power to
rotate the femur ventrally, which, when the knee is partly flexed, tends
to raise the ankle and foot from the ground. As flexion progresses,
however, from 50° extension on to complete flexion, the capsule of the

1 Gap: Verhandlungen der physikalischen medicinischen Gesellschaft in
Wiirzburg, N. F., 1884, xviii, p. 44.

40 Warren P. Lombard and F. M. Abbote.

joint would inhibit ventral rotation more and more, and if the pelvis
had been carried dorsally would help the obturator internus to pro-
duce the dorsally rotated position of the thigh belonging to the sitting
position. In harmony with this is the fact that as the pelvis has been
tilted dorsally the origin of the iliacus internus is carried dorsally, and
the power of the muscle to rotate the femur ventrally is lessened.
If the femur has been rotated dorsally, however, the muscle can rotate
it ventrally.

Although the iliacus internus acts directly to rotate the thigh ven-
trally, if the pelvis has been tilted dorsally, the iliacus, like any flexing
muscle, may make use of the tendon action of the obturator internus
and cause dorsal rotation of the thigh. It of course could not do this
in case its power to flex were not greater than its power to rotate
ventrally, and the power of the obturator to rotate dorsally were not
greater than the power of the iliacus to rotate ventrally. Naturally
the obturator internus would have to be under tension, in order that
this effect should be produced.

As the thigh is flexed, the iliacus externus and iliacus internus
would, because of the nearing of their points of attachment, have less
and less flexion action. On the other hand, as the thigh is rotated
dorsally by the obturator internus, those muscles which rotate it ven-
trally are put under greater tension, and so acquire a condition ‘favor-
able to their action, when the thigh has to be rotated ventrally during
the extension of the hip by the leap. Undoubtedly the contraction
power of a muscle is greatest when it is under tension and has its
greatest length, and is least when its tension is least and it is short-
est, as is the case when it has completed a movement which it is its
function to produce. Thus the dorsal rotators of the femur have the
least power when dorsal rotation is complete, while the ventral rota-
tors are at this time in the most favorable position for action. Since
the dorsal rotation of the femur which takes place at the close of the
flexion of the thigh, when the pelvis has been tilted dorsally, would
restore the tension of those flexors which are also ventral rotators,
one would be inclined to think that at the moment of the leap they
would by their flexor action interfere with the leap in case they con-
tracted. It is probable, however, that the extensors of the thigh are
sufficiently powerful to overcome the flexion action of these muscles,
and to cause the whole force of their contraction to be directed to ro-
tating the thigh ventrally, Indeed, we have here probably another
example of tendon action. The extensors of the thigh, by the act of

Individual Muscles of the Thigh of the Frog. AI

extension which they produce, increase the tension of the contracting
ventral rotators and thus help in producing the important movement
of ventral rotation.

Ilio-femoralis. — This muscle carries dorsally and rotates ventrally
in all positions of the thigh. These are the only effects when the
limb is extended 80°; if the extension is less than 80°, the muscle
will flex, and if more than 80° will extend, as well as carry dorsally
and rotate ventrally. Carrying the femur dorsally decreases, and
carrying ventrally increases, the tension on the muscle, with corre-
sponding effects on the strength of its action. If the pelvis be tilted
dorsally 40°, and the femur be horizontal, the muscle extends, carries
dorsally, and rotates ventrally in all positions of the bone. The pelvis
is tilted ventrally by this muscle if the femur is flexed, and is tilted
dorsally if the femur is extended at the time contraction occurs.

llio-fibularis. — The ilio-fibularis is a slender two-joint muscle which
lies in the middle of the dorsal surface of the thigh, between the
glutzus magnus and the semimembranosus. It arises as a tendon
just caudad to the point of origin of the glutzus magnus, from the
root of the wing of the ilium just below the processus superior. Thus
the point of origin is dorsal and slightly cephalad and median to the
acetabulum, and to the axis of rotation for flexion and extension in
the head of the femur. The method of insertion at the knee brings
the point of attachment dorsal and median to the distal end of the
femur. When the femur is extended, the muscle lies dorsal and
median to the shaft; when the femur is flexed, the muscle lies diag-
onally across the shaft, because in the process of flexion the tendon
of origin is carried across the head of the femur from the dorsal,
median, and caudad side to the dorsal, lateral, and cephalad side.
This shifting of the tendon of origin from one side to the other of
the head of the femur of course markedly alters the function of the
muscle.

When the pelvis, femur, and os cruris are suspended horizontally,
the muscle is found to flex the femur on the pelvis when the hip is
extended go° or less, but if it is extended 100° or more, the muscle
extends the femur. If the pelvis be tilted dorsally, the femur being
horizontal, the point of origin of the muscle is carried a little more
caudad, 2. ¢., towards the extensor side of the hip joint, and the ex-
tending power of the muscle is increased at the expense of the flex-
ing power, and flexion is rarely seen, even when the hip is completely
flexed.

42 Warren P. Lombard and F. M. Abbote.

The ilio-fibularis can carry the femur dorsally, and so raise the
knee from the ground, this action being most marked when the knee
is extended more than 9o°, and especially when the femur has been
carried ventrally out of the horizontal plane. Moreover, the muscle
can rotate the femur ventrally, this effect being greatest when the
femur has been rotated dorsally, as in the sitting position. The power
to carry dorsally and to rotate ventrally is most evident when flexion
of the knee has been prevented by a pin thrust through the tendons
on the extensor side of the knee, and when the muscle is excited
electrically. The power of the muscle to carry dorsally shows best
when the flexion and extension movements are prevented.

In the following description lateral condyle and lateral ligament are
spoken of, although these in the frog lie dorsally to the knee joint.
The ilio-fibularis is inserted at the knee by a tendon which divides
into two branches which span the dorsal (lateral, Gaupp) side of the
knee joint and lie superficially to the lateral ligament, acting as a sec-
ond lateral ligament to bind the femur to the os cruris. The distal
branch has a broad insertion on the dorsal side of the proximal end
of the shaft of the os cruris, just distal to the lateral condyle, and the
proximal branch, which forms a broad, delicate ribbon, passes round
the dorsal surface of the lateral condyle of the femur and is inserted
well towards the lateral surface of the condyle. The distal branch
of the tendon flexes the os cruris on the femur. At the same time
it has a distinct tendency to carry the crus ventrally and rotate it
dorsally, this tendency being the greater the more the crus is carried
dorsally and rotated ventrally by overflexion. In these respects it is
therefore the antagonist of the semimembranosus, gracilis major, and
semitendinosus. Since the crus is carried dorsally in the process of
overflexion, this muscle would be of service in carrying the crus ven-
trally into the plane of the femur during the extension of the hind
limb. The proximal branch of the tendon, through its attachment
to the dorsal surface of the lateral condyle of the femur, enables the
muscle to carry the femur dorsally.

Gad states that the resting length of the muscle is so small that, if
the hip is flexed, the knee is passively flexed. This: latter statement
is questionable, for the origin of the muscle lies dorsally to the hip
joint, and the tendon slides over the dorsal side of the joint when the
hip is flexed, instead of being wound round the joint, as is the case
with some of the muscles of the median side of the femur.

Obturator externus. — The origin of this muscle is a continuation

Lndividual Muscles of the Thigh of the Frog. 43

of that of the pectineus on to the ischium, z. ¢., lies ventrally and
slightly caudad with respect to the hip joint. The thigh is carried
ventrally by this muscle in all the positions of the limb, the effect
being the more marked the more dorsal the thigh at the time. If
the thigh and pelvis are horizontal, and, the supporting thread being
held, the thigh is prevented from moving ventrally, the muscle acts as
an extensor when the thigh is extended more than 70°, and acts as a
flexor when the thigh is flexed less than 70° extension. When the
thigh is carried dorsally or ventrally, like effects are observed, except
that when it is carried ventrally the action is less marked.

If the pelvis be tilted dorsally to the sitting position, the origin of
the muscle is carried somewhat cephalad with respect to the head of
the femur. If now the muscle be excited, the thigh is carried ven-
trally ; or if this is prevented, it is flexed if extended and extended
if flexed, the dead point being at about 110° extension.

If the femur is fixed, the muscle tilts the pelvis dorsally if the
femur is flexed, and ventrally if it is extended.

Obturator internus. This muscle surrounds two thirds of the
acetabulum, from the cephalad and ventral to the dorsal side. The
fibres take origin from the lateral surface of the pelvic disk, just in-
side the border, all the way from the anterior to the posterior spine.
They all act on a tendon which is inserted in a slight groove which
runs transversely across the dorsal flattened surface of the head of
the femur, just proximally to the point of attachment of the tendon
of the iliacus externus. The tendon is attached, not, as Gaupp says,
to the capsule, but to the head of the bone. It soon penetrates the
synovial membrane and the capsule, which surrounds it for a short
distance as a sleeve, and spreads out upon the capsule over the head
of the bone.

The tendon consists of a short and relatively broad ribbon, which
is prolonged ventrally as a strip of fibrous tissue, which runs half-way
round the joint on the caudal and ventral sides, in the form of a thin
collar. The cephalad, ventral, and caudad fibres of the muscle
run obliquely to join this collar, forming a whirl about the joint. If
the fibres be cut away from their origin on the disk of the pelvis and
the muscle be raised up, it will be seen that the fibrous collar on
which the fibres are inserted lies in a groove on the cartilago re-
manens. The groove is formed by the ventral border of the acetabu-
lum, where it projects laterally from the disk of the pelvis. As the
ventral fibres of the muscle contract, the fibrous collar draws over the

AA Waysen £. Lomiitd aad Fee

smooth cartilage forming the groove as round a pulley. The dorsal
and some of the caudal muscle fibres take a more direct course, and,
converging from their point of origin, are inserted on the short broad
head of the tendon. This latter group of fibres differs both in its ana-
tomical relations and its physiological action, so greatly from the
former group as almost to deserve another name. We shall refer to
the two groups as the ventral and dorsal fibres.

The tendon glides over the capsule and the head of the femur as it
is flexed and extended. Its position on the head of the bone is se-

cured in part by loose bands of fibrous tissue which, though connect- -

ing it with the capsule, permit it to shift its position upon it within
certain limits. It is also held in place by the fibres of the muscle
itself, especially the dorsal and caudal fibres. Indeed the chief func-
tion of these fibres seems to be to act as guy ropes to prevent the
tendon from sliding forwards over the head of the bone when the
ventral fibres contract, and so enable the ventral and cephalad por-
tions of the muscle to do its important work of rotating the thigh
dorsally.

If the femur is horizontal and is extended, the tendons of the ob-
turator internus and iliacus externus are seen to be attached to the
median side of the dorsal flattened surface of the head of the bone,
the obturator being inserted in a slight groove which lies median and
somewhat cephalad to the little trochanter which forms the point of
insertion of the iliacus externus.

As the extended bone is flexed to 90° extension, one’ sees the
points of insertion of these muscles describe an arc about the centre
of rotation for flexion and they come to lie caudad to this centre. As
the bone is flexed to complete flexion, the points of insertion come to
lie laterally from the centre of the head of the bone, and the groove
where the obturator is inserted is seen to be caudad and median to
the trochanter, instead of cephalad and median as when the femur is
extended.

This change of position of the point of insertion of the obturator
with respect to the axis of rotation for flexion and extension neces-
sarily alters the action of the muscle. Moreover, if the femur remains
in the horizontal plane, and the pelvis be rotated about its transverse
axis, z. ¢., be tilted ventrally or dorsally, the tension and the direction
of pull of both the ventral and dorsal fibres of the muscle is

changed with respect to the femur, and the action of the muscle is
altered.

Lndividual Muscles of the Ti high of the Frog. 45

The muscle helps to bind the bone in the socket (see iliacus ex-
ternus).— This action is very important, because the acetabulum is
very shallow, taking in, indeed, less than a third of the head of the
bone, and the capsule is so loose that, aside from securing the air-
tight closure of the joint, it does little to prevent the head of the
bone from leaving the socket, except when it has been put under
tension by marked ventral or dorsal rotation.

In the flexed position of the femur the cephalad and ventral fibres
of the obturator internus by contracting draw the fibrous collar to
which they are inserted firmly against the head of the bone, bind it
in the socket, and oppose backward and dorsal displacement. The
dorsal and caudad fibres not only assist by preventing the fibrous col-
lar from sliding forwards over the head of the bone, but act directly
to prevent ventral and forward displacement. This action of the
muscle is most marked in the sitting position, since tilting of the
peivis dorsally puts the fibres under tension.

Rotation effects. To study the action of the obturator internus,
make a pelvis-femur preparation, with the os cruris cut off just below
the knee, and all the other muscles divided or removed. Fix the pelvis
in the horizontal position and suspend the femur so that it can move
freely in the horizontal plarie. When the preparation is thus placed,
it will be found that the femur can be flexed and extended without
any opposition being encountered, and with but very little rotation
occurring. If, however, the pelvis is tilted dorsally 40° and the femur
is passively flexed, a dorsal rotation of the bone is observed, this be-
ginning when the flexion process has carried the femur beyond 90°,
and increasing as complete flexion is approached. Tilting the pelvis
dorsally puts the ventral fibres of the obturator internus on the
stretch, and flexion of the joint increases the tension on the muscle.
Any muscle or force, therefore, which will flex the femur horizontally
when the pelvis is in the sitting position will tend to act through the
ventral fibres of the obturator internus to rotate the femur dorsally,
that is, help to produce the rotation which is normally present when
the frog is in the ordinary squatting position.

The amount to which the ventral fibres of the obturator internus
can rotate the femur by contracting differs in different positions of
the hip joint. It is greatest when the femur is flexed and the
pelvis is tilted dorsally, z. ¢., in the squatting position, where the dor-
sal rotation may be 70°, as much as the ligamentum ventrale will
permit. If the pelvis be brought into the horizontal plane, the dorsal

46 Warren P. Lombard and F. M. Abbote.

rotation can only be 50°. If the femur be extended, the amount of
dorsal rotation possible lessens, until by complete extension it is
almost absent.

The importance of the dorsal rotation produced by the ventral
fibres of the obturator internus during the latter part of the flexion
of the femur, when the thigh is being given the squatting position, is
very great. It is through the rotating action of this muscle that the
distal end of the femur comes to be supported on the central end
of the os cruris instead of lying beside it. This muscle must act,
moreover, during the early part of the extension of the thigh by the
leap in order that the muscles extending the thigh may cause the
trunk to be raised up and thrown forward, instead of driving the leg
out sidewise.

The dorsal fibres may rotate ventrally when the femur is flexed,
and dorsally when the femur is extended. The ventral rotation of the
flexed femur is best seen when the femur is prevented from being
carried dorsally. These fibres are weak, and neither of these rotation
effects is strong.

Flexor and extensor action. — These effects can be ascertained best
by electrical excitation of different parts of the muscle. If the pelvis
and femur are horizontal, the muscle is found to extend the femur if
it be flexed and to flex it if it be extended. The dead point is at from
110° to 120°. Both the ventral and the dorsal fibres act thus, and the
reversal of action is due to the change in the position of the point of
insertion of the muscle on the head of the bone with respect to the
axis, about which flexion and extension occur. It is doubtful whether
the muscle acts either as flexor or as extensor strongly enough to be
of much service.

Carrying the femur dorsally. —The dorsal fibres of the obturator
may produce this effect, and act the more, the more the femur has
been carried ventrally. In the flexed position, as the bone is rotated
dorsally, the part of the tendon connected with the dorsal fibres draws
more and more across the central portion of the head of the femur,
which is now directed backward and medially, and helps to carry the
distal end of the bone dorsally.

Pectineus. — The muscle arising from the pelvic disk, ventrally and
somewhat cephalad to the hip joint, runs obliquely across the ventral
surface of the femur, and is inserted in a long line on the crista femo-
ris almost as far as the middle of the bone.

The chief action of the muscle is to carry the thigh ventrally; in

Individual Muscles of the Thigh of the Frog. 47

addition to this it is a flexor. The power to carry ventrally is greatest
when the thigh has been carried dorsally and extended, since carry-
ing the thigh ventrally and flexing it relaxes the muscle. The flexor
action comes more and more into prominence as the extended leg is
carried ventrally.

If the pelvis is carried dorsally, the line of origin of the muscle
comes to lie more cephalad to the head of the femur, and in the sit-
ting position the flexor action of the muscle is increased at the ex-
pense of the power to carry ventrally.

Pyriformis. — This feeble muscle acts to carry the thigh dorsally
and to depress the urostyle. It has in addition to this a slight ex-
tensor action. .

Quadratus femoris. — This weak muscle lies caudad to the hip joint,
and acts as an extensor of the thigh in all of its different positions.

Sartorius. — The sartorius is a long-fibred, ribbon-like muscle which
lies along the middle of the ventral surface of the thigh. It arises by
a thin, broad tendon from the ilium, on the forward part of the ventral
border of the pelvic symphysis, and terminates on the ventral side of
the knee in a narrow, thin tendon, which spreads out to form a tri-
angular expansion. The more proximal part of the expanded portion
passes beneath the median border of the tendon of the triceps, with
which it is connected by bands of fibrous tissue, and is inserted on
the ventral surface of the median condyle of the crus, well forwards
and towards the lateral side, close to the insertion of the gracilis
major, the tendon of which is covered by the tendon of the sartorius
(see Fig. 2, Plate 1). The more distal part of the tendon of the sar-
torius is inserted on the middle and proximal part of the ventral sur-
face of the expansion of the tendon of the semitendinosus. This
method of insertion enables the muscle to make use of the peculiar
tendon of the semitendinosus, and should give it a similar action to
flex the extended crus, and, as it flexes, to carry it dorsally and rotate
it ventrally.

In studying the action of this slender muscle by electrical excita-
tion, the lower leg was cut off at about its upper third, so as to get
rid of most of the weight of the leg and foot. The effect of the
spread of current was avoided, in some cases by freeing the muscle
from its neighbors and placing a strip of rubber membrane between
the sartorius and the adjacent muscles. In other cases, all the other
thigh muscles were cut away, except that the one-joint muscles of
the hip and the central end of the cruralis where it is attached to the
capsule were left, so that the direction of the strain of the contract-

48 Warren P. Lombard and F. M. A bbote.

ing sartorius should not be altered. Of course, care was taken to

watch for spread of current. The following effects were obtained also
by electrical excitation of the nerve:

Effect on the hip. —The muscle can flex the femur on the pelvis in
all positions of the femur. If the muscle be cut across the middle
when it is at rest and the hip and knee are extended, the cut ends
slightly overlap, and when the hip and knee are flexed, the cut ends
overlap markedly. Were it not that the fibres of the muscle are very
long, and consequently capable of great shortening, the muscle would
have little or no power as flexion of the hip and knee neared com-
pletion. As the flexion of the hip takes place, the thigh is at the
same time carried ventrally. This latter action of the muscle is the
stronger, the more the thigh is flexed and the more it has been carried
dorsally. When the thigh has been carried dorsally, the power of the
muscle to carry it ventrally is greater than its power to flex the thigh
on the pelvis. Tilting the pelvis has no appreciable effect on the
character of the movements produced.

The effect on the knee joint. —In the extended position of the knee
the strain from the contracting sartorius is transmitted through the
tendinous expansion of the semitendinosus tendon to the crus. The
line of attachment of the semitendinosus tendon on the crus extends
from distal and median, proximally and laterally, diagonally across the
proximal part of the ventral surface of the crus, and when the muscle
is excited it is seen to act like the semitendinosus, to flex the knee,
and to carry the flexing crus dorsally, at the same time that it rotates
it ventrally. The more proximal part of the triangular expansion of
the sartorius tendon being inserted towards the lateral border of the
ventral surface of the median condyle of the crus gives to the muscle
a marked rotation effect when the knee is over-flexed.

Semimembranosus. — This is a strong two-joint muscle on the dorso-
median side of the thigh. The dorsal border covers the proxi-
mal part of the ilio-fibularis, which in the distal part of the thigh
shows itself between this muscle and the dorsal head of the
triceps, the gluteus magnus. The median border is in contact with
the gracilis major, and partly covered by the gracilis minor. The
semimembranosus arises from a broad insertion from the lateral
surface of the pelvic disk, chiefly from the ischium, and dorsal and
slightly median to the acetabulum. The muscle lies more median
_ than dorsal to the shaft of the femur, and is inserted towards the
median side of the knee on the proximal end of the os cruris and the
ligaments within the joint (see Plate II, Figs. 1, 2).

Individual Muscles of the Thigh of the Frog. 49

This muscle extends the thigh on the pelvis, when the pelvis and
thigh are in the horizontal plane, or when either of these has been
carried either dorsally or ventrally out of this plane. It also tends to
carry the thigh dorsally in all positions of the thigh, this action
showing itself most markedly when the thigh is prevented from
extending and when it has been carried ventrally.

The action on the pelvis is not quite so simple. As has been
explained elsewhere (p. 17), when the thigh is extended 90°, carrying
the thigh dorsally has the same effect at the hip joint as rotating the
pelvis about its long axis: when the thigh is flexed, carrying the
thigh dorsally has the same effect as tilting the pelvis ventrally; and
when the thigh is extended, carrying the thigh dorsally has the same
effect as carrying the pelvis dorsally. Since this muscle always
carries the thigh dorsally when it is free to move, one might expect
that when the thigh was fixed and the pelvis was free, this muscle
would produce the movements of the pelvis just described. Experi-
ment shows that in all positions of the thigh the muscle tends to
extend the pelvis on the thigh. It also tends to rotate the pelvis
about its long axis, and, in most positions of the thigh, to tilt the
pelvis dorsally; but if the thigh be flexed beyond 50° extension, the
muscle will tilt the pelvis ventrally.

In the sitting position, that is, with the thighs in the horizontal
plane and extended to 40°, and the pelvis tilted dorsally 40°, the
muscle acts strongly to extend the thigh, to carry the thigh dorsally,
and totilt the pelvis ventrally. In the early part of the leap it would
propel the body forwards, would raise the knee off of the ground, and
would tend to tilt the pelvis ventrally, opposing those muscles which
act to tilt it dorsally. By carrying the pelvis ventrally in the early
part of the leap, it would help to keep the weight of the trunk well
forward and oppose the action of the powerful gastrocnemius
muscle, which by vigorously extending the ankle joint tends to
throw the frog over on his back. When the thigh was extended
more than 50°, and especially in the latter part of the leap, the
muscle while still acting to extend the thigh, and to raise the knee
from the ground by carrying the thigh dorsally, would act to carry
the pelvis dorsally and thus give the body a position favorable to
landing on the hind feet.

In addition to the above effects the muscle tends to rotate the
thigh ventrally, since it winds round the femur somewhat in passing
from the point of origin dorsal to the head of the bone, to its point of

50 Warren P. Lombard and F. M. A bbote.

insertion on the median condyle of the os cruris. This rotation effect
is most when the femur is rotated dorsally in the sitting position.
The action of this muscle on the knee is a very remarkable one,
and requires a brief statement concerning the relation of its tendon
to the joint. The tendon of insertion divides into two short
branches, these forming a Y, the arms of which pass to either side of
the median condyle of the os cruris. One of the arms of the Y runs
in the dorsal (lateral, Gaupp) direction, and is inserted beneath the
tendinous arch forming the origin of the plantaris longus, in the
most proximal part of the fossa intercondylia and on the ligaments
within the joint, while the other arm runs in the ventral direction
and is inserted on the most proximal part and to the ventral
(median, Gaupp) side of the median condyle of the os cruris. These
two branches of the tendon form an arch round the median condyle,
which plays between them as it rolls on the distal end of the femur
during the early part of the act of flexion. Still another branch
which is given off dorsally (laterally) from the tendon is closely
connected with the internal ligaments of the joint and passes to the
most proximal and dorsal (lateral) side of the lateral condyle.
When the crus is completely extended, the dorsal branch, especially of
the Y tendon, which is attached to the os cruris in the most proximal
part of the fossa intercondylia, slides the proximal end of the os
cruris on the distal end of the femur and flexes the completely
extended crus a short distance to 150° extension. If the crus is
passively flexed beyond this amount, the rotation surfaces of the
two condyles come to bear on the end of the femur, and the points of
attachment of the tendon of the muscle are wound under them and
carried into the joint. The tendon now draws over the distal end of
the femur as round a pulley, and when the muscle contracts it causes
the condyles of the os cruris to roll on the head of the femur and
extends the os cruris as far as 140 extension. This happens if the
thigh holds a position between 100° and 140° extension. If the os
cruris be still further passively flexed to less than 100° extension, the
flattened median surfaces, instead of the proximal rounded rotation
surfaces of the condyles of the os cruris, come in contact with the
head of the femur, which now lies in the intercondyloid fossa, and
the muscle causes the end of the os cruris to slide on the femur in
such a way as to flex the knee, rotate the os cruris ventrally, and
carry it dorsally with respect to the femur. These results are largely
due to the shape of the joint surfaces and the action of the lateral

Individual Muscles of the Thigh of the Frog. 51

ligaments of the knee, for, if either the dorsal or the ventral branch
of the Y be cut, the other branch acting alone can produce all of
these movements, — flexion, when the extension present is 150° or
more, extension when the extension present is between 100° and
140°, and flexion, rotating ventrally and carrying dorsally, from 90°
extension to complete over-flexion. These figures differ slightly by
different preparations.

In the early part of the leap the semimembranosus acts to extend
the hip, but opposes the extension of the knee. This is of use,
because, if the knee were extended before the hip, the leg would be
driven out sidewise instead of backwards. This opposition is
overcome when the thigh is partly extended by the powerful ex-
tensors on the lateral side of the knee; and when the knee has been
extended beyond about 100°, a dead point is passed, and the flexor
action of the semimembranosus is suddenly changed to an extensor
action. Up to this point the flattened median surfaces of the fossa,
intercondylea of the os cruris were sliding on the median surface
of the head of the femur, but now the proximal rotation surfaces
of the condyles suddenly come in contact with and roll on the distal
end of the femur. In passing the dead point the contracting semi-
membranosus is put under elastic tension, and when its flexor action
is changed to an extensor action, it behaves like a spring that has
been stretched and is suddenly released and gives a sudden impulse
to the extending leg. Towards the close of the leap, when extension
has passed 150°, the flexor action of the muscle again shows itself and
helps to check the extension movement, lessen the strain on the knee
joint, and favor the beginning of the recovery of the leg to the flexed
position needed for the next leap.

Gad! says that the semimembranosus and rectus internus (the
gracilis major) cause, like the two-joint semitendinosus and ilio-
fibularis (biceps), besides extension of the hip, flexion of the knee.
He goes on to say that they do not have to be contracted to cause
the latter movement, for their natural length even in the resting
position is so small that by the dead animal the flexion of the hip
produced by the ilio-psoas (iliacus internus) without anything
further is associated with flexion of the knee. This effect belongs in
the case of dead muscles only to the semitendinosus, and this can
passively flex the knee by being wound up round the hip when that
joint is flexed. If the pelvis of a frog be held with the cephalad end

2 GAD: “0c. cit., DP: 172:

52 Warren P. Lombard and F. M. Abbott.

downwards, the thigh is flexed by its own weight, and the knee is
seen to be flexed at the same time, provided the foot has been cut off
and the muscles of the lower leg have been removed, to get rid of the
stretching caused by their weight. If now the tendon of the semi-
tendinosus be cut, the knee immediately straightens to the position
of almost complete extension.

Semitendinosus. — This is a two-joint muscle which lies beneath
the gracilis major on the median side of the thigh. It arises from
the pelvis by two heads, one of which lies to the dorsal and the other
to the ventral side of the head of the femur. The two bellies of the
muscle come together distally, and are inserted by asingle tendon on
the median surface of the proximal end of the os cruris.

Both of the heads of the muscle act to extend the femur on the
pelvis in all positions of the bone. The dorsal head, at the same time
that it extends, carries the femur dorsally, and the ventral head, at
the same time that it extends, carries the femur ventrally. When the
two heads act together, simple extension is seen. The power of the
two heads to carry the femur out of the horizontal plane is best seen
when extension is prevented.

Both heads of the muscle act through the common tendon to flex
the os cruris on the femur. The method of attachment of the tendon
on the os cruris is so remarkable, and reveals the function of the
muscle so clearly, that it deserves a special description (see Plate I
and Plate I, Fig. 2). Toappreciate the adaptation of the method of
insertion to the work which the muscle has to do, one must recall that
the flexion movement of the knee of the frog is not limited to 180° or
less, as in the case of most joints, but that, as the flexing os cruris
approaches the femur, it is carried dorsally and rotated ventrally, so
as to be able to pass by the femur, and that flexion may be carried on
into an over-flexion of 40—50°.

As the crus passes from extreme extension to extreme over-flexion,
it is first simply flexed, and then, as it approaches the thigh, is flexed,
carried dorsally and rotated ventrally. The attachment of the semi-
tendinosus is such as to enable it to act in harmony with the joint sur-
faces of the knee to produce all of these movements of flexion of the
crus, of carrying it dorsally and rotating it ventrally, and to produce
each of these effects most strongly at the time that it is most needed,

The slender tendon of this muscle divides near its termination, one
branch being attached close to the proximal end, and well towards
the lateral border of the ventral side of the os cruris, and the other

Individual Muscles of the Thigh of the Frog. 53

further down the bone and towards the median border of the ventral
surface. Between these two branches there is a membranous expan-
sion, which is attached along the line connecting the points of inser-
tion, and inside of this there is a second membrane, the two
membranes forming a pocket. The line of insertion runs, therefore,
from before backwards diagonally across the ventral side of the crus,
and from the lateral towards the median border. The relation of the
triangular expansion of the semitendinosus tendon to the bone is such
that when the knee is flexed the more distal fibres are slack, and when
the knee is extended the more proximal fibres are slack. As the
knee is extended, the tension is transferred from the proximal and
lateral to the distal and median fibres, and as it is flexed, from the
distal and median to the proximal and lateral fibres. By this arrange-
ment the main tendon is kept always close to the joint. By extended
leg the conditions are favorable to simple flexion, and, the power being
applied to the os cruris at a considerable distance from the joint, the
flexor action is a strong one. As the leg flexes, the power of the
muscle to rotate the crus ventrally and carry it dorsally increases
more and more at the same time that the flexor action lessens. _

The method of action of the semitendinosus is of advantage during
extension as well as flexion, and although the muscle is to be classed
chiefly as a flexor of the crus, it probably contracts vigorously when the
crus is being extended in the leap; the advantage of this has been
explained in part in the description of the action of the semi-mem-
branosus. When the leg is flexed in the sitting position, the muscle
acts strongly to carry the crus dorsally with respect to the femur and
to produce a ventral rotation at the knee, but has only a feeble flexor
action; thus it can contract vigorously during the leap and help to
hold the lower leg away from the thigh as it slides past it. As the
extensor action approaches completion, however, the flexor action of
the muscle becomes greater and greater, and it tends to lessen the
strain which would be brought on the knee joint at the end of the
leap, and aid in the rapid recovery of the limb to the flexed position
required for the next stroke.

As Gad observed, the resting length of this two-joint muscle is so
short that the knee is passively flexed by it when the hip is flexed. The
flexion is not complete, however, or rather the muscle does not cause
over-flexion unlessin tonus. This action of the muscle can be readily
seen by holding a pelvis limb preparation with the pelvis pointing
downward. The thigh is then flexed by its own weight. If the foot

54 Warren P. Lombard and F. M. Abbott.

and muscles of the lower leg are present, their weight is sufficient to
stretch the dead muscle and prevent its tendon action from showing
itself; but if the foot and muscles of the lower limb are cut away, the
os cruris is seen to flex to nearly complete flexion. That this effect is
produced by the semitendinosus and not by the other two-joint mus-
cles which may flex the knee, is proved by cutting the tendon of the
semitendinosus, when the os cruris immediately extends.

Tensor fascie late. — The tensor facie late, the median and weakest
head of the triceps group, lies on the lateral side of the thigh. It
arises from the ventral border of the wing of the ilium, a little cephalad
to the belly of the iliacus externus. With pelvis and thigh horizon-
tal, the fibres take a slightly ventral course, covering the dorsal border
of the cruralis, and are inserted into the common fascia of the triceps.
When the limb is in complete extension, the muscle is quite tense,
but in complete flexion the fibres are very lax, and could exert but
little traction power.

With the pelvis and thigh in the horizontal plane, and the leg in
the resting, semi-extended position, the muscle causes the femur to
be flexed and the os cruris to be extended until the limb forms an
angle of about 90° with the long axis of the pelvis. If the crus be
flexed on the femur and be kept from extending, the muscle will cause
the femur to be completely flexed. If the limb be carried ventrally
and the femur be prevented from flexing, the muscle will carry the
thigh dorsally above the horizontal plane. Tilting the pelvis dorsally
40° increases the power of the muscle to carry the limb dorsally, but
does not prevent the flexor action. In the sitting position the muscle
acts the same as the cruralis, only is much weaker. This muscle tends
to carry the pelvis ventrally, and opposes the action of the chief ex-
tensors of the hip, which tend to carry the pelvis dorsally.

Triceps. — The triceps consists of three muscles, glutzus magnus,
cruralis, and tensor fasciz late, which cover the lateral and a part
of the dorsal and ventral aspects of the thigh. They all arise from
the pelvis, cross over the hip and knee joints, and are inserted by a
common tendon on the os cruris. The action of each of the three
parts of the muscle has been described separately.

Lndividual Muscles of the Thigh of the Frog. 55

SUMMARY OF ACTION OF THE MUSCLES OF THE THIGH ON THE FEMUR, PELVIS,
AND Os CRURIS.
Adductor longus.
Flexes femur and to a less extent carries ventrally (feeble).
Adductor magnus.
Ventral head.
Effect on femur.
Pelvis and femur horizontal.
Flexes femur if it is extended less than 45°.
Extends femur if it is extended more than 45°.
Carries femur ventrally (strong).
Pelvis tilted dorsally, and femur horizontal.
Flexes femur if it is extended less than go°.
Extends femur if it is extended more than go°.
Carries femur ventrally (stronger).
Effect on pelvis (strong).
Tilts pelvis dorsally if femur is flexed.
Tilts pelvis ventrally if femur is extended.
Dorsal head.
Effect on femur.
Pelvis and femur horizontal.
Extends femur (strong).
Rotates femur dorsally (strong).
Effect to carry femur dorsally or ventrally slight.
Femur flexed.
If it is already dorsal, carries it dorsally.
If it is already ventral, carries it ventrally.
Femur extended.
If already dorsal, carries it ventrally.
If already ventral, carries it dorsally.
Pelvis tilted dorsally and femur horizontal.
Extends femur.
Rotates femur dorsally (strong).
Carries femur ventrally.
Cruralis.
Effect on hip.
Flexes femur (not strong).
Binds head of femur in acetabulum from cephalad side.
Effect on knee.
Extends os cruris on femur in all positions of the limb.

56 Warren P. Lombard and F. M. Abbott.

Gemellus.
Effect on femur.
Extends femur ; most when pelvis tilted dorsally.
Carries femur dorsally if it has been carried ventrally.
Effect on pelvis.
Tilts pelvis ventrally if femur flexed.
Tilts pelvis dorsally if femur extended.
Gluteus magnus.
Effect on femur.
Pelvis and femur horizontal.
Flexes femur in all degrees of extension.
Carries femur dorsally (strong).
Pelvis tilted dorsally and femur horizontal.
Flexes femur if pelvis tilted dorsally less than 20°.
Extends femur if pelvis tilted dorsally more than 20°.
Carries femur dorsally.
Effect on pelvis (strong).
Tilts pelvis ventrally if femur flexed.
Tilts pelvis dorsally if femur extended.
Effect on knee.
Extends os cruris on femur in all positions of the limb.
Gracilis major.
Effect on femur.
Extends femur on pelvis (strong).
Carries femur ventrally ; most when pelvis tilted dorsally.
Rotates femur ventrally ; most when femur rotated dorsally.
Effect on pelvis.
Tilts pelvis dorsally when already dorsal.
Effect on knee.
Flexes os cruris into over-flexion when extended less than 110°.
Extends os cruris when extension between 120° and 140°.
Flexes os cruris if extension more than 140°.
Rotates os cruris ventrally in over-flexion.
Carries os cruris dorsally in over-flexion.
Gracilis minor.
Puts the skin under tension.
Tliacus externus.
Binds head of femur in acetabulum from dorsal and caudad sides.
Effect on femur.
Flexes femur (strong) ; less when pelvis tilted dorsally.
Carries femur dorsally ; most when femur ventral and extended.
Rotates femur ventrally (strong) ; most when pelvis is tilted dorsally
and the femur flexed.

Individual Muscles of the Thigh of the Frog.

Tliacus externus (comtinued).
Effect on pelvis.
Tilts the pelvis ventrally if femur flexed.
Tilts pelvis dorsally if femur extended.
Tliacus internus.
Effect on femur.
Pelvis and femur horizontal.
Flexes femur, strong.
Carries femur ventrally if it is dorsal.
Carries femur dorsally if it is ventral.
Rotates femur ventrally.
Pelvis tilted dorsally and femur horizontal.
Flexes femur (strong).
Carries femur dorsally.
Rotates femur ventrally (but effect not strong).
Effect on pelvis.
Tilts pelvis ventrally if femur flexed.
Tilts pelvis dorsally if femur extended.
Tlio-femoralis.
Effect on femur.
Pelvis and femur horizontal.
Flexes femur if it is extended less than 80°.
Extends femur if it is extended more than 80°.
Carries femur dorsally.
Rotates femur ventrally.
Pelvis tilted dorsally and femur horizontal.
Extends femur.
Carries femur dorsally.
Rotates femur ventrally.
Effect on pelvis.
Tilts pelvis ventrally if femur flexed.
Tilts pelvis dorsally if femur extended.
Tlio-fibularis.
Effect on femur.
Pelvis and femur horizontal.
Flexes femur if extended less than go.
Extends femur if extended more than go°.
Carries femur dorsally ; most if extended more than go°.
Rotates femur ventrally.
Pelvis tilted dorsally and femur horizontal.
Extends femur.
Carries femur dorsally.
Rotates femur ventrally.

SF J

58 Warren P. Lombard and F. M. Abbott.

Tlio-fibularis (continued ).
Effect on pelvis.
Tilts pelvis ventrally if femur flexed.
Tilts pelvis dorsally if femur extended.
Effect on knee.
Flexes os cruris of femur.
Carries os cruris ventrally when flexed.
Rotates os cruris dorsally when flexed.
Obturator externus.
Effect on femur.
Pelvis and femur horizontal.
Flexes femur when it is extended less than 70°.
Extends femur when it is extended more than 70°.
Carries femur ventrally.
Pelvis tilted dorsally and femur horizontal.
Action much the same, except the dead point between flexion
and extension is at 110.°
Effect on pelvis.
Tilts pelvis dorsally if femur flexed (strong).
Tilts pelvis ventrally if femur extended (strong).
Obturator internus.
Ventral head.
Binds head of femur in acetabulum, from ventral and caudad sides.
Effect on femur.
Pelvis and femur horizontal.
Flexes femur if extended more than 110°.
Extends femur if extended less than 110°.
Carries femur dorsally, most if it has been carried ventrally.
Rotates femur dorsally (weak).
. Pelvis tilted dorsally, and femur horizontally.
Flexes and extends about the same.
Rotates femur dorsally when extension go° or less; the
chief action, and the stronger the greater the flexion.
Effect on pelvis.
Tilts pelvis dorsally if femur flexed.
Tilts pelvis ventrally if femur extended.
Dorsal head.
Effect on femur.
Flexes femur if extended more than 110°.
Extends femur if extended less than 110°.
Carries femur dorsally.
Rotates femur ventrally if it is flexed.
Rotates femur dorsally if it is extended.

Individual Muscles of the Thigh of the Frog. 59

Pectineus.
Effect on femur.
Flexes femur ; most when femur carried ventrally ; action increased
when pelvis tilted dorsally.
Carries femur ventrally ; most when femur extended and carried
dorsally ; action less when pelvis is tilted dorsally.
Effect on pelvis.
Tilts pelvis dorsally if femur flexed.
Tilts pelvis ventrally if femur extended.
Pyriformis.
Effect on femur.
Extends femur (feeble).
Carries femur dorsally.
Effect on urostyle.
Depresses.
Quadratus femoris.
Extends femur.
Sartorius.
Effect on femur.
Flexes femur on pelvis in all positions. é
Carries femur ventrally ; most when flexed and carried dorsally.
Effect on pelvis.
Tilts pelvis dorsally if femur flexed.
Tilts pelvis ventrally if femur extended.
Effect on knee.
Like semitendinosus, flexes the os cruris, carries it dorsally and
rotates it ventrally in over-flexion.
Semimembranosus.
Effect on femur.
Extends femur.
Carries femur dorsally.
Rotates femur ventrally when rotated dorsally.
Effect on pelvis.
Tilts pelvis dorsally if extension more than 50°.
Tilts pelvis ventrally if extension less than 50°.
Effect on knee.
Flexes os cruris into over-flexion when extension less than roo®,
Extends os cruris when extension between 100° and 140°.
Flexes os cruris when extension more than 150°.
Carries os cruris dorsally in over-flexion.
Rotates os cruris ventrally in over-flexion.

60 Warren P. Lombard and F. M. Abbot.

Semitendinosus.
Effect on femur.
Extends femur.
Carries femur dorsally by dorsal head.
Carries femur ventrally by ventral head.
Effect on pelvis. ,
Ventral head.
Tilts pelvis dorsally if femur flexed.
Tilts pelvis ventrally if femur extended.
Dorsal head.
Tilts pelvis ventrally if femur flexed.
Tilts pelvis dorsally if femur extended.
Effect on knee.
Flexes the os cruris in all positions (strong). —
Carries the os cruris dorsally in over-flexion.
Rotates the os cruris ventrally in over-flexion.
Tensor fasciae latae.
Effect on femur.
Pelvis and femur horizontal.
Flexes femur.
Carries femur dorsally.
Pelvis tilted dorsally and femur horizontal.
Flexes femur.
Carries femur dorsally more strongly.
Effect on pelvis.
Tilts pelvis ventrally if femur flexed.
Tilts femur dorsally if femur extended.
Effect on knee.
Extends os cruris on femur in all positions of limb.

BURTHER: OBSERVATIONS ON THE RESUSCITATION
OF THE RESPIRATORY NERVOUS MECHANISM.

By. Gs N. STEWART ann). F >, H., PIKE.
[From the Hull Physiological Laboratory of the University of Chicago. |

I. THe AUTOMATISM OF THE RESPIRATORY CENTRE.

N a recent number of this Journal we! published experiments on
the resuscitation of certain of the bulbar centres, including ob-
servations on the automatism of the respiratory centre. We endeav-
ored to isolate the centre from all afferent impulses by subjecting the
brain, including the bulb and cervical spinal cord, to varying periods
of anemia. After restoration of the circulation a stage can be found
during which the respiratory centre is discharging itself rhythmically,
while as yet excitation of such afferent nerves as may be presumed
to affect it most easily is ineffective. Although at this stage there
is every reason to believe, both on general grounds and on certain
special evidence already dealt with, that the connections of the bulbar
centre with the upper parts of the brain have not yet become capable
of conduction, it seemed worth while to supplement the experiments
referred to by another series in which, in addition to eliminating
totally, as was believed, both upper and lower paths by anzmia, the
anatomical elimination of the most important of them was accom-
plished by the knife. In the previous paper the general result of
these observations is mentioned, but only in the briefest way.
The experiments were performed on eleven cats and two rabbits.
In all the animals (anzsthetized by ether) the upper paths were
severed by dividing the brain stem above the level of the respiratory
centre by means of a thin probe passed through a trephine hole.
The level of the section varied in different experiments from about
the middle of the fourth ventricle to the upper border of the pons.
Where shock is avoided by making the section sufficiently late in the

1 STEWART and PIKE: This journal, 1907, xix, p. 328.
61

62 G. N. Stewart and F. H. Pike.

occlusion period, or sufficiently early in the period of resuscitation,
we find, without exception, that the behavior of the respiratory
nervous mechanism is the same as that previously described when
anemia alone is relied on to temporarily isolate the centre from
afferent impulses. For instance, if the brain stem be divided at a
stage in the occlusion when, after the first cessation of respiration, the
secondary gasps have returned and have been going on for some time,
while stimulation of the vagus or brachial plexus has ceased to influ-
ence them, the gasps go on unaltered either in rhythm or in depth.
The same is true if the section be made in the resuscitation period,
soon after the return of respiration, when the afferent nerves tested
are still ineffective. Section of both vagi, in addition to section of
upper paths, is equally without influence on the respiratory move-
ments at this time. We have already shown that elimination of the
lower part of the spinal cord by ligation of the thoracic aorta does
not essentially affect the return of respiration in resuscitation. In
two experiments the cord was also divided with the knife in the lower
cervical region.

It is completely in agreement with our view (that at the moment
when respiration returns in resuscitation the respiratory centre is
still isolated from afferent impulses, coming either from above or from
below) that the initial rate of the respiratory movements in resusci-
tation is the same as after actual section of the upper paths and the
vagi without anemia. This initial rate is apparently dependent upon
a property of the respiratory centre so fundamental that it is estab-
lished as soon as the automatic respiratory discharge emerges in
resuscitation, independently of the length or completeness of the pre-
vious occlusion. The character of the respiration reminds one strongly
of that seen after double vagotomy, the inspiration being prolonged,
but not exhibiting the exceedingly long spasmodic contractions of the

diaphragm described by Marckwald. There is considerable disagree-.

ment among investigators as to the effect upon the respiratory
movements of division of the higher paths without section of the
vagi. Nikolaides! gives the literature. As a rule, anatomical section
of the upper paths, when the respiratory centre has not been isolated
by anzemia, causes a diminution in the respiratory rate, which is fur-
ther diminished by subsequent section of both vagi, usually to be-
tween 4 and 5 a minute. Occasionally the upper paths do not seem
to be exerting any influence on the respiratory movements, and then

' NIKOLAIDES: Archiv fiir Physiologie, 1905, p. 466.

Resuscitation of the Respiratory Nervous Mechanism. 63

section of the vagi alone brings down the rate at once to the mini-
mum. Under experimental conditions it may in rare cases be ob-
served that division of the upper paths, p/ws one vagus, slows the
respiration to the minimum rate, section of the second vagus pro-
ducing no effect on it.

It may be noted that all the characteristic reflex respiratory effects
of stimulation of the vagus or brachial nerves are obtained through
the bulbar centre after anatomical separation of the higher paths.
The following extracts from the protocols of Experiments 1 and 13
illustrate some of these points :

Experiment 1.— Adult male cat. Ether. Tracheotomy. Central ends of

right vagus and left brachial prepared for stimulation.

2.38 P.M. Stimulated vagus. Complete inhibition of respiration in
expiration. Stimulated brachial; respiration much quickened.

2.39. Occluded head arteries. Trephined skull immediately.

2.45.30. Divided brain just below the tentorium. The cat is gasping
spontaneously. Pupils well dilated.

2.48. No gasps for some time. None later so long as occlusion
lasted.

2.50. Heart beat strong; 188 in a minute.

2.51.30. Stimulated brachial. No respiration.

2.52. Released head arteries.

2.53. Cut left vagus. Nose and tongue pink. Corneal tension in-
creased. Pupils somewhat contracted.

2.57- Stimulated brachial. No respiratory effect.

3.01. First gasp. Stimulated brachial. No effect on respiration, but
movements of hind legs occur.

3.03. Gasps 54 a minute.

3.04. Stimulated brachial. Distinct acceleration of respiration.

3.05. Stimulated vagus. No noticeable effect. Respiration, 7 in a
minute.

3.10. Stopped artificial respiration.

3-11. Stimulated brachial. Marked effect on respiration, especially
increase in depth.

3-12. Stimulated vagus. Inhibition of the deeper gasps, while the
shallower intermediate gasps go on, or are even increased in frequency.

3-13-30. Occluded head arteries again.

3-14.30. Left pupil dilating rapidly.

3-15. Gasps going on regularly ; deep, 12 a minute.

3-20. Last gasp.

3-25. Released arteries.

G. N. Stewart and F. H. Prke.

3.39. Stimulated brachial. No respiratory effect. No respiratory
movements have returned, although occasional movements of the lower
abdominal wall occur, associated with hind-limb movements. These
abdominal movements can also be elicited by striking the hind legs.

3.53. Pupils more slit-like. No corneal reflex.

4.05. Pulling on the tracheal cannula causes strong expiratory move-
ments of the abdominal and intercostal muscles. Stimulation of the
brachial causes contraction of the abdominal muscles.

4.08. Stimulated vagus. No effect. The reflexes from the upper part
of the spinal cord are now good, but no respiratory movements have yet
returned. Sucking air from the trachea does not give the expiratory
phenomenon, which can be got every time the trachea is pulled out by
means of the cannula. When traction on the trachea is continuous, the
contraction of the abdominal and intercostal muscles is prolonged, but
eventually relaxes.

4.20. No change. Asphyxia produced at this time causes general
spasms, but no respiratory movements.

Autopsy. — Section complete, 8 mm. above the calamus scriptorius on
the dorsal surface, but slanting further forward on the ventral surface, to
pass through the pons just below the trigeminal roots.

Experiment 13.—Cat. Ether. Tracheotomy. Prepared and ligated (for

stimulation) central ends of right vagus, and left brachial nerves.

3.27 P.M. Stimulated vagus. Complete inhibition of respiration in.

expiration. Stimulated brachial; great acceleration of respiration.

3-30. Occluded head arteries. Respiration and corneal reflex dis-
appear promptly.

3-32-20. Pupils at maximal dilation. A gasp.

3-33. Gasps.

3-36.30. Divided brain posterior to tentorium. No respiratory move-
ments for some time before this, and none after so long as occlusion
lasted.

3-37-30. Struggles of hind end of animal.

3-40. Released head arteries.

3.44. First gasp; good jaw movements also. Gasps continue. Divis-
ion of the left vagus has no effect.

3.45.30. Stimulated vagus. No effect on gasps, which continue
regularly.

3.46. Stimulated brachial. Effect doubtful.

3-48. Gasps are at rate of 1 in thirteen seconds. :

3-48.30. Stimulated vagus. Gasps are 1 in thirteen seconds, exactly
the same as before stimulation.

3-50. Stimulated brachial. Gasps 1 in seven seconds.

3-51. Respirations are 1 in eight and one-half seconds.

Resuscitation of the Respiratory Nervous Mechanism. 65

3-52. Stimulated vagus. Gasps are 1 in sixteen seconds during
stimulation.

3-53. Respirations 1 in ten seconds.

3-54. Stimulated vagus. Respiration distinctly inhibited.

3-55. Respirations 1 in nine seconds.

3.58.30. Stopped artificial respiration.

4.00. Respirations, 11 to the minute. Good regular gasps, including
jaw movements, although these are somewhat feebler than before.

4.02. Stimulated vagus. Respiration 1 in twenty-eight seconds.

4.03. Stimulated brachial. Distinct inhibition in expiration.

4.05 All respirations include movements of the jaws. Respirations
are slow, with prolonged inspiration, like those seen after double vagotomy.
Spontaneous respiration was well established, and no doubt would have
continued for a considerable time.

4.12. Divided spinal cord. Clamped aorta below diaphragm. Right
pupil well contracted ; left, wide.

4.21. Twitching of whiskers.

4.23. Stimulated vagus. Contraction of diaphragm occurs each time
the vagus is stimulated; also, opening the mouth, protrusion of tongue,
and contraction of cesophagus as in swallowing.

4.25. Twitching of right axilla. Asphyxia does not produce respira-
tory movements, but when artificial respiration is started again, the dia-
phragm contracts independently of the artificial respiration (¢. e. with a
different rhythm), as shown by feeling the diaphragm with the finger.

4.35. Left pupil still wide, right very narrow. Right fore limb gives
reflex contraction on striking it.

4-55. Stopped artificial respiration. Movements of diaphragm begin
after a time and continue regularly. Fair respiration, rate 3 in twenty
seconds, apparently unaffected by stimulation of vagus or brachial. A little
earlier, stimulation of central end of vagus caused contraction of diaphragm,
while stimulation of brachial did not. Verified several times. Contraction
of diaphragm caused by vagus stimulation was not tonic, but a single twitch.
Not due to escape of current, since it could not be obtained later on.

5-25. Spontaneous respirations still occur at intervals, involving dia-
phragm, thorax, and jaw. Stimulation of the vagus now causes contraction
of diaphragm, accompanied by jaw and throat movements. Animal has
lived thirty minutes without artificial respiration. Weak respiratory move-
ments of diaphragm and jaw alternate with strong, one strong respiration
and one weak occurring in fifteen seconds. Sometimes there are several
weak respirations to each strong one, but the rate of the strong movements
is practically constant, about 4 in the minute.

5-31. Condition much the same. Stimulation of the central end of
the vagus still causes respiratory movements; spontaneous movements
also occur. Stopped experiment.

66 G. N. Stewart and F. H. Ptrke.

Autopsy. — Section of spinal cord was at level of the seventh cervical
pair of nerves. Complete. Upper section was through pons above roots
of trigeminus. Section was complete except for a thin layer on the
dorsal surface.

The effects of stimulation of a nerve like the brachial on the respi-
ratory muscles through the reflex centres of the cord are to be
sharply discriminated from the effect produced on the bulbar respi-
ratory centre. We have seen abundant instances of reflex excitation
of the intercostal muscles and the muscles of the abdominal wall
through the cord at a time when there was no evidence whatever
that the bulbar respiratory centre had resumed its functions. For
example, stimulation of the central end of the brachial very frequently
causes a strong and long-continued contraction of these muscles when
no spontaneous respirations have appeared, and even in cases where
respiration never returns. The diaphragm does not participate in
these movements, and their character, although undoubtedly they
may be capable of altering the intrathoracic pressure, and therefore
of aiding pulmonary ventilation, is totally different from that of the
true respiratory movements. One characteristic is their tendency in
the earlier stages of resuscitation to involve only the lower abdom-
inal muscles and the hind legs, movements of which are also easily
elicited at this time by stimulation of the sciatic. It is possible that
the strong and long-continued expiration which we have frequently
observed, for example, on making steady traction on the tracheal
cannula is a reflex effect of this nature. It can be obtained under
conditions which preclude the activity of the bulbar respiratory
centre.

As a rare phenomenon, an apparently reflex contraction of the dia-
phragm may be caused by stimulation of the central end of the vagus
at a time when no spontaneous respirations are occurring. The con-
traction consists of a single twitch at the beginning of stimulation °
involving both halves of the muscle, and is not due to escape of cur-
rent on to the phrenic (see Experiment 13). Here respiration had
returned after section through the pons and occlusion.

II. THE MECHANISM OF SPINAL SHOCK.

We have said that shock must be avoided in making the upper sec-
tion. We can entirely confirm the statement of Asher and Liischer,!

? ASHER and Liscuer: Zeitschrift fiir Biologie, 1899, xxxvili, p. 499.

Resuscitation of the Respiratory Nervous Mechanism. 67

who used Kronecker’s method of paraffin emboli, that elimination by
anzemia of the brain and cervical cord does not cause shock in the
portion of the cord whose circulation is still maintained, and can
add that complete section of the encephalon at the level mentioned,
after a certain duration of the anemia, is also innocuous in this re-
gard. The same is true at an early period in resuscitation. Where
the section is made so early in occlusion as to be accompanied by
visible signs of stimulation it produces shock. The experiments
already quoted illustrate the avoidance of shock by dividing the brain
stem six and one-half minutes after the beginning of occlusion.. In
Experiment 4 the section was made four minutes after the begin-
ning of occlusion; in Experiment 3, one and one-half minutes from
beginning of occlusion, while Experiments 5 and 6 show the result
of a division at the very beginning of occlusion, and Experiment 10
the effect of section before occlusion. The condensed protocols of
these experiments follow:

Experiment 8. — Adult cat. Ether. Tracheotomy. Prepared central ends of
right vagus and left brachial nerves. ;
3-25 P. M. Stimulated vagus. Respiration inhibited in expiration.

3.30.30. Occluded head arteries. Respiration stopped in about
forty-five seconds.

3-32. Pupils well dilated. Some secondary gasps. Divided brain.

3-41.30. Released arteries. Hind end very active. No respiration
ever returned, either spontaneously or in response to stimulation of nerves,
4.10. Produced asphyxia. No contraction of upper limbs or dia-
phragm even in two minutes. Pupils remain narrowly contracted, but, on
resuming artificial respiration, they dilate partially and then soon become
slit-like again.

Experiment terminated.

Autopsy. — Section was complete; passed through the pons above
trigeminus roots, and dorsally through posterior part of the posterior
corpora quadrigemina.

Experiment 4.— Adult cat. Ether. Tracheotomy. Prepared for stimulation
central ends of right vagus and left brachial. Divided both vagi.

10.38 A.M. Stimulated brachial. Quickening of respiration. Stimu-
lated vagus. Quickening of respiration, but gasps are shallower.

10.41. Occluded head arteries.

10.45. Divided brain.

10.46. Spontaneous respiration going on, not spasmodic. Pupils
maximal. No corneal reflex. Intra-ocular tension low.

10.49. Last gasp.

68 G. N. Stewart and F. H. Pike.

10.51. Released the head arteries.

11.24. First gasp.

11.30. Deep gasps going on. Stimulated brachial. Strong tonic
contraction of the abdominal muscles follows. Respirations 4 in fifty-five
seconds without stimulation, and exactly the same during stimulation.

11.32. Stimulated vagus. ‘There is exactly the-same rate of respira-
tion, 4 in fifty-five seconds, as before stimulation. Strong tonic contrac-
tion of abdominal muscles.

11.39. Now got complete inhibition of respiration. After stopping
stimulation, respiration returned, but was shallower than before.

I 1.40.30. Spontaneous respiration ceased.

No true respiratory movements occurred after this, either spontaneously
or in response to stimulation of nerves.

11.55. Produced asphyxia. No true respiratory movements. Spasms
of hind end.

Autopsy. — Section complete. Situated just above tentorium. Passed
above the upper edge of the pons, nearly 5 mm. above the trigeminus roots.

Experiment 5. — Cat. Ether. Tracheotomy. Prepared central ends of right
vagus and left brachial nerves.

2.43 P.M. Stimulated vagus. Complete inhibition of respiration in
expiration. Stimulated brachial; great increase in respiration. ~

2.45: 30. Occluded head arteries. Divided brain and spinal cord
immediately, the bone having been removed before occlusion.

2.53. Released arteries.

3-27. Stimulated vagus. Good retraction of nictitating membrane,
and bulging of eye, but no dilation of pupil unless stimulation is very
strong, when a slight dilation occurs, increased somewhat by increasing the
strength of stimulation. Confirmed. No respiratory effect. No respira-
tion has returned as yet.

Slight reflex movements of hind limbs on striking them. Also, on
striking hind leg, get fairly good contractions of abdominal muscles on
the same side, extending to the median line but not crossing. The reflex
contraction of the hind limbs crosses slightly.

3-51. Clamped abdominal aorta below the diaphragm. The head
arteries are now much better filled. Intra-ocular pressure increases
almost at once. The nictitating membrane, which had covered the eyes,
retreated, and the eyes, which had been strongly rotated inward and
downward, returned to their normal position.

4.12. Stimulated vagus; slight dilation of pupil; no respiratory effect.
No respiratory movements ever returned. Stopped experiment.

Autopsy. — The lesion of the cord involved the dorsal half, but only a
little of the ventral half, at level of the sixth cervical pair. Section through
the brain was complete, through pons just above trigeminus roots.

Resuscitation of the Respiratory Nervous Mechanism. 69

Experiment 6.— Cat. Ether. Tracheotomy. Central ends of right vagus
and left brachial prepared for stimulation.

10.45 A.M. Stimulated brachial. Great acceleration of respiration.
Stimulated vagus ; stronger expiration.

10.47. Occluded head arteries. Divided brain immediately, the skull
having been trephined before occlusion.

10.53. Released arteries.

10.54. Divided left vagus.

12.12. Stimulated right vagus. No effect except on eye. Stimulated
brachial. Great contractions of lower end of animal; no true respiratory
effect.

12.45. Reflex movements of right fore limb are got on pulling or strik-
ing it, or pinching toes. Strong spasms of almost the whole animal occur
from time to time, with opisthotonus, the head being strongly drawn back.
Respiration did not return. Experiment discontinued. Asphyxia caused
a strong general spasm.

Autopsy. — Section complete, through posterior part of anterior corpora
quadrigemina.

Apart from the problem with which they were more particularly
connected, these observations have a certain interest in relation to
the mechanism of spinal shock. Sherrington, in discussing the
causation of spinal shock, sums up as follows: “I think, therefore,
that spinal shock is neither due to irritation by trauma nor in the
main a phenomenon of inhibition. The rupture of certain aborally
conducting paths appears to induce it. Which these paths exactly
are is matter for research.” It is difficult to reconcile our results with
this conclusion. For undoubtedly these hypothetical aboral paths are
physiologically ruptured by cerebral anemia of a certain duration,
and they are anatomically ruptured when at this stage of anemia the
bulb or pons is divided, yet shock is not produced in the part of the
spinal cord below the anemic region. If, however, an earlier stage
in the anzmia be chosen for the anatomical lesion, or if this be pro-
‘ duced prior to occlusion of the head vessels, the symptoms of bulbar
and spinal shock promptly appear. If we admit the validity of Sher-
rington’s argument, that the negative result of a second spinal trans-
section, made after the shock produced by the first has passed off, is
inconsistent with the idea that the stimulation caused by the trauma
is the effective factor, then we would seem shut up to the conclusion

1 SHERRINGTON: The integrative action of the nervous system, New York,
1906, p. 246.

70 G. N. Stewart and F. Ht. Prke.

that it is neither the rupture of conducting paths by itself nor the
traumatic stimulation by itself, but the combination of the two, which
determines the production of shock.

So far as the respiratory centre is concerned, we have observed
that section of the upper paths with the knife, when their conduc-
tivity and excitability were still intact or but slightly depressed by a
very short total occlusion or by a longer imperfect occlusion, acted
prejudicially on the power of resistance of the centre to a continu-
ance of the anemia for a period which under normal circumstances
would have left it capable of prompt and perfect resuscitation. An
example is seen in Experiment 10:

Experiment 10.— Cat. Ether. Tracheotomy.

3.29 P. M._ Respiration g in fifteen seconds.

3.30. Divided brain. ‘The respiration goes on for a little, somewhat
slower, then stops. Started artificial respiration. Some spontaneous
respiration soon came back.

3.40. Fair spontaneous respiration.

4.00. Respiration 7 in sixty seconds, deep and prolonged inspiration,
just like the respiration following double vagotomy.

4.17. Respiration 2 in sixteen and one-half seconds. Prepared cen-
tral ends of right vagus and left brachial for stimulation.

4.18. Stimulated vagus. Strong inhibition of respiration in expiration.
Stimulated brachial. Strong inhibition in expiration with shallow gasps
toward end of stimulation. Two swallowing movements occur.

4.20. ‘Occluded head arteries.

4.25. Released head arteries.

4.36. First respiration occurs (including movement of diaphragm).
Only one gasp seen. ‘Tongue is protruded occasionally.

4.46. Produced asphyxia. No respiratory movements.

4.54. Reflex can be got in shoulder by striking fore limb of same side.
No respiration has returned.

4.58. Stimulation (pinching) of right ear causes strong scratch move-
ments of right hind limb.. No movements of fore limb occur on pinching
ear, although good contractions of the fore limb can be obtained on
striking it.

5.08. Stimulated vagus. No effect except on eye. The pupil be-
comes maximal, with bulging of the eye and retraction of the nictitating
membrane.

5-09. Stimulated brachial. Strong contractions of hind end of animal,
including abdominal muscles.

5.10. Cut left vagus (in case absence of respiratory movements should
be due to apncea). No respiratory movements occur. Reflex contrac-

Resuscitation of the Respiratory Nervous Mechanism. 71

tions of abdomen and thorax can be got from any part of the abdomen or
thorax on striking that part.

5-55- Stopped artificial respiration. Struggles occurred, but no true
respiration. Discontinued experiment.

Autopsy. — Section was between the anterior and posterior corpora
quadrigemina on the right side, but spared the greater part of the left
posterior corpus quadrigeminum, slanting backwards from right to left.
Ventrally, the section passed through the pons above the level of the
trigeminus roots. The connection of the left corpus quadrigeminum with
the bulb was not completely destroyed.

In our experience, section of the cord as lowas the upper and mid-
dorsal region at a time in the resuscitation when the bulbar centres
are just beginning to function is always fatal, the bulbar discharge
ceasing in a very short time. It is not, perhaps, surprising that this
should be the case where the blood pressure is not maintained after
section by ligation of the thoracic or abdominal aorta. But in one
experiment (13) section of the cord at the seventh cervical seg-
ment, although accompanied by occlusion of the aorta just below the
diaphragm, was followed by immediate cessation of the respiratory
movements, which had been thoroughly re-established after section of
the pons and occlusion. It was not the cutting off of (hypothetical)
impulses from below, normally indispensable to the activity of the
respiratory centre, which was responsible for the stoppage of respira-
tion, since, more than forty minutes later, regular diaphragmatic
respiration was established upon stopping the artificial respiration.
It would seem that, under the conditions of the experiment, the still
more or less crippled respiratory centre succumbed temporarily to
shock, spreading in this instance up the cord from the lesion. It is
generally stated that spinal shock involves only the nervous apparatus
peripheral to the lesion. But it may very well be that even under
normal circumstances the influence which we call shock is propagated
in both directions, although less readily centripetally. Normal, un-
deteriorated centres may not show any obvious ill effect of the rela-
tively weak centripetal shock influence, while centres previously
- crippled by anemia may do so; just as the normal spinal cord is
rendered more excitable by strychnia, whereas a portion of the cord
recovering from anzmia is, at a certain stage in its resuscitation,
rather depressed than excited.!

1 STEWART, GUTHRIE, BuRNS, and PIKE: Journal of experimental medicine,
1906, viii, p. 309.

72 G. N. Stewart and F. H. Ptke.

In Experiment 10, although respiration did not return, good re-
flexes could be elicited from the previously anemic cervical cord.
The conduction of impulses by this part of the cord and by the bulb
was early established (twelve minutes after release). Stimulation
of the left brachial at this time, although it caused no movements of
the opposite fore limb, caused strong contractions of the hind limb
and movements of the tongue. Thirty-three minutes after release,
pinching the right ear was followed by strong scratch movements of
the right, and feeble movements of the left hind limb. The right
fore limb did not contract at all, although it gave good movements on
being struck. Asher and Arnold! observed that, at a certain stage
in total anaemia of the spinal cord below the upper thoracic region,
excitation of the central end of the sciatic caused movements of the
fore limbs without exciting intervening spinal segments. In this
case the impulses were possibly carried up the long afferent paths in
the posterior column without passing over a synapse in the anzmic
area. In our observations the impulses must have passed over at
least one synapse in the previously anzemic region before reaching
the long proprio-spinal fibres in the lateral column described by Sher-
rington ? as the path through which the scratch reflex is liberated in
the spinal dog when the skin of the shoulder is stimulated, if indeed
this is the path concerned under the conditions of our experiments.

In Experiment 9, where the brain was divided in a cat, above the
tentorium, without occlusion, and respiration stopped permanently
after a few breaths, tickling either ear (thirty-five minutes after section
of the brain) caused strong scratch movements of the corresponding
hind limb, not involving the contralateral foot. No movements of the
fore limbs were caused during the passage of these impulses down the
cervical cord, whose reflex functions were indeed still almost sub-
merged by the shock, as only feeble movements of the fore limb could
be elicited by stimulating it mechanically.

We® have previously described the severe tonic convulsions which
appear in the period of recovery from cerebral anemia. It is of
interest to note in this connection that similar spasms, involving
the muscles of the trunk limbs, neck, and jaws, may occur in the
period of recovery from cerebral anemia combined with transsection
of the brain stem above the medulla oblongata. The same hyper-

* ASHER and ARNOLD: Zeitschrift fiir Biologie, 1900, xl, pp. 271-287.

? SHERRINGTON: Loc. cit., p. 52.

* STEWART and PIKE: Loc. cét., Journal of experimental medicine, p. 307.

ee

ee yee ee a

Resuscitation of the Respiratory Nervous Mechanism. 73

excitable condition of the spinal reflex mechanisms, simulating that
seen in strychnine poisoning, previously described, is met with also
in animals whose brain has been divided. Whatever share the higher
parts of the brain may take in maintaining the hyperexcitability of
the lower parts of the central nervous system at a particular stage in
the resuscitation, they are not indispensable for the condition in
animals subjected to anemia combined with section of the higher
paths, although these centres may add to the effects produced by
the lower centres in animals subjected to anzmia but whose brain
has not been divided.

An observation (in Experiment 5) on the effect of stimulation of
the vago-sympathetic on the pupil during resuscitation, after an occlu-
sion of the head arteries and division of the brain stem and cervical
spinal cord, is worthy of mention. While retraction of the nictitating
membrane and protrusion of the eyeball were readily obtained, and
to the maximum degree, no dilation whatever of the pupil was caused
unless the current was very strong, when it dilated slightly. The di-
latation was increased somewhat when the stimulation was strength-
ered still more, but nothing like maximal dilatation could be obtained
with any strength. This is an instance in which the statement of
Mulert,! that the “all or nothing” law holds approximately for the
dilator pupilla, could not be verified. It is possible that the pupillo-
dilator fibres somewhere on their course had suffered relatively more
from the anzemia than the other groups.

1 MULERT: Archiv fiir die gesammte Physiologie, 1894, lv, p. 550.

THE ACUTE EFFECTS OF GASTRIC AND PERITONEAL
CAUTERIZATION AND IRRITATION ON THE BLOOD
PRESSURE AND RESPIRATION.

By TORALD SOLLMANN, E. D. BROWN, anp W. W. WILLIAMS.

[From the Pharmacological Laboratory of the Medical Department of Western Reserve
University, Cleveland, Ohio.]

|e the course of a series of experiments on the treatment of carbolic

acid poisoning, it was observed that the administration of concen- .

trated phenol by a stomach tube, to dogs, produces a very prompt,
considerable, and lasting fall of blood pressure. It became necessary
for us to decide whether this fall is caused by a direct action of the
phenol, following its absorption; or whether it is a condition of reflex
shock, resulting from the gastric irritation. As one means to this
end, we experimented with other corrosives and irritants, and found,
to our surprise, that the acute effects of gastric irritation are abso-
lutely negative, at least with anesthetized dogs.

We employed violent corrosives, and strong and mild irritants, all
in large doses, administered either by the stomach tube, or painted
on the inner and outer surface of the exposed stomach, and on the
denuded submucosa, and on the parietal peritoneum. We also per-
forated the stomach with actual cautery. A number of these meas-
ures were applied successively to the same animal. When given by
the stomach tube, the simple irritants were made to precede the
corrosives; with the local applications, a fresh surface was taken for
each application.

The experiments were made on dogs, anzsthetized ‘with morphin
and ether and having the carotid connected with a mercury ma-
nometer, The anzsthesia was made light in the case of the milder
irritants, such as mustard; otherwise it was sufficiently deep to ex-
clude all expressions of pain, although the lid reflex was often
preserved. The publication of the protocols and tracings seems
superfluous, since they are so largely negative.

74

Acute Effects of Gastric and Peritoneal Cauterization.

75

The results of the experiments are classified and summarized in

Table I.

TABLE I.

1. Administration by stomach tube.

Substance No. of Effect on
administered. tracing. blood
pressure.
Water at about 20° C.:
Stomach filled
«evacuated
“ filled 49 No effect.
“ evacuated
500 c.c. in 3 min. 53 No effect.
250 c.c. evacuated
in 5 min. 56 conse
250 c.c. evacuated
Hot water:

fiGeratsOc C;
250 ce at 65°:C.

51 Slight rise.

|
|
l
|

CLASSIFICATION OF EXPERIMENTS AND SUMMARIZED RESULTS.

On respi-

ration. Remarks.

There is a gradual rise of pres-
sure throughout the proce-
dure, but we are inclined

\ to attribute this to the with-
drawal of the ether.

No effect.

The blood pressure rises slowly
by 10 mm., but this is prob-
| ably due to lessening of ether.

No effect at first,

in 14 min. 57 No effect. later decrease in
235 c.c. evacuated excursions.
200 c.c. at 60° C. in
: Decrease of
ue 37 = e2:Cursions.
Evacuated
Cold water:
300 c.c.at 10° C.
in 3 min. 57 Noeffect. No effect.
Evacuated
NaOH:
10 per cent, 12 c.c. 32 Insignificant fall ....
aU) OS a with recovery.
Be Ae BR. 32 =Noeffect. ; Vagi divided.
ee i Loy. 49 Slight vagus stimulation and
slight irregularity of blood
H.SO,: pressure.
25 per cent, 12 c.c.
concentrated, 12c.c. 46 No effect.
e: Sue 51 Slight fall Excursions slightly
(15 mm.). increased.
HCI:
Concentrated,12c.c. 46 No effect.
“ 10 “ fe ep
“ 25 “6 j 50
Phenol :
95 percent, 12 c.c. 47 No effect.
meee Sh res 47 Comte Vagi divided.

76 Torald Sollmann, E. D. Brown, and W. W. Withams.

TABLE I (continued).

Effect on :
Substance No. of blood On respi- Remarks.
administered. tracing. pressure. ration.
Phenol (continzed ) : .

95 per cent 12 c.c. 48 No effect.
ee Some | Sd

i Ee ae pia Vagi divided.
95 “ “ 10 iis
Formaldehyde :
13 per cent, 15 c.c. 49 Noeffect? Sooo There is a rise due to with-

drawal of ether.
Slight irregularity.

40 “« « ey 49 | vase stimulation ?

AO met em Ola 50 No effect. sooC Vagi divided.
tO eo eel Orcs 51 mf oe NOlettect.
Mustard spirits:
6 C.c. 47 No effect.
9 cc. ' 49 ~~ Slight irregularity ?
6 C.c. 51 No effect. No effect.

2. Local applications.

(2) Tip of tongue.
HCl: Concentrated, drop 49 No effect.

(6) Gastric mucosa.

REAhORGOO 34 Rio roe Nore acer Applied twice, once to perfo-
58 ration.
NaOH, 40 per cent 58 se ce “ + Local change manifest.
#250), ‘concentrated | Ke Ue Sk ee Immediate corrosion
* 58 :
HNO, ¥ 54 cs a is id Turns yellow at once.
Phenol, 95 per cent oy)
ign Oona ke 55 uo oH ‘¢ Blanches instantly.
fs 95 “c “ 58
F 53 ; P ;
ormaldehyde, 40 per cent 58 : ‘ : :
; in] 53 “ ce “ “
Mustard spirits 58
Peppermint spirits. 58 s 4 ce i

Alcohol, 95 per cent
Acetic acid, 5 per cent

(c) Gastric submucosa.

Melted sealing-wax 52 No effect.
Red-hot iron 54g) OX s No effect.
Formaldehyde, 40 per cent 53 Slight rise. Stimulation.
xs 40 “ ““ 53 Noeffect. | No effect.
Mustard spirits 54 Very slight Slight decrease.
fall.
H,SO,, Conc. 54 No effect. No effect.

Acute Effects of Gastric and Peritoneal Cauterization. 77

TABLE I (continued).

Effect on
Substance No. of blood On respi- Remarks.
administered. tracing. pressure. ration.
(d@) Gastric serosa.
Red-hot iron, twice a
NaOH, 40 per cent 58 No effect. No effect. Local change manifest.
“ 10 td “ce 54
H,SO,, concentrated 51 Slight rise? probably no effect.
re ie 7 No effect. No effect. Strong contraction of muscle.
HNO; i 54 es a) ee Turns yellow.
f Probably The blood pressure rises, ,but
Pace tlehyde, 40 per cent. 51 no effect. this is doubtless asphyxial.
ei ate OS Slight rise. Stimulation.
“s 40 ce “ 54
4 40 «© « 58 No effect. No effect.
Mustard spirits 53 Veryslightrise. Very slight increase.
a. “ 58 No effect. No effect.

Alcohol, 95 per cent

Peppermint spirits |
Acetic acid, 5 per cent

(¢) Parietal peritoneum.

Red-hot iron

NaOH, 40 per cent
“ec 10 “e “cc

H,SO,, concentrated

HNO, ‘¢

“ e
Phenol, 95 per cent
Formaldehyde, 40 per cent
Mustard spirits

“ “

Acetic acid, 5 per cent

55}

54
58
55
54

No effect. No effect.
Slight mo-
mentary fall.
54)
58
55
55

Momentary arrest.

No effect. No effect.
54
54
58

58

Arranging the experiments into three groups, namely :

Group I.

Cautery.
Group II.
Group III.

Corrosives. — NaOH, H,SO,, HNO;, HCl, Phenol, Actual

Strong Irritants. — Formalin, Spir. Mustard, Hot Water.
Mild Irritants. — Alcohol, Spir. Peppermint, 5 per cent Acetic

Acid, Cold Water, Simple Distention and Evacuation.

We obtain the following totals:

78 Torald Sollmann, £. D. Brown, and W. W. Williams.

Blood pressure. Respiration.

Slight | Slight
de- in-
crease. | crease.

No of No | Slight | Slight | No. of
exp’s. | effect. | fall. rise. | exp’s.

Corrosives
Strong irritants .

Mild irritants

Total

It will be seen that in about nine tenths of the experiments no
effect whatsoever was obtainable, either on the blood pressure or on
respiration. In the remaining tenth the effect was insignificantly
small, and as often in the direction of an increase as of a decrease.
It may be interesting to remark, incidently, that very considerable
distention of the stomach with water, or sudden evacuation of the
fluid, is also without effect.

In the above summaries each application is considered as a sepa-
rate experiment, together with the very short after-period elapsing
before the application of another irritant. This after-period was
generally less than five minutes ; —a very short time, it is true, but
ample for any ordinary reflex effects, such as might affect the general -
blood pressure or the respiration, or produce shock. In fact, how-
ever, a number of these experiments were made on each animal. Even
should we grant the possibility of a delayed reflex action, this would
therefore also be indicated on the curves; for the experiments were
continued for half an hour to an hour, and during this entire period
there was no sudden change. Furthermore, the successive application
of the irritants to the same animals would be an ideal method of em-
phasizing any cumulative action which might be assumed. A brief
enumeration of the procedures on each animal, together with the
initial and final blood pressure, is of interest in this connection, and
will also serve to emphasize the absolutely reckless severity (as we
should have supposed a fvior7) of irritation to which the stomach
may be subjected.

Acute Effects of Gastric and Peritoneal Cauterization. 79

IRRITANTS APPLIED TO EACH ANIMAL.

Dog No. 12. Procedures: By stomach tube: 12 c.c. of to percent NaOH ;
46 c.c. of 40 per cent NaOH; Vagotomy. Blood pressure at the begin-
ning, 132 mm.; at end, 80mm. Duration of experiment, one hour.

Dog No.18. Procedures: By stomach tube: 12 c.c. of 25 per cent H,SO,;
12 c.c. of conc. H2SO,; 6 c.c. of Spts. Mustard; 12 c.c. of conc. HCl;
12 c.c. of conc. Phenol; Vagotomy ; 12 c.c. of conc. Phenol ; Distention
of stomach ; 24 c.c. of conc. Phenol. Blood pressure at beginning, 115
mm.; atend, 75 mm. Duration of experiment, one hour.

Dog No. 19. Procedures: By stomach tube: g c.c. of Sp. Mustard; 5 c.c.
of Formalin; 10. c.c. of go per cent NaOH. Gastric lavage; 5 c.c. of
Formalin ; 35 c.c. of conc. HCl; Vagotomy; 10 c.c. of Formalin; 20
c.c. of Phenol. Blood pressure at beginning, 150 mm.; at end, 140
mm. Duration of experiment, thirty-five minutes.

Dog No. 20. Procedures: By stomach tube: 6 c.c. of Sp. Mustard; 10 c.c.
of Formalin; 8 c.c. of conc. H2SO,; Hot water (70°C.). Abdomen
opened ; Formalin, H2SO,, and melted sealing-wax on outer coat of
stomach. Stomach cut open, mucosa partly removed; melted sealing-
wax on submucosa. Blood pressure at beginning, 105 mm. ; at end, 77
mm. Duration of experiment, twenty-six minutes.

Dog No. 21. Procedures: Abdomen and stomach cut open. Local applica-
tions to gastric mucosa, submucosa, and serosa and to parietal peritoneum,
of Formalin; Sp. Mustard ; Red-hot iron to perforation ; conc. H2SQ, ;
Formalin; 1o per cent NaOH; conc. Phenol; conc. HNO;. Blood
pressure at beginning, 132 mm.; at end, 78 mm. Duration of experi-
ment, one hour.

Dog No. 22. Procedures: Distention of stomach with hot (65° C.) and cold
(10° C.) water. Abdomen and stomach cut open. Local application to
gastric mucosa and serosa and parietal peritoneum of 95 per cent Alcohol ;
Sp. Peppermint; Sp. Mustard ; 5 percent Acetic Acid ; Formalin; conc.
He2SO,; 40 per cent NaOH; conc. Phenol; Actual cautery. Blood
pressure at beginning, 90 mm.; at end, 45 mm. Duration of experiment,
one hour. Sciatic stimulation at end causes rise of pressure and increase
of respiration.

Dog No. Ig is especially interesting. It would be difficult to imag-
ine a more formidable array of corrosives than those which were poured
into the stomach of this animal during the course of half an hour;
and notwithstanding, the experiment was concluded with a blood
pressure of 140 mm., only 10 mm. lower than the initial pressure.
The other five animals showed a very gradual fall, the final pressure

89. 6 Zorald Sollmann, E. D. Brown, and W. W. Williams.

ranging from 45 to 80 mm., with an average of 71 mm.; representing
a fall of 28 to 54 mm., with an average of 33 mm. This, however, is
still far removed from a condition of shock; indeed, in the animal
with the lowest blood pressure (dog No. 22) the vaso-motor and res-
piratory centres responded to sciatic stimulation in a perfectly nor-
mal manner, at the conclusion of the experiment. In view of the
behavior of dog No. 19, the gradual fall of pressure is, in our opinion,
fairly attributable to the exposure, the operation of opening the abdo-
men, and especially to the duration and the generally increasing
depth of the anzsthesia, and not to the irritants. In brief, there-
fore, it would seem that the most violent corrosion and irritation of
the stomach or parietal peritoneum, in the anesthetized dog, pro-
duces no acute reflex effects. It is, of course, very probable that the
later effects would be more serious. This is a portion of the subject
which our experiments do not touch. It is self-evident that the
measures to which we subjected these animals would inevitably have
led eventually to intense and probably fatal inflammatory reactions.

Experiments now under way tend to show that this insensibility to
strong corrosives is not shared by the mucous membrane of the
mouth and by the larynx and trachea. These seem to respond by
much more violent reflexes to all kinds of irritation. We mention
this fact at the present time, so that our negative results from irri-
tants administered by the stomach may not be misconstrued as
applying to irritants taken by mouth.

CONCLUSIONS.

Corrosion or violent or mild irritation of the gastric mucosa, sub-
mucosa, or serosa, or of the parietal peritoneum, has generally no
acute reflex effect upon the blood pressure or respiration in anesthe-
tized dogs. In the few cases in which a response was obtained this
was slight, and about as often in the direction of an increase as of a
decrease.

A succession of violent irritant measures applied to these struc-
tures is also without definite effect on circulation and respiration, the
observation extending over an hour.

CHEMICAL STUDIES ON GROWTH.—I. THE INVERTING
PNZYMES OF THE ALIMENTARY JRACT,. ESPE-—
Sisley IN THE EMBRYO,

Be LAFAYETTE B: MENDEL anp PHILIP H. MITCHELL.
[From the Sheffield Laboratory of Physiological Chemistry, Yale University.]

N. the ordinary study of nutrition the phenomena of growth re-
ceive only incidental consideration, since developmental changes
are not usually conspicuous features in the activities investigated.
From a chemical standpoint growth is distinguished by a preponder-
ance of synthetic, or anabolic, reactions leading to the replacement of
disintegrated parts or the accumulation of new tissue constituents.
There are periods in the life of every organism which are especially
characterized by processes of growth. Any comprehensive appre-
ciation of nutrition in its broadest sense, therefore, calls for adequate
recognition of the nature of the metabolism which distinguishes
them.

A knowledge of the chemico-physiological peculiarities of growing
tissues and organisms finds direct application to the practical prob-
lems of nutrition. What are the materials at the expense of which
growing structures build? What is the equipment of a developing
organism for the special metabolism which it experiences? What
chemical and physiological adaptations contribute to a relative pre-
ponderance of constructive activities? To what extent does growth
have unusual biochemical facilities at its disposal? Questions like
these remain for the most part unanswered in the familiar literature
on nutrition. They demand special researches which appear timely
at present, in view of the profound interest awakened among physiol-
ogists in the study of metabolism. It is this which has encouraged
the present series of experiments, the expenses of which have been
met largely by a grant from the Carnegie Institution of Washington.

The data, most of which have been obtained through a study of
81

82 Lafayette B. Mendel and Philip H. Mitchell.

embryonic material, will be presented for publication as contributions
to the solution of specific problems.}

It is well established that the familiar disaccharides, maltose,
sucrose and lactose, can be inverted to six-carbon sugars before or
during the act of absorption from the alimentary tract. It is, further,
understood that the hydrolysis is accomplished by enzymes which act
specifically upon the individual disaccharides. Whether such agencies
—maltase, sucrase, lactase —are always present in the intestinal
secretions as well as in the secretory cells, and act within the lumen
of the digestive tube upon the sugars; or whether the hydrolytic
reactions occur mostly within the cells of the lining mucosa, is by no
means so clearly demonstrated. The fact that some of the disaccha-
rides referred to are rarely, if ever, introduced into the intestinal tract
of certain species, gives a new interest to the significance. of the
occurrence of the enzymes which invert them, z. ¢., to the question as
to whether specific enzymes are developed only in response to the
individual requirement. The problems regarding the adaptation of
secretory glands are likewise here suggested. We shall see that the
specific inverting enzymes furnished in the same species may vary
at different periods in its life history, while the distribution also
varies in the different groups of animals. It is too early, however,
to offer an adequate explanation, not to say a teleological theory, for
these facts.

The sucrose-inverting properties of the intestinal mucosa were the
first of these enzyme activities to be pointed out, in 1871. We shall
refer briefly to some of the literature indicating the distribution of
the individual enzymes. The bibliography in the appendix includes
many of the more significant contributions.

Maltase.— The most universal of the sugar-inverting enzymes,
maltase, was first recognized in animals in the mucosa of the small
intestine of the pig by Brown and Heron. Bourquelot soon there-
after clearly distinguished between maltase and sucrase, and showed
that the activity of intestinal extracts is lost by filtration through
- porous clay. Since then maltase has been shown to have a very wide

' Brief notes regarding some of the studies have already been published. Cf
MENDEL: “ The alimentary enzymes of the embryo,” This journal, 1906, xv, p. xiii;
‘“Chemical studies on growth,” British medical journal, 1906, p. 1787 (December) ;
‘““Embryo-chemical studies — the purine metabolism of the embryo,” This journal,
1907, xix, p. xvii; Journal of biological chemistry, 1907, iii, p. xxxiv.

Chemical Studies on Growth. 83

distribution in many tissues and secretions of animals as well as
their blood serum. In extracts of the stomach alone was it missed
by Fischer and Niebel. Maltase has been found in the intestinal
juice of man (by Tubby and Manning, and others) and animals (by
Rohmann, Mendel, and others); although, according to Rohmann
and Nagano, the mucosa extracts are far more active than the secre-
tion itself. Bierry has reported the occurrence of maltase in the
embryonic intestines of sheep and cattle.

Sucrase.—It is interesting to note that Paschutin, the earliest
investigator of the sucrose-inverting power of the intestinal mucosa,
reported distinct differences in different species of animals. Whereas |
extracts prepared from the dog, pig, rabbit, and rat were active,
those from the calf and sheep were not. Precisely similar observa-
tions regarding the latter animals were made much later by Fischer
and Niebel.! The occurrence of sucrase as described above has been
verified by the work of Brown and Heron, Bourquelot, Grtnert,
Rodhmann, Mendel, and others, and extended to include the intestine
of birds, the horse, both young and old (Fischer and Niebel), and
the newly born infant (Miura). It is present in the intestinal se-
cretion itself in the dog (R6hmann, Bastianelli, Mendel). In the
intestinal extracts of the embryos of cattle and sheep sucrase has
been missed by Bierry, and its occurrence regarded as doubtful by
Kriiger. It is not ordinarily present in the blood and tissues as are
maltase and amylase.

Lactase. — With reference to the distribution of lactase the state-
ments are less concordant; and about this enzyme much of the
recent controvérsial literature regarding adaptation in glands has
turned. Observations of R6hmann and Lappe indicated its absence
in the mucosa of the small intestine of the cow, although it is present
in that of the calf; and in the case of the dog they likewise found
the enzyme present in greater abundance in young animals. Similar
distinctions between young and old individuals of the same species
have been claimed for the rabbit by Orbdn and Weinland, for the
guinea pig and cat by Plimmer. In birds and amphibia lactase has
been uniformly missed. A survey of the numerous data now avail-
able in the literature which is compiled in the appendix indicates a
preponderance of lactase in younger animals wherever comparative
investigations are at hand. In the succus entericus itself lactase has

1 Positive results with the succus entericus of sheep have been reported by
PREGL. The experiments were conducted without adequate antiseptic precautions.

84 Lafayette B. Mendel and Philip H. Mitchell.

always been missed (Pregl, Mendel, Rohmann and Nagano, Frouin
and Porcher). Its presence in the embryos of various species, as
well as in the faeces of sucklings, has been reported.

Plimmer, who has published a careful study of the distribution
of lactase since most of our experiments were completed, reaches
these conclusions: “ Neither frog nor fowl have lactase in their in-
testine, and we may conclude that animals lower than mammals do
not possess this ferment.” ‘Of the mammals the carnivora and
omnivora have lactase present during the whole of their lives, but the
herbivora only when they are young, with the exception of the
rabbit.” With reference to this animal, divergent views have been
expressed. In the case of the adult pig, also, Plimmer’s positive
results differ from those reported by Portier. As an illustration of
the influence of age, Plimmer found that lactase appears in the in-
testine of the rat between two days and twelve hours before birth.
In the guinea. pig it is lost already five weeks after birth. Bierry
and Salazar found lactase in the foetus of the cow at the fourth
month; in that of the sheep at the end of the second month. It
is interesting to note that Frouin and Thomas! have found the
intestinal contents of the foetus capable of splitting various glucosides.
With regard to artificial adaptation by feeding lactose or milk to.
animals Plimmer concludes, contrary to Bainbridge and Weinland:
“Neither the pancreas nor the intestine of animals can be made to
"2 Lactase is not present
in the tissues in general, according to Portier.

Attention has been directed primarily to the occurrence of these
enzymes in the embryonic intestine. The cells lining this portion of
the alimentary tract are among the earliest to develop into a char-
acteristic and distinct membrane with many of the structural features
of the tissue subsequently formed. In the earliest stages they exist
as a single layer of simple columnar epithelial cells lining the dorsal
wall of the yolk sac, later becoming differentiated into the special-
ized varieties of cells characteristic of the future alimentary canal.

Methods of investigation. — Before describing the technic of our
experiments certain factors contributing to the discrepancies between

adapt themselves to any particular diet.

' Frouin and THomMas: Comptes rendus de la société de biologie, 1907, xii,
Pp. 227. :

? In the case of the salivary glands in relation to amylase, MENDEL and UNDER-
HILL have reached a similar conclusion. Cf. Journal of biological chemistry, 1907,
iii, p. 135; also GARREY: 767d., p. xl.

Chemical Studies on Growth. 85

different investigators, or essential to any satisfactory procedure, may
be pointed out. Some of these have been particularly emphasized by
Plimmer (1907). Competent observers, for example, ROhmann and
Nagano (1903), have maintained that the inverting enzymes may be
unequally distributed in the different portions of the small intestine,
not to mention the remainder of the alimentary tract. This suggests
the desirability of examining the entire membrane, where possible.
The succus entericus contains these enzymes in weak concentration,
if at all. The epithelium of the intestine, however, is decidedly
richer in active ferment. A reasonable time is required for any
satisfactory extraction. The enzyme will not pass through a Berkfelt
filter; in view of the endocellular character of the inverting ferments
it is advantageous to use strained extracts without attempting a per-
fect separation of the cellular débris. Obviously, with weak extracts
a reasonable digestion period must be allowed in order to obtain
convincing results.

To determine whether inversion has taken place, the osazone
method cannot be depended upon where the degree of hydrolysis
is slight.1_ We have employed this only as a confirmatory test. For
Jactose the changes in rotatory power in the solutions are not always
satisfactory indications of digestion, owing to the small variations
which even a considerable conversion of lactose ((a)p = ~52.5°) to
dextrose ((a)p = *+52.7°) and galactose ((a)p =*80.3°) brings about.
More striking in any case is the increased reducing power of the
digested solutions, upon which we have especially depended. Like
Plimmer, we prefer the use of the gravimetric Allihn method in place
of volumetric copper processes. For filtering the cuprous precipitate
Gooch crucibles are regularly employed in this laboratory, the copper
being weighed as cupric oxide.

The tissues examined include the small intestine of the pig at
embryonic, suckling, and adult ages; the unhatched chick and adult
hen; and the newly born and suckling dog. The extracts were uni-
formly prepared by treating the finely comminuted tissue with about
three volumes of 2 per cent sodium fluoride solution at room tempera-
ture during twenty-four hours. In the case of the embryo material the
entire gut was used; in the adult tissues the mucosa was scraped off.
The comparatively-long period of extraction was selected in order to
make certain that the enzymes present would be removed. In the

1 Cf Brerry: Comptes rendus de la société de biologie, 1905, lviii, p. 700;
also PLIMMER : Journal of physiology, 1907, xxxv, p. 24.

86 Lafayelle B. Mendel and Philip Hl. Mitchell.

earlier experiments the extracts were filtered through paper; later
they were merely strained through cloth, following the procedure
recommended in the most successful investigations on the inverting
enzymes.

Most of the digestion trials, unless otherwise stated, were carried
out as follows: 5 c.c. of the intestinal extracts were added to 100 c.c.
of the sugar solutions, the latter of approximately 1 per cent
strength and containing 2 per cent of sodium fluoride. Control ex-
periments with boiled extract were simultaneously carried out in
every case under precisely comparable conditions. The digestions
were allowed to proceed at 38° C. during from forty-eight to seventy-
two hours, a period which the experience of previous investigators
has shown to be adequate to indicate digestive changes. The digest-
ing mixtures were then heated to coagulate the proteins present,
slightly concentrated on a water bath, filtered, the coagulum washed,
and the filtrates made up to 100 c.c., z.¢., the original volume of the
sugar solution used. The occurrence of inverting enzymes was
determined by a comparison of the reducing power of the sugar
solution before and at the end of the digestion. The losses inciden-
tal to the manipulation, retention of sugar in the coagulum, etc., can
be estimated from the data obtained with the (boiled) control
digestions.

The reducing power of the solutions was estimated in some early
preliminary trials by Pavy’s volumetric method.! Our conclusions
are, however, based upon data obtained with the gravimetric Allihn
method. The results are expressed in terms of cupric oxide obtained
from an aliquot portion of the final solution. Typical protocols
are summarized below.

Intestinal mucosa of the adult pig.— The extracts were prepared as
already described. The letters (a, 3, c, etc.) refer to extracts obtained
from different animals (Table I).

In another series mucosa from an adult pig was extracted during ¢ez
days with 2 per cent sodium fluoride solution before the digestion
trials were begun, in order to obtain any lactase present without fail.

* Sixty cubic centimetres of PAvy’s solution were employed in each estimation
concordant duplicate titrations being made in every case. About 6 to 7 c.c. of the
sugar solutions were necessary in the case of maltose and lactose. The variations
in duplicate experiments were expected not to exceed 0.3 c.c. We do not regard
the volumetric method as entirely satisfactory, and have always repeated the experi-
ments, using the gravimetric Allihn process.

Chemical Studies on Growth. 87

TABLE I.

ADULT PIc.

CuO from 25 c.c.
Source of

the extract. Results.

Sugar used.

Before digestion. | After digestion.

Maltose Pig a2 0.1200 0.2160 Maltase present

Pig a (boiled) 0.1200 0.1190 in the intestine

Sucrose Pig a none heavy reduction

Pig 4 ‘

iced Sucrase present

a (boiled) in the intestine
é (boiled)

Pig ¢ (boiled) ‘ ‘

Lactose Pig a |

Pig a (boiled)

NVo lactase found
Pig 4

Pig ¢

In view of the positive results recorded by others for the existence
of lactase in the gut of the adult pig, we have considered the possi-
bility of a localization of the enzyme in some portion of the small
intestine. To determine this we did not employ our usual method
as described above, but substituted the procedure suggested by
Plimmer. The entire length of this organ was divided into six parts,
each of which was separately examined. The mucosa was scraped
from six feet of gut from each section so as to furnish 75 gm. of

TABLE IL.

Solution required
to reduce 60 c.c.
Pavy’s solution,

CuO from

Solution. ae.

gm.
Original lactose solution . . . 0.406

Digestion solution . . .. . 0.392

“ “(boiled extract)

88 Lafayette B. Mendel and Philip H. Mitchel.

material. This was ground with clean sand and extracted twenty-
four hours with 200 c.c. water and 2 cc. toluene. After straining
through cloth, boiled and unboiled portions (50 c.c.) of each extract
were digested with 100 c.c. 4 per cent lactose solution, in the pres-
ence of toluene (2 c.c.). At the end of the experiment the solutions
were made up to 200 c.c., and 20 cc. of mercuric nitrate solution
were added. After six to twelve hours’ standing, 125 c.c. of clear
filtrate were exactly neutralized with potassium hydroxide, the un-
boiled extract and its control experiment with boiled extract being
kept at the same volume throughout. After filtration of the neutral-
ized mixtures, 100 c.c. of the filtrate were treated with hydrogen
sulphide gas. After removal of the mercuric sulphide, just sufficient
copper sulphate solution was added to 75 c.c. of the solution to com-
pletely remove the excess of hydrogen sulphide. The volumes of the
mixtures were then equalized with water and the reducing power
determined by Allihn’s gravimetric method in duplicate in aliquot por-
tions (20 c.c.) of the final filtrates! The experimental results are
summarized in Table IIT.

TABLE III.

LACTASE IN INTESTINE OF ADULT Pic.

Portion of intestine used CuO obtained

for extracts. (averages). Remarks.

Animal.

gm.
Upper six feet, next to pylorus (boiled) 0.2828 Control expt.

~ 0.3804 Lactase present
Second 0.5738
Third 0.3797
Fourth 0.3722
Fifth 0.3274
Sixth 0.2928 Lactase?

In view of these positive results, portions of small intestine just
beyond the pylorus and just before the coecum were used for
comparison (see Table IV).

In these experiments the positive results were verified by the osazone
method, phenyl-glucosazone crystals being obtained only in those

‘ Cf. PLIMMER: Journal of physiology, 1907, xxxv, p. 23.

Chemical Studies on Growth. 89

cases where the increased reducing power also gave evidence of an in-

version of the disaccharide.
TABLE: IV.

Portion of intestine used | CuO obtained

for extracts. (averages). Remarks.

Animal.

gm.
Upper part (boiled) 0.3157 | Control expt.

0.4050 Lactase present.

Pig x

Lower part (boiled) 0.3145 Control expt.
: 0.3164 No lactase.
Upper part (boiled) 0.3470 Control expt.
0.4241 Lactase present.
a er part (boiled) 0.2763 Control expt.

“¢ ss 0.2806 Lactase ?

|
Pig

Inasmuch as lactase was regularly found missing in the lower por-
tions of the small intestine, we are inclined to attribute the discrep-
ancies between the results already reported by Portier, Weinland,
and Plimmer regarding the presence of lactase in the adult pig’s gut,
to the use of different parts of the alimentary tube. We have no
record regarding the exact location of the portions of intestine used
in our earlier negative experiments. The subsequent experience
makes it appear likely that the lower portions where lactase is usu-
ally missing were used in those cases. A similar explanation may
apply to the experiments of the investigators quoted above.

Intestinal mucosa of suckling pigs. — An extract was prepared from
the small intestine of a suckling pig seven weeks old. The analyti-
cal data are summarized in Table V, page go, the analyses being
made in duplicate. The presence of all three enzymes, maltase,
sucrase, and lactase, was thus demonstrated in the young pig.

Experiments with embryonic pigs. — The experiments upon embry-
onic pigs were conducted with extracts from embryos of varying
ages. In order to afford an approximate idea of the degree of devel-
opment the specimens were measured in length of body, in which
terms they are referred to in the protocols. The variations in size
between the individuals of the same litter are not inconsiderable.
We are indebted to our colleague, Professor Coe, for the following
statement regarding the size of the pig embryo at various stages
of development:

90 Lafayette B. Mendel and Philip Hl. Mitchell.

“We have not been able to find in the literature exact data con-
cerning the relation between length of body and age of embryo in
the pig, so that our estimate of the actual age of the embryos of any
particular size has been influenced by the available data concerning
the rate of development of the embryos of other mammals. Further-
more, it is well known that the size of the pig embryo at any given

TABLE V.

SUCKLING PIc. -

CuO from 25 c.c.

Sugar | Character of
used. | the extract.

Results.
Before digestion. After digestion

gm.
Maltose boiled 0.2678 0. 4441
Maltase found.

unboiled 0.2678 0.2570

Sucrase found.
unboiled

Lactose boiled .2692 0.3613
Lactase found.

Sucrose boiled no reduction heavy reduction ms

unboiled } 0.2569

age is subject to great variation, doubtless depending to some extent
on the number of young in the litter and the size and general condi-
tion of the mother; and, moreover, the individual embryos of a single
litter exhibit a remarkable variation in size. Within certain wide
limits, however, it is possible to estimate the age of the embryo
of any size from a measurement of its body length.

“The figures show the estimated age of the embryos used

Average length of body Estimated age of embryo

mm. days.
25 2
a2 44
75 54

100 62

125 68

i775 80

200 88

230 96

280 IIo

300 112

y

Chemical Studies on Growth. QI

TABLE VI. (Embryo Pic.)

CuO from 25 c.c. 60 c.c. Pavy’s
olution. solution require.
Sugar | Source of the | Size of | ; Rawat!
used, extract. embryo.| Before | After | Before | After
diges- | diges- | diges- | diges-
tion. tion. tion. tion.
mm. gm. gm. | ¢.c. c.c,
Maltose | intestine 230 0.2429 | 0.2469 fs Ae Maltase ?
«boiled ee 0.2429 | 0.2339 ae As
£ 200 0.3215 | 0.3250 6.5 5.9 Maltase ?
Ke boiled u 0.3215 | 0.3080 6.5 6.8
3 175 eres ae cke Tell 5.9 Maltase
ss boiled “e Eves note fel Tel
ef 120 0.3215 | 0.363 6.5 4.7 Maltase
& boiled s 0.3215 | 0.313 6.5 6.9 ,
ss 120 refer 2 shavers fil 5.3 Maltase
* boiled os wales oar Hel! a2,
my 120 0.2429 | 0.2704 ws Ae Maltase
a boiled oe 0.2429 | 0.2360 oa of
oy 75 sak ne dicen 6.0 52 Maltase
AS boiled e sisters stasis 6.0 6.0
2 50 0.3215 | 0.310 6.5 6.4 No maltase
ce boiled as bsiehe Asi 6.5 6.5
liver 200 0.3215 | 0.331 6.5 6.1 No mattase ?
<s boiled ae 0.3215 asters 6.5 6.6
kidneys 120 0.3215 | 0.311 6.5 6.7 No maltase
as boiled fe Steet Sone 6.5 6.9
Sucrose | intestine 280 none none a ae No sucrase
fs boiled a none none
v 175 none none ae ite No sucrase
ee boiled “ none none
a 120 none none oe a8 No sucrase
.< boiled ue none none oe ..
Lactose | intestine 230 0.2811 | 0.3674 a AG Lactase
“boiled ‘ 0.2811 | 0.2760 a oi
«¢ 200 0.4060 | 0.4255 5.3 4.3 Lactase
« boiled se 0.4060 | 0.3950 Ses 5.4
rE 175 Sayers Siete 5.4 3.6 Lactase
de boiled < wares So 5e 5.4 5.6
oe 120 seiae orate 6.0 4.5 Lactase
«boiled cs SORE aoe 6.0 6.1
. 120 0.2811 | 0.3621 ve Ae Lactase
of! boiled “ 0.2811 | 0.2930 AP AG
a 75 era Lae 5.6 4.8 Lactase
Y boiled “ slayeis saws 5.6 5.4
s 50 0.4060 | 0.3990 5.3 5.4 No dactase
«boiled + 0.4060} .... 5:3 a3
liver 200 0.4060 | 0.4000 5.3 3-3 No /actase
€ boiled ss Are Sone 5.3 5:3
kidneys 120 0.4060 | 0.3900 oe ad No /actase

92 Lafayette B. Mendel and Philip 1. Mttchell.

in the present experiments, the length of the body being the meas-
urement from tip of snout to base of tail. The period of gestation is
about one hundred and twelve days @ixteen weeks).”

Experiments were also made with extracts of tissues other than
the intestine, in order to determine to what extent, if at all, the
specific inverting enzymes are early developed in various parts of the
embryo. The protocols appear in Table VI, page 9I.

The figures give evidence of the presence of lactase and maltase
in the embryonic intestine of the pig at an early age, these enzymes
being missed only in the smallest embryos (50 mm.) examined. The
constant failure to find sucrase in the embryonic intestine is in strik-
ing contrast with the presence of the other inverting enzymes and
with the conditions prevailing in the pig after birth. The negative
results with embryo liver and kidney extracts suggest the specific
localization of the inverting enzymes in the embryonic intestine.

Intestinal mucosa of the suckling dog. — In view of the differences
just recorded in the occurrence of the inverting enzymes at various
stages in the life history of the pig, we have made experiments with
intestinal extracts from the puppy in order to complete the data
already available in the literature. Quantitative variations in the con-
tent of lactase in young and old dogs have been noted by Ro6hmann
and Lappe, and Orban. We are unaware of comparative observations
respecting the other inverting enzymes.

The entire small intestine was extracted during forty-eight hours.’
In each digestion trial 100 c.c. of the sugar solution + 5 c.c. extract
were used. Duration of digestion = 3 days.

TABLE VII. (NEwty Born Doe.)

CuO from 25 c.c. solution.

Source of
extract.

Results.
In boiled control

Before digestion. | After digestion. after digestion.

gm

: gm. gm.
2 puppies / Maltose 0.2431 0.3938 0 2428 Maltase.

6 hours
after
birth

~ Sucrose none much none Sucrase.

\Lactose 0.3322 0.4360 0.3281 Lactase.

7 * | (Maltose 0.2428 0.3881 0.2460 Maltase.
2 puppies

old

2 weeks 1 Sucrose none much none Sucrase.

.Lactose 0.3058 0.3936 0.3050 Lactase.

Chemical Studies on Growth. 93

All three enzymes were found as in the suckling pig (Table V).
Pautz and Vogel, Miura, and Weinland have described the occurrence
of them in newly born infants also.

Intestinal mucosa of the full-grown hen. — The entire intestine,
cleaned and comminuted immediately after death, was extracted
thirty hours. In each trial 100 c.c. of the sugar solution + 10 c.c.
extract were digested forty hours.

TABLE VIII.

FULL Grown HEN.

CuO from 25 c.c. solution.

Sugar used. Results.
In boiled control

Before digestion. | After digestion. after digestion.

gm. gm. | gm.
Maltose. . 0.1201 0.2093 0.1218 Maltase.

Sucrose. . none much none Sucrase.

Eactose . . 0.3153 0.3052 asiere No /actase.

Corresponding with the observations of Weinland, Portier, and
Plimmer on birds, no lactase was detected (Table VIII).

Intestinal mucosa of the chick at birth. — The entire intestine of
nineteen chicks at birth was extracted in the usual way during
seventy-two hours. In the digestion trials 100 c.c. of sugar solution
+5 c.c. extract were kept at 38° during forty-eight hours (Table IX).

TABLE IX.

NEWLY Born CHICK.

CuO from 25 c.c. solution.

Sugar used. Results.
Before After In boiled control
digestion. digestion. after digestion.
RIMGEOSE) «oe none 0.1890 none Sucrase.
Weaectose: «<2. 0.2458 0.2422 No /actase.

These results on the newly born chick, quite contrary to the find- °
ings in the embryonic pig and other newly born animals, give addi-
tional evidence of the biological individuality of this group of animals,

94 Lafayette B. Mendel and Philip Hl. Mitchell.

and warn against broad generalizations on the basis of observations
made on single species. We are not aware that the presence of
sucrase in embryo birds has previously been noted.

SUMMARY AND CONCLUSIONS.

The early appearance of inverting enzymes in the intestine of
the embryo corresponds with the relatively early specialization and
histological development of the portion of the alimentary tract here
investigated. The alimentary proteolytic enzymes, like the special
glands which elaborate them, come into evidence at a comparatively
late period. Maltase is the most universally distributed of all invert-
ing enzymes. In the embryo pig maltase and lactase are found in
the intestine, while sucrase is missing. After birth all three enzymes
are present. In the full-grown pig lactase is not regularly found in
all portions of the small intestine. We attribute the differences re-
ported by the various investigators (Plimmer, Weinland, Portier) to
the probability of their having used different portions of the small
intestine, since the experiments recorded by us show an unlike dis-
tribution of the enzyme in various regions of the intestine. In the
newly born puppy all the enzymes are found.

In birds other conditions prevail. Lactase is not found at any
period; sucrase, on the other hand, is uniformly present in the newly
hatched chick and the adult hen. One might be inclined toward a
teleological explanation for the absence of lactase from the intestine
of non-mammalian animals, and similarly for the absence of sucrase
from the embryos of the pig, sheep, and cattle. Such considerations
apply with less force, however, to the subsequent formation of su-
crase, or its embryonic occurrence in birds. For the present, the
statistics of the occurrence of the alimentary inverting enzymes must
await a more adequate interpretation with respect to their functional
significance. At any rate, it is safe to conclude that the alimen-
tary tract of the young mammal is, as a rule, even more adequately
equipped to digest and utilize the sugar of the milk than are the
adults of the same species.

BIBLIOGRAPHY.
MALTASE.

1880. Brown and Heron: Annalen der Chemie, cciv, p. 243 (pig).
1883. BOURQUELOT: Comptes rendus de l’académie des sciences, xcvii, pp. 1000
and 1322; Journal d’anatomie et physiologie, 1886, xxii, p. 161 (rabbit).

1892.

1892.
1894.
1895.
1895.

1895.
1896.

1896.
1898.
1goo.

1903.

1871.

1880.
1883.
1889.
1891.
1891.

1892.
1892.
1895.
1895.
1896.
1896.
1goo.
1903.
1904.

1890.
1895.
1895.

Chemical Studies on Growth. 95

Tussy and MANNING: Guy’s Hospital reports, xlviii, p. 271 ; Centralblatt
fiir die medizinischen Wissenschaften, 1892, p. 945 (human intestinal
juice).

SHORE and TEBB: Journal of physiology, xiii, p. xix; TEBB: zdzd., 1894,
XV, p. 421 (pig, rabbit, ox, sheep).

ROHMANN: Berichte der deutschen chemischen Gesellschaft, xxvii, p. 3251
(dog’s intestinal juice).

HAMBURGER: Archiv fiir die gesammte Physiologie, lx, p. 543 (dog’s in-
testinal juice).

PREGL: Archiv fiir die gesammte Physiologie, lxi, p. 359 (sheep’s intes-
tinal juice).

Pautz and VOGEL: Zeitschrift fiir Biologie, xxxii, p. 304 (dog, infant).

MENDEL: Archiv fiir die gesammte Physiologie, Ixiii, p. 425 (dog’s intes-
tinal juice).

FISCHER and NIEBEL: Sitzungsberichte der Berliner Akademie, i, p. 73
(horse, ox, calf).

DAVENIERE, PORTIER and POZERSKI: Comptes rendus de la société de
biologie, 1, p. 515 (mammals).

BIERRY: Comptes rendus de la société de biologie, lii, p. 1080 (embryos
of sheep and cattle).

ROHMANN and NaGAno: Archiv fiir die gesammte Physiologie, xcv, p. 533
(dog’s intestinal juice and mucosa).

SUCRASE.

PASCHUTIN: Archiv fiir Anatomie und Physiologie, p. 305 (pig, rabbit,
dog, rat, calf, sheep).

Brown and HERON: Cf Maltase.

BouRQUELOT: Cf. Maltase.

BASTIANELLI: Moleschott’s Untersuchungen zur Naturlehre.

GRUNERT: Centralblatt fiir Physiologie, v, p. 285 (dog).

KRUGER: Die Verdauungsfermente beim Embryo und Neugeborenen,
Wiesbaden, p. 71.

Tupsy and MANNING: Cf. Maltase.

KRUGER : Centralblatt fiir die medizinischen Wissenschaften, No. 32.

PREGL: Cf. Maltase.

MiuRrA: Zeitschrift fiir Biologie, xxxii, p. 266 (infant). °

MENDEL: Cf. Maltase.

FISCHER and NIEBEL: Cf. Maltase (horse, ox, calf, sheep).

BIERRY: Cf. Maltase.

ROHMANN and NaGAno: Cf. Maltase.

HAMBURGER and HeEkma: Journal de physiologie, vi, p. 40 (human intes-
tinal juice).

LACTASE.

DAsSTRE: Archives de physiologie, p. 103 (dog’s intestinal juice).

Pautz and VOGEL: Cf Maltase.

ROHMANN and Lappe: Berichte der deutschen chemischen Gesellschaft,
XXVvili, p. 2506 (cow, calf, dog, puppy).

96

1895.
1896.
1896.
1898.
1899.
1899.
1900.

19ol.

1903.
1904.

1904.
1904.
1904.

1905.
1906.

1906.
1906.
1907.
1907.
1907.

1907.

Lafayette B. Mendel and Philip Hl. Mitchell.

PREGL: Cf Maltase.

MENDEL: Cf. Maltase.

FISCHER and NIEBEL: Cf Maltase.

PORTIER: Comptes rendus de la société de biologie, 1, p. 387 (pig, rabbit,
dog, calf, bird).

WEINLAND: Zeitschrift fiir Biologie, xxxviii, p. 16 (pig, rabbit, dog, horse,
cow, calf, sheep, lamb, goat, hen, infant).

ORBAN : Jahresbericht fiir Thierchemie, xxix, p. 384 (rabbit, dog, infant).

BiERRY: Cf Maltase. ‘

PorTIER and BIERRY: Comptes rendus de la société de biologie, liii,
p. 810 (duck).

RGOHMANN and NAGANO: Cf. Maltase.

PORTIER: Comptes rendus de la société de biologie, lvii, p. 205 (press
juice of organs).

HAMBURGER and HEKMA: Journal de physiologie, vi, p. 40 (human intes-
tinal juice).

BIeERRY and SALAZAR: Comptes rendus de la société de biologie, lvii,
p. 181 (rabbit, dog, calf, embryos).

Bierky: Comptes rendus de la société de biologie, lviii, p. 700 (critique of
methods).

PORCHER: Comptes rendus de l’académie des sciences, cx], p. 1406 (goat).

PORCHER : Comptes rendus de la société de biologie, Ix, p. 1114 (feces of
sucklings),

FROUIN and PORCHER: Comptes rendus de la société de biologie, 1xi,
p- Ioo (intestinal juice).

LANGSTEIN and STEINITZ: Beitrage zur chemischen Physiologie, vii,
p- 575 (feces of children).

PLIMMER: Journal of physiology, xxxv, p. 20 (pig, rabbit, guinea pig, frog,

fowl, cat, dog, monkey, calf, sheep, rat, embryo rat).

BIERRY and SCHAEFFER: Comptes rendus de la société de biologie, 1xii,
p- 723 (dialysis of lactase).

Sisto: Archivio di fisiologia, iv, p. 116 (adaptation to lactose in adult
mammals).

MARTINELLI: Zentralblatt fiir die Physiologie des Stoffwechsels, ii, p. 481
(adaptation trials).

ore

CHEMICAL STUDIES ON GROWTH.—II. THE ENZYMES
INVOLVED IN PURINE METABOLISM IN THE EM-
BRYO.,

By LAFAYETTE B. MENDEL anp PHILIP H. MITCHELL.
[From the Sheffield Laboratory of Physiological Chemistry, Yale University.]

“HE progress of physiological research has demonstrated that
the metabolism of nucleoproteins is characterized by features
which are specific and. peculiar to these compounds, in distinction
from the simple proteins. To this may be added the recognition that
the chemical changes involved are facilitated by unique enzymes
found widely distributed in both animai and plant tissues. The ex-
perimental studies up to the present time have been concerned almost
entirely with the chemical transformations of the purine derivatives
associated with the nucleoproteins.? Sufficient data have already
been accumulated in such investigations to encourage the belief that
diverse pathological conditions may be connected with disturbances
in the metabolism of the purines and referable to alterations in the
occurrence or activities of the enzymes just referred to. This view
is fortified by the increasing importance which is being attributed to
the 7é/e of enzymes in the nutritive exchanges of living organisms.

The purine metabolism of developing tissues has scarcely been in-
vestigated. It is generally conceded that a synthesis of purines is
unusual, if it occurs at all, in adult mammals.2 In growing animals
and during embryonic development other conditions prevail.* The
growth and metamorphosis of tissues during developmental periods

1 This research was conducted with the aid of a grant from the Carnegie Insti-
tution of Washington.

2 The recent extensive and almost complete résumé of the literature on this
subject, by BLocH: Biochemisches Centralblatt, 1906, v, pp. 521, 561, 817, 873,
makes a detailed reference to individual papers unnecessary.

8 The evidence on this point has beenreviewed by MENDEL: The Harvey lectures
1905-6, p. 208; Journal of the American Medical Association, March 31, 1906.

* The experimental results of our study of purine synthesis in the embryo will
be reported in a subsequent paper of this series.

97

98 Lafayette B. Mendel and Philip FH. Mitchell.

are likely to be attended by an appropriate capacity to alter the avail-
able materials. The methods of study which have been applied to
the adult consist in an investigation of the changes in the body as a
whole, the end products of purine metabolism being considered; or
in observations on the action of individual organs or tissue extracts
upon nucleoproteins and purines. We have employed the latter
method in the present research upon the equipment of the embry-
onic organism (of the pig) in respect to the purine-transforming
enzymes.

Those transformations of the nucleoproteins in which the purines
are involved are at present referable to several distinct. types of en-
zymes which appear to be specific:

1. Nuclease -—— which liberates the purines from the nucleic acid molecule.

2. Deamidizing enzymes; adenase and guanase — which act upon amino-
purines to liberate ammonia by hydrolysis, forming oxypurines (hypo-
xanthine and xanthine).

3. XNantho-oxidase — transforming these oxypurines to uric acid by oxidative
change.

4. Uricolytic enzyme — effecting a destruction of uric acid, and forming
allantoin as one of the probable intermediary products.

Nuclease. — Enzymes of this class unquestionably have a wide dis-
tribution. They have been described in animal and plant tissues.!
Jones showed the close association of nucleases with nucleoproteins,
and Sachs has given evidence that they are not identical with tryp-
sin. In the blood nuclease appears to be missing. Numerous experi-
ments upon autolysis have, as is well known, given evidence of the
liberation of purines from their nucleic acid complexes. The specific
character of the nucleases is, however, not proved thereby. With
respect to the participation of nucleases in the possible degradative
changes in embryonic tissues ? little is known. Sachs has reported
the presence of nuclease in the pancreas of newly born dogs.

1 Cf. ARAKI: Zeitschrift fiir physiologische Chemie, 1903, xxxviii, p. 84;
IwANOFF: /é7d., 1903, Xxxix, p. 31 (the term zzclease first used here); PLENGE:
Lbid., 1903, XxXxix, p. 190; JONES: /é7d., 1904, xli, p. ror, xlii, p. 35; NAKAYAMA:
lbid., 1904, xli, p. 360; SCHITTENHELM : /éid., 1904, xlii, p. 257; SHIGA: Jézd.,
1904, xlii, p. 502 (yeast); SacHs: J/ézd., 1905, xlvi, p. 337 (nuclease and trypsin) ;
ABDERHALDEN and SCHITTENHELM: Jé7d., 1906, xlvii, p. 452 (digestive juices) ;
KIKKoJI: /ézd., 1907, li, p. 201 (fungi); Foa: Zentralblatt fiir Physiologie, 1907,
aol uP: 24 Gatesnel juice).

- The extent of these autolytic changes zal be discussed in a subsequent
paper.

Chemical Studies on Growth. 99

Adenase and guanase.— The significance of these enzymes which
transform adenine and guanine into hypoxanthine and xanthine,
respectively, involves the question as to what purine groups occur
preformed in the nucleic acid molecule. Although in the earlier
studies the four familiar purines mentioned above were reported as
decomposition (cleavage) products of nucleoprotein materials, the
refinement of analytical methods has tended more and more to limit
the yield to the two aminopurines, adenine and guanine. The very
recent investigations of Jones and Austrian! give convincing evidence
in this direction. They find that thymus nucleic acid fails to produce
xanthine by the action of nuclease, and show that hydrolysis at high
temperature (as conducted by most previous experimenters) gives
results that are misleading.?, The writer is quite in accord with the
views of these authors; and the earlier experience of Levene ? on
the hydrolysis of fresh glands points in a similar direction. The
appearance of hypoxanthine and xanthine under conditions favorable
to enzyme activity is therefore referable to the aminopurine precursors.
Such reactions have been abundantly demonstrated by the work of
Schittenhelm,* Jones and Levene. Active enzyme preparations have
already been obtained from extracts of various tissues. Upon one
point the views are somewhat conflicting. Jones postulated the
existence of distinct adenase and guanase enzymes because extracts
‘of certain organs employed by him failed to convert guanine into
xanthine, although adenine was readily transformed into hypoxan-
thine.® The contrary results of Schittenhelm ® were shown by Jones?
to be due to the differences in the animal species employed in their
experiments. Thus, the spleen of the pig lacks guanase, which is
abundant in the same organ of cattle. Without further review of the
controversial discussion on this point ® we may add that our own ex-

1 JoNEs and AusTRIAN: Journal of biological chemistry, 1907, iii, p. 1.

2 The destruction of purines (with the probable formation of pyrimidines)
under certain conditions of hydrolysis with strong acids has been shown quite
lately by BuRIAN: Zeitschrift fiir physiologische Chemie, 1907, li, p. 438.

8 LEVENE: This journal, 1904-5, xii, p. 276.

* See their numerous papers in the Zeitschrift fiir physiologische Chemie since
1904.

Tans and PARTRIDGE: Zeitschrift fiir physiologische Chemie, 1904, xlii,
p- 343 (guanase) ; JONES and WINTERNITZ : /ézd., 1905, xliv, p. I (adenase).

6 SCHITTENHELM : Zeitschrift fiir physiologische Chemie, 1905, xlv, p. 152.

7 Jones: Zeitschrift fiir physiologische Chemie, 1905, xlv, p. 84.

8 Cf. JONES and AUSTRIAN: Zeitschrift fiir physiologische Chemie, xlviii,
p- 110; SCHITTENHELM : /6zd., 1900, xlviii, p. 571; SCHITTENHELM and SCHMID:
Tbid., 1906, 1, p. 30.

100 Lafayette B. Mendel and Philip H. Mitchell.

periments corroborate the experience of Jones in showing the absence
of guanase and the presence of adenase in the same species and
organs. Both investigators have contributed valuable data to our
knowledge of the distribution of the purine-transforming enzymes in
various animals. The unlike equipment oi different organisms in
this respect is indicative of noteworthy variations in the purine me-
tabolism of different species, and at once dispels the idea that these
characteristic enzymatic properties are common to all active tissues.

Xantho-oxidase. — When glandular tissues are allowed to autolyze,
or aminopurines are acted upon by tissue extracts or appropriate
enzyme solutions in the absence of oxygen, the final purine products
are oxypurines (hypoxanthine and xanthine). With an abundant sup-
ply of air, however, uric acid may be formed by some tissues. The
facility with which this reaction can be accomplished depends in turn
upon a specific enzyme to which Burian ! has applied the name xantho-
oxidase, xanthine presumably being the intermediary product which
experiences oxidation, thus:

C;H,N,O ——> C;H,N,O, ——> C;H,N,0;

§ hypoxanthine xanthine uric acid

Uricolytic enzyme. — The destruction of uric acid by organ pulp and
tissue extracts is likewise well established. Since uric acid is quite
readily oxidized in pure solution in the presence of alkalies, some
doubt has been expressed? as to how far, if at all, the destruction ex-
perimentally observed by previous investigators — notably Schitten-
helm,’ who used alkali solutions of uric acid in his experiments — is
attributable to enzyme action rather than the destructive action of the
alkali present in excess. We have given careful consideration to this
criticism in relation to the present experimental work. In another
place* one of us has already considered the action of alkalies upon

' BurIANn: Zeitschrift fiir physiologische Chemie, 1906, xliii, p. 497. This
mode of formation of uric acid had, of course, long been recognized. Cf
SPITZER: Archiv fiir die gesammte Physiologie, 1899, Ixxvi, p. 192; WIENER:
Archiv fiir experimentelle Pathologie und Pharmakologie, 1899, xlii, p. 373;
SCHITTENHELM : Zeitschrift fiir physiologische Chemie, 1904, xlii, p. 251, etc.

2 Cf. AuSTIN: Journal of medical research, 1906, xv, p. 309; 1907, xvi, p. 71.

® SCHITTENHELM: Zeitschrift fiir physiologische Chemie, 1904, xliii, p. 239;
1905, xlv, p. 121 (term “uricolytic enzyme” introduced); p. 161 (enzyme isolated)
Cf. also WIECHOWSKI and WIENER: Beitrage zur chemischen Physiologie, 1907,
1X, P. 247.

* MITCHELL : Journal of biological chemistry, 1907, iii, p. 145.

Chemical Studies on Growth. IOI

uric acid, and found, in conformity with previous observations, that
the compound can easily be destroyed in alkaline solution. It was
shown, however, that this does not occur when protein is present,
since alkali-protein is relatively inert. Furthermore, the quantities
of alkali used to dissolve the uric acid in all our experiments quoted
below was minimal in excess of that necessary to form soluble alkali
urate. The evidence in favor of a specific uricolytic enzyme is strength-
ened by the fact which we have repeatedly verified, namely, that uric
acid is not destroyed by extracts of certain embryonic organs, although
comparable extracts from adult tissues readily destroy it. Finally,
the destructive action is lost when the extracts are boiled, — an ob-
servation quite in accord with the behavior of enzymes. Levene
and Beatty! have observed uricolysis by tissue extracts in acid
solution.
EXPERIMENTAL PART.

reliminary experiments were undertaken to ascertain whether any
of the familiar enzyme transformations of the purines can be demon-
strated in embryonic tissues. The livers of embryo pigs were em-
ployed for this purpose. In the adult liver of this species Jones and
Austrian! noted adenase and xantho-oxidase, while guanase was miss-
ing. Their results may be indicated by the following graphic scheme:

Adult pig’s liver.

guanine adenine

|
. NG
uric acid <—— xanthine <——— hypoxanthine

in which the enzyme reactions observed are indicated by an arrow,
those missing by the interrupted line. The purines present in fresh
embryo livers and autolyzed embryo livers were at first isolated by
the procedure suggested by Levene, after hydrolysis of the materials
with acid. This method is at best unsatisfactory and only approxt-
mate from a quantitative standpoint.

Method.t— The finely pulped livers were mixed with two or three
parts of water, toluene was added in abundance, and the mixture

1 LEVENE and Beatty: Proceedings Society for Experimental Biology and
Medicine, 1907, iv, p. 109.

2 Jones and AvsTRIAN: Zeitschrift fiir physiologische Chemie, 1906, xlviii,
p: 120.

8 LEVENE: This journal, 1904-5, xii, p. 276.

4 The experiments upon the fresh and autolyzed organs were carried out by
Mr. C. S. LEAVENWORTH.

102 Lafayette B. Mendel and Philip H. Mitchell.

subjected to autolysis, with occasional shaking, in a closed vessel at
38°. After a period varying from three weeks to two months the
mixtures, which showed decided change in the direction of solution,
were treated with sulphuric acid to the extent of about 5 per cent of
the entire solution. The mixture was-heated until the so/u¢zon failed
to give a biuret reaction. This usually required about forty-two
hours. After cooling the mixture was filtered, the residue was thor-
oughly washed with water, and the united solutions precipitated with
Hopkins’ mercuric sulphate solution. The mercury precipitate was
decomposed in the manner followed by Levene.! The guanine was
removed from the resulting solution of the purine bases by precipita-
tion with ammonia. The guanine was converted into the hydro-
chloride, weighed and analyzed in this form. The remaining purine
bases were separated by means of an ammoniacal solution of silver
chloride, and the silver-purine precipitate was decomposed according
to the well-known directions of Kriiger and Salomon.? Hypoxanthine
was identified as the nitrate, adenine as picrate (m.p. 276°-283° C.),
and xanthine was searched for in the residues. Fresh tissues were
decomposed directly with 5 per cent sulphuric acid and aneaed
in precisely similar manner.

Adult pig liver.— For purposes of comparison estimations of the
purines in 200 gm. (64 gm. of dry substance) of the liver of a full-grown
pig were made in fresh tissue and after seventeen days’ autolysis |
(Table I). These autolyses were subsequent to those reported below

TABLE I.

Found. Fresh liver. Autolyzed liver.

gm. gm.
Guanine hydrochloride 0.1814 0.135

Hypoxanthine nitrate 0.058 0.046

Adenine picrate = 03228? none

found calculated
1 €s3HN,O.* HCl - 2H,0 2'H,0 = 15.9% 16.1%
after drying at 105°: IN Seca 37.3
ASH oy CAS (Ee

1 LEVENE: This journal, 1904-5, xii, p. 278.
* KRUGER and SALOMON: Zeitschrift fiir physiologische Chemie, 1899, xxvi,
P- 373-

Chemical Studies on Growth. 103

on embryo livers, and the precipitation of the purines was made by
the copper sulphate-sodium bisulphite method in this experiment
instead of by the procedure just described.

Averages calculated for 100 gm. of dry tissue substance, in grams:

TABLE II.

Purine bases. | Before autolysis. | After autolysis.

Gianiness -be) =) sis | 0.191 0.142

Adenine: oS hea ee 0.125 none

Hypoxanthine .. . 0.057 0.045

Embryo pig liver.— The livers were obtained from embryos of 50
mm., 75 mm., and 100 mm. in length. The variations in the size of
pig embryos and their relation to age have been discussed in an
earlier communication.! The average weight of the embryo pig liver
at various ages is estimated from our data as follows:

TABLE III.

| Length of embryo

} | Water content

from crown of head | Weight of liver. | a ees
| to base of tail. ;

mm. 4 per cent
5

80
79
$0

Adult pig abe snot 68

The isolated purines were identified by analysis in cases where the
quantities obtained permitted. The analytical data are summarized
in Table IV, page 104.

The figures presented plainly indicate the preponderance of guanine
and adenine in the native nucleoproteins of the embryo liver. The
small quantities of hypoxanthine isolated from the fresh livers may
well be attributed to incipient enzymatic changes already begun, to
purine material accumulated free in the liver tissue (as occurs in
adult muscle), or to the action of the reagents in the method of de-
composition employed.2, The changes effected by autolysis are char-

1 MENDEL and MITCHELL: This journal, 1907, xx, p. 90.

2 Cf. the criticisms of Jones and AUSTRIAN: Journal of biological chemistry,
1907, ill, p. I.

104 Lafayette B. Mendel and Philip H. Mitchell.

TABLE IV. (Empryo Pic LIVER.)

Fresh livers. Autolyzed livers.

Size of embryo | 50mm./75 mm.) 100mm./ 100mm.) 50mm. /75 mm. | 100mm. | 100mm.
|

Estimated age 44 da. | 54da.| 62da. | 62da. | 44da.| 54da. | 62da. | 62 da.

Number of livers! 999) 100 | 46 440 | 202]/ 225 1 ogee

used
Dry substance | 56 gm. | 49 gm.| 37 gm. | 112 gm. | 56 gm.| 111 gm.| 37 gm. | 132 gm.
Purines isolated:

Guanine hydro-) ¢ 3351] 9.2762/ 0.179 | 0.5275| 0.390 | 0.745 | 0.14419] 0.998

chloride

Adenine picrate| 0.617 | 0.3702| 0.2504 0.676% | 0.1728| 0.2769] 0.04014] 0.09812
Hypoxanthine | 996g | 0.04023, 0.00623| 0.1167] 0.138 | 0.311 | 0.20523| 0.490

nitrate

Xanthine |
|

1 C;H;N;O*HC1:‘2H,O H,0=16.15%; calculated 16.1%. In the dehydrated
salt: C,H;N;O - HCl N= 37.1%; calculated 37.3%.
2 C;H;N;O ‘HCl: 2H,O H,O = found 15.8%; calculated 16.1%. The salt was
converted to the free base and analyzed as C;H;N;0 N = 45.7%; calculated 46.3%.
8 C;H;N; ° CgH,(NO,); OH: H,O m. p. 2829-2839 C. N = 29.05%; calculated
939. 4 m. p. 282-283° C.
C;H;N;0*HCl:2H,O0 H,O = 16.15%; calculated 16.1%. In the dehydrated

bo

sait: C;,H;N;0: HCl N =37.2%; calculated 37.3%.

6 m. p. 280°-282° C. Analyzed, after drying, as C;H;N, ° C,H.(NO,.)3° OH N=
30.3%; calculated 30.7%

7 C;H,N,O* HNO3" H,O N = 32.26%; calculated 32.2%.

8 m. p. 280° C. 9 m. p. 278°-280° C.

1” C;H;,N;,O° HC1'2H,O H,O = 15.95%; calculated 16.1%. The dehydrated
salt was analyzed as C;sH;N;0* HCl N = 37.4%; calculated 37.3%.

11 m. p. 282°-283° C. 12 m. p. 278°-280° C.

13 The hypoxanthine nitrate from three experiments was united and analyzed after
recrystallization. C;H,NsO* HNO;*H,O H,O = 8.7%; calculated 8.2%. In the
dehydrated salt: C;H,NsyO* HNO; N = 35.0%; calculated 35.2%.

Averages calculated for 100 gm. of dry tissue substance, in grams:

Embryos 50 mm. Embryos 75 mm. Embryos 100 mm.

Purine bases.

Before After Before After Before After
autolysis. | autolysis. | autolysis. | autolysis. | autolysis. | autolysis.

(Guanine eee 0.405 0.470 0.380 0.453 0.323 0.385
Adenine ... 0.389 0.109 0.266 0.088 0.225 0.031
Hypoxanthine . 0.076 0.154 0.051 0.176 0.038 0.208

Xanthine .

Chemical Studies on Growth. 105

acteristic in each case, and correspond with the experience with adult
pig’s liver. The content of guanine is scarcely altered, especially
when the unsatisfactory adaptation of the methods used for gzantz-
tative work is taken into account. Corresponding with this is the
failure to find xanthine. We have united the residues obtained from
1127 livers (488 fresh and 639 after autolysis) in the treatment of
the silver-purine precipitates by the Kriiger-Salomon method, under
the conditions in which xanthine is usually separated by this process,
without obtaining any of this base whatever from the combined traces
of insoluble material. Guanase is therefore undoubtedly missing in
the liver of the pig embryo, —a finding which we have verified by
other methods to be described later. This peculiarity of the tissues
of the pig thus fully corroborates the results already described by
Jones and Winternitz! for the adult pig liver, as does our observa-
tion on the occurrence of adenase. The existence of this enzyme is
plainly indicated in the experiments recorded above, by the marked
disappearance of adenine with simultaneous increase in the content
of hypoxanthine. The divergent results obtained by Levene? are
undoubtedly attributable to the use of the glands of some other
species than the pig.

The specific characteristics of the purine-transforming enzymes of the
liver are thus already demonstrable in the embryonic organ. Nuclease
and adenase are present in the liver at a very early stage in embryonic
life.

EXPERIMENTS WITH EXTRACTS OF EMBRYONIC TISSUES.

In extension of the observations already recorded we have searched
for nuclease, adenase, guanase, xantho-oxidase, and uricolytic enzymes
in some of the tissues of the embryo pig at various stages of develop-
ment. The materials used were always quite fresh, never more than
two or three hours elapsing until the experiments were begun. . Asa
rule, the embryo livers were used in one set of experiments, while the
remaining viscera, including lungs, heart, alimentary tract, spleen,
pancreas, and kidneys, were used together in a comparable series.
The material was always comminuted in a small -hashing-machine,
and treated with five times its weight of water, with an abundance
of toluene. The extraction was continued with frequent stirring at
room temperature for periods varying, in different experiments, from

1 JONES and WINTERNITZz: Zeitschrift fiir physiologische Chemie, 1905, xliv,

p. 8.
2 LEVENE: This journal, 1904-5, xii, p. 295.

106 Lafayette B. Mendel and Philip H. Mitchell.

twenty-four to forty-eight hours. The mixture was allowed to settle,
the supernatant fluid strained through cloth, and then filtered quite
clear by suction through paper pulp. The digestions were conducted
at about 38°, an excess of toluene always being present.

Analytical methods. — The purine compounds were separated
by the copper sulphate-sodium bisulphite process of Kruger and
Schmid.' Reprecipitation was usually adopted after decomposition
of the purine precipitate with hydrogen sulphide. We have found
the use of aluminium acetate to effect clear solutions in filtration very
helpful.2 The quantitative estimations of total purines were con-
ducted according to the directions of Kruger and Schmid.? When
purine compounds were precipitated by the silver method, the sepa-
ration of the individual constituents followed the process of Kruger
and Salomon. Guanine was weighed as the hydrochloride or the
free base; adenine as the picrate; hypoxanthine, xanthine, and uric
acid were separated and identified in various ways indicated below:

Nuclease. —

Lixperiment 1.— The livers of 150 mm. pig embryos were extracted twenty-four
hours at room temperature. The filtered extract was analyzed in 500 c.c.
portions for free purine-base nitrogen (by double precipitation with the
Kriiger-Schmid reagents with subsequent Kjeldahl-N estimation), as
follows :

(@). Presh extraet’ 437 3 ¥ . + 0.036 gm. purine N.
(6) Extract after three days’ pees
with air drawn through . . 0.046

Experiment 2.— The same experiment was repeated with an extract of livers

from 175 mm. embryos, using 500 c.c. portions.

ee ce “

(2) ‘Fresh extract “2 2% . . 0.038 gm. purine N.
(6) Extract after three days’ aaealyae
with air drawn through . . o.o52 “ ara

Experiment 3.— In this experiment an extract of adu/t pig’s liver was used
iN 500 C.c. portions.

(a) Fresh-extract .. . . ./ .» o:e26 pm. purime aE
(az) “ec ce i Fs 0.026 vc cc cé
(2) Extract after ieee aye fauer

with air drawn through . . o.o16 “ eee
(b.) 66 “c 7 2d Seong, “6 “c «“

1 Cf Hoppe SEYLER-THIERFELDER: Handbuch der chemischen Analyse,
p- 435.

* Cf. SCHITTENHELM: Archiv fiir klinische Medizin, lxxxi, p. 429, 1904.

* KRUGER and ScHMID :/Zeitschrift fiir physiologische Chemie, 1905, xlv, p. I.

* KrUGER and SALomon: Zeitschrift fir physiologische Chemie, 1898, xxvi,
p- 350.

Chemical Studies on Growth. 107

Experiments 1 and 2 show an appreciable increase in the free
purines of the embryo liver extracts, and thus zvdicate the presence of
nuclease. Wespite the current of air steadily maintained through the
solutions, there is no loss of free purines, — quite in contrast with
what was observed with extracts of adu/¢ livers in Experiment 3.
This speaks against the presence of uricolytic activity in the emdryo
extracts, — an inference which is justified by other experiments spe-
cially directed towards this question. Calculated on the basis of
xanthine (36 per cent N), the quantity of free purines contained in
500 c.c. of the embryonic extracts was about 0.I gm.

Guanase. — It will be recalled that the failure to find guanase in the
liver of the pig was one of the observations which led Jones ! and his
co-workers to postulate the distinction between the two deamidizing
enzymes, adenase and guanase. Schittenhelm? admits that the action
of extracts of the liver of the adult pig upon guanine is feeble at most,
although he is apparently unwilling to allow the specific distinction
made by Jones. Our experience with adult livers of pigs corresponds
with that of Jones.

For example: One kilo of the finely comminuted liver of an adult pig was mixed
with 2.5 litres of water to which liberal amounts of chloroform and toluene
were added, and allowed to autolyze during thirty days at room temperature.
The mixture was then heated to boiling, treated with a little acetic acid to
effect complete coagulation, and filtered. In the filtrate purine bodies
were precipitated by copper sulphate and sodium bisulphite and liberated
in the usual way with sodium sulphide. The material was then evaporated
to dryness, and the residue was thoroughly extracted with 1 per cent
ammonia. The portion thus remaining undissolved after further washing
with dilute ammonia was taken up in sodium hydroxide, filtered, and
treated with a slight excess of acetic acid. The precipitate thus formed
weighed (after drying at 105°C.) 0.388 gm. It was shown to be guanine
by the fact that by proper recrystallization 0.35 gm. of pure guanine
hydrochloride was prepared from it. In the ammoniacal filtrate from
the separation of guanine, examination for other purine bases disclosed
no adenine, mere traces of hypoxanthine, and an amount of xanthine nitrate
equivalent of 0.85 gm. of xanthine.

This result confirms the presence of adenase and oxidase in the liver
of the pig, but indicates the absence of guanase. Identical results were
obtained in the above-mentioned experiments by Jones and Winternitz.

1 Jones and WINTERNITZ: Zeitschrift fiir physiologische Chemie, 1905, xliv,
p- 8; Jones and AustRIAN: J/did., 1906, xlviii, p. 120.
2 SCHITTENHELM: Zeitschrift fiir physiologische Chemie, 1905, xlvi, p. 368.

108 Lafayette B. Mendel and Philip H. Mitchell.

The data already reported in the preliminary autolysis experiments
pointed strongly to the absence of guanase in the liver of the embryo
pig, since the guanine content of the autolyzing organs was unaltered
after several weeks’ digestion in the absence of air. In what follows
it will be shown how extracts of embryonic organs act towards gua-
nine added directly to them.

Experiment 4. Lxtracts of embryo livers. — Livers of 150 mm. and 175 mm.
embryos were extracted at room temperature during thirty-six hours.
Portions of 500 c.c. of filtered extract were allowed to digest (a) with
and (4) without added guanine four days without air supply. The purines
were then isolated by a modification of the Kriiger-Salomon process. The
copper-free filtrates were evaporated to dryness and the residues extracted
twenty-four hours with 2 per cent ammonia. After repeated washing with
ammonia the undissolved material was dissolved in sodium hydrate solu-
tion and guanine precipitated with acetic acid. It was converted into the
beautifully crystalline hydrochloride and weighed as such.

(a) 500 c.c. extract + 0.19 gm. guanine hydrochloride ‘ (converted into
the free base by repeated evaporation with water and alcohol and then dis-
solved in a little dilute sodium hydrate solution).

(4) 500 c.c. extract alone.

Guanine was recovered : °
(az) 0.18 gm. guanine hydrochloride.
(3) 0.04 «cc “cc “

The salt from (a) dried at 105° C. was analyzed:
N = 37.1 per cent; calculated 37.3 per cent.

Making allowance for the guanine present in the extract itself, at
least 0.14 gm. of the hydrochloride was recovered. We attribute
the discrepancy to the inevitable losses incidental to the analytical
process used rather than to any enzyme activity. The filtrates from
the guanine were examined by the silver precipitation method for
other purines. No adenine picrate could be separated, but hypo-
xanthine was obtained as the nitrate. Adenase was therefore un-
doubtedly present.

Experiment 5, — Filtered liver extracts from embryos 120 mm. to 200 mm. in
length were used. The digestions, without air supply, lasted one week.
The guanine was weighed as the free base.

(2) 500 c.c. extract + 0.30 gm. guanine hydrochloride (representing
0.225 gm. guanine) dissolved in 15 c.c. 2 per cent sodium hydrate solu-
tion. Recovered: 0.26 gm. guanine.

? Prepared from an acid hydrolysis of pancreatic glands.

Chemical Studies on Growth. 109

(6) 1000 c.c. extract alone. Recovered: o.10 gm. guanine.
The guanine was converted into the hydrochloride, dried at 105°C.
and analyzed :
(2) 37-1%N; calculated: 37.3 %N.
(4) 37-5 %N.

. : 0.10
Correcting for the guanine content of the extract (Is = 6.05 gm.

the quantity recovered was 0.21 gm., or 93 per cent.

Experiment 6. Extracts of other viscera. — The viscera (except the livers)
of 150 mm. and 200 mm. embryos were extracted 36 hours. Filtered
extracts were employed, without air supply, and digested during 6 days.

(@) 700 c.c. extract + 0.21 gm. guanine (dissolved in sodium hydrate
solution).

(6) 700 ¢.c. extract alone.

The purines in 4 were negligible in amount. From @ 0.198 gm. of
xanthine was obtained. It was converted into the nitrate, then the free
base, dried at 105°C. and analyzed: N = 36.7%; calculated: N =
36.8 %.

- The conversion to xanthine was therefore practically quantitative.
Guanase ts present. ;

Experiment 7. Control experiment. —To a boiled and strained extract of
similar viscera 0.15 gm. of guanine (dissolved in sodium hydroxide solu-
tion) was added and a digestion of six days’ duration carried out.

The purine bases were separated according to the method of Kriiger
and Schittenhelm,’ which is more delicate for small quantities of xanthine.
No xanthine was obtained. 0.136 gm. of guanine was recovered and
converted into the characteristic hydrochloride.

Experiment 8.— The viscera of 120 mm. embryos were extracted forty-eight
hours, and the extract digested five days with 0.23 gm. guanine hydro-
chloride. Analysis as in Experiment 7.. Recovered:

0.04 gm. guanine hydrochloride.
0.157 ‘‘ xanthine.

The xanthine was converted into the nitrate and then the free base.
N = 36.6 % ; calculated: 36.8 %.

Experiment 9. — An extract of 230 mm. embryos was digested four days with
0.36 gm. guanine hydrochloride (=o.27 gm. guanine). Recovered:
0.26 gm. xanthine.

Under comparable conditions, therefore, the presence of guanase in
embryonic tissues other than the liver has thus been demonstrated.

1 KRUGER and SCHITTENHELM ; Zeitschrift fiir physiologische Chemie, 1902,
XXXV, Pp. 153.

110 Lafayette B. Mendel and Philip Hl. Mitchell.

Adenase. — The adult liver, spleen, and pancreas! of the pig have
been shown to contain an enzyme capable of converting adenine into
hypoxanthine. Our preliminary autolysis experiments indicated the
existence of adenase in the embryoliver also. We have made further
studies regarding its distribution.

Experiment 10. Extracts of embryo livers. — An extract of the livers of
200 mm. embryos was allowed to undergo autolysis during five days.
One litre of the filtered extract, equivalent to about 200 gm. tissue, yielded
0.075 gm. guanine.

NVo adenine could be separated with picric acid. Aypoxanthine was
isolated as the silver nitrate salt, and converted into the characteristic
nitrate. Recovered: 0.12 gm. Dried at 110° C. : N = 34.7 per cent;
calculated: 35.2 per cent.

Experiment 11.— The preformed bases present in the fresh livers of 200-230
mm. embryos were separated after hydrolysis in a steam sterilizer during
sixteen hours with 5 per cent sulphuric acid. 200 gm. of tissue, equiva-
lent to the quantities used in some of the extraction experiments, were
decomposed. The purine bases recovered were:

0.09 gm. guanine hydrochloride.
0.16 gm. adenine picrate, m. p. 280° C.

These data may be compared with some of those following.
Experiment 12.— Livers of 150-200 mm. embryos were extracted sixty hours.
The filtered extracts were digested without air supply during five days.

(a) 1000 c.c. of the extract alone.
(6) 500 c.c. extract + 0.19 gm. adenine sulphate.”

The products isolated were:

(2) (2)
guanine 0.18 gm. o.II gm.

hypoxanthine nitrate
calculated > 0.12 (N = 34.6%) 0.17 (N = 35-7%)

C;H,N,O “HNO; N= 252%,
0.12

Making allowance for 0.06 gm. (=) hypoxanthine nitrate obtainable

from 500 c.c. of extract alone, o.11 gm. hypoxanthine nitrate was obtained
from o.19 gm. adenine sulphate added. The guanine remained practically
unchanged, as has repeatedly been found in the preceding section.
Experiment 18. — Extracts from livers of embryos of 120-230 mm. length.
(2) 1000 c.c. of the extract alone.
(6) 500 c.c. extract + 0.10 gm. adenine’® (dissolved in sodium
hydrate solution).

1 Cf. JONES and WINTERNITZ: Zeitschrift fiir physiologische Chemie, 1905,
xliv, p. 1 (liver); SCHITTENHELM: /d7d., 1905, xlvi, p. 354 (spleen) ; JONES and
PARTRIDGE : /é7d., 1904, xlii, p. 343 (pancreas).

* Prepared by Dr. Mitchell from thymus glands.

® Furnished by C. F. BOEHRINGER und SOHNE, Mannheim.

Chemical Studies on Growth. Ill

Recovered : ; (2) (2)

Hypoxanthine nitrate 0.11 gm. 0.21 gm.
calculated

Nitrogen in C,H,N,O - HNO, - H.0 G ai eS SHAG eel,

Lxperiment 14.— 500 c.c. of a filtered extract of livers of 150,mm. embryos
were digested with 0.17 gm. adenine (dissolved in sodium hydrate solution).
Recovered: 0.25 gm. hypoxanthine nitrate.

Experiment 15. Extracts of other viscera. — From 500 c.c. of an extract of
the viscera (excluding the liver) of 230 mm. embryos, digested five days at
38°, no adenine could be separated. 0.04 gm. hypoxanthine nitrate was
isolated.

A similar experiment with 120 mm. embryos furnished 0.05 gm. hypo-
xanthine nitrate and xo adenine.

Adenase ts therefore present in both the liver and other viscera of the
embryo pig.

Oxidase. — The uric-acid-forming xantho-oxidase has been demon-
strated in the liver of the adult pig, but missed in the pancreas,
spleen, and lung.! The experiments must, of course, be conducted in
the presence of an abundant oxygen supply. In our experiments an
air current was driven through the solutions during about ten hours
of each day. Uric acid was separated by the method of Kruger and
Schmid.?

Experiment 17. Extracts of embryo livers. —A filtered extract was pre-
pared from the livers of 175 mm. and 230 mm. embryos. 800 c.c. were
digested three days, a current of air being driven through the solution
during thirty hours of that period. Vo uric acid or xanthine could be
tsolated. 0.07 gm. guanine and o.0og6 gm. hypoxanthine nitrate were
isolated. The latter was identified air-dry as C;sH,N,O - HNO;° H.O.
N = 32.1 per cent; calculated: 32.3 per cent.

Experiment 18.— Livers of 175-230 mm. embryos were extracted twenty-four
hours. The filtered extract was then used for digestions, air being driven
through during twenty hours of the two and one-half days’ experiment.

(a) 1000 c.c. extract alone.
(4) 500 c.c. extract + 0.20 gm. adenine (dissolved in sodium
hydrate solution).
No uric acid or xanthine was obtained from either solution.
(a) yielded a small quantity of guanine and 0.108 gm. hypoxanthine
nitrate.

1 Cf Jones and AvustTRIAN: Zeitschrift fiir physiologische Chemie, 1906,
xlviii, p. 120 (liver, pancreas); SCHITTENHELM: /67d., 1905, xlvi, p. 354 (lung,
spleen).

? KRUGER and SCHMID: Zeitschrift fiir physiologische Chemie, 1905, xlv, p. 1.

112 Lafayette B. Mendel and Philip H. Mitchell.

(2) yielded a small amount of guanine, 0.275 gm. adenine picrate
(equivalent to 0.102 gm. adenine) and o.208 gm. hypoxanthine nitrate.
The failure to effect a complete change of the adenine to hypoxanthine
was doubtless due to the brief period of digestion. Over 95 per cent of
the adenine which disappeared was recovered as hypoxanthine.

Experiment 19. — A comparable extract (500 c.c.) from the livers of 150 mm.
and 200 mm. embryos was digested with 0.17 gm. adenine (dissolved in
sodium hydroxide solution) during five days, air being driven through nine
hours each day. No adenine or uric acid were found at the end of the
experiment. 87 per cent of the adenine added was recovered as hypo-
xanthine (0.299 gm. hypoxanthine nitrate, air dry, N = 32.0 per cent;
calculated: 32.3 per cent).

Experiment 20. — 500 c.c. of a boiled and strained extract of the livers of
200 mm. embryos + 0.16 gm. adenine (dissolved in sodium hydrate
solution) were digested five days, air being driven through during thirty
hours of that time. A very small xanthine fraction was isolated, together
with a trace of guanine and 0.35 gm. adenine picrate (m. p. 281° C.)
representing 0.13 gm. adenine, z.¢., over 80 per cent of the added adenine
could be isolated, despite the difficulties of exact quantitative recovery.

NXantho-oxidase ts therefore missing in the liver of the embryo pig
during a considerable period of embryonic life. Adenase is present
and adenine can be converted almost quantitatively into hypoxanthine.
Uric acid is not formed, even in the presence of oxygen. That the fail-
ure to detect uric acid is not due to simultaneous uricolytic action,
by which any of the compound formed is at once destroyed, will be
shown in experiments recorded below.

Experiment 21. Extracts of other viscera.— 1000 c.c. of a filtered extract
from 230 mm. embryos + 0.38 gm. guanine hydrochloride (representing
0.285 gm. guanine) were digested four days, air being driven through
during nine hours of each day. Vo uric acid was found. ‘The xanthine
fraction (weighing 0.278 gm.) yielded 0.256 gm. xanthine nitrate which
was converted into the free base and analyzed. N = 36.5 per cent;
calculated: 36.8 per cent.

Experiment 22.— 1000 c.c. of a filtered extract of viscera of 150-200 mm.
embryos + 0.4 gm. guanine hydrochloride (equivalent to 0.30 gm. guanine)
were digested (with air) during four days. Analysis yielded:

No uric acid.
0.07 gm. guanine.
0.22 gm. xanthine.
The latter was purified by conversion into nitrate, and then the free
base, and analyzed : .

N = 36.4 per cent; calculated: 36.8 per cent.

Chemical Studies on Growth. ie

Guanase ts present in the extracts of the embryontc viscera other than
the liver. Uric acid could not be detected.

The changes in the equipment of enzymes incidental to growth are
shown by comparison of the data furnished by a suckling pig about
seven weeks old.

Experiment 23. Extracts of the liver of the suckling pig. —(a@) 500 c.c.
filtered extract + 0.24 gm. adenine sulphate. (4) 1000 c.c. extract alone.
The digestion proceeded five days, air being driven through at intervals
during forty hours.

Recovered. (a) (4)
WUriemeid §. = = <5 .62:050 pm. 0.007 gm.
N in other purines . . 0.037 gm. 0.024 gm.

The uric acid was crystalline and gave the murexide test. The forma-
tion of this compound and the marked disappearance of the other purine
bases give evidence of the occurrence of xantho-oxidase in the liver in
early adult life.

Experiment 24. Extract of the liver of the adult pig. — 1000 c.c. of a filtered
extract were digested with 0.25 gm. adenine (dissolved in sodium hydrate
solution) with an air current during three days. There were recovered:
No uric acid or adenine, a trace of xanthine, a small amount of guanine,
and only 0.03 gm. hypoxanthine nitrate. The disappearance of the ade-
nine and the failure to find a comparable quantity of hypoxanthine or
xanthine indicate the presence of a xantho-oxidase. The absence of
uric acid is attributable to the well-known uricolytic action of extracts
of adult pig’s liver. The experiment was carried out under conditions
strictly comparable with those pertaining in the case of the embryo
organs ; the differences are thus emphasized by contrast.

Uricolytic enzyme. — This enzyme has been demonstrated in the
liver of the adult pig by Schittenhelm ! and Wiener,? and in the kid-
ney by Pfeiffer.2 We have failed to detect its occurrence in extracts
of embryonic tissues. In searching for the enzyme tissue extracts
have been digested with uric acid (dissolved in a minimal excess of
sodium hydrate solution) at 38°, with a current of air driven through
the mixtures, following the procedure adopted by Schittenhelm.* The

1 SCHITTENHELM : Zeitschrift fiir physiologische Chemie, 1905, xlvi, p. 354.

2 WIENER: Archiv fiir experimentelle Pathologie und Pharmakologie, 1899,
hip. 375:

8 PFEIFFER: Beitrage zur chemischen Physiologie und Pathologie, 1906, vii,
p- 463.

* SCHITTENHELM : Zeitschrift fiir physiologische Chemie, 1905, xlv, p. 121.

114 Lafayette B. Mendel and Philp H. Mitchell.

criticisms of this method, with the possibility of destruction of uric
acid by the reagents employed, have already been considered on
page 100.!

Extracts of embryo livers. — 500 c.c. of filtered extract + 0.15 gm. of uric
acid were used in each case. The details are summarized below:

Experiment. Size of embryos. Uric acid recovered. _ Remarks

per cent,

mm. gm.
200 0.139 92

0.137

0.122
Four days’ digestion
0.127 N = 33. %;) ealeu-
ulated: 33.3 %

At later stages the liver contains a uricolytic enzyme.

Experiment 30. Extracts of the liver of suckling and adult pigs. — 500 c.c.
filtered extract of the liver of suckling pigs about two months old + 0.17
gm. uric acid were dissolved and digested in the usual way during four
days. 0.095 gm. uric acid = 55 per cent only was recovered.

Experiment 31. — 500 c.c. filtered extract of adu/t liver + 0.15 gm. uric acid
were employed in a series of trials.

Conditions. Uric acid recovered.
gm. per cent,
(a) Digestion one day 0.048 32
(b) «© three days 0.021 14
(c) = wiour sls none
(d) ce “ec “ec none

(e) Control, boiled ike
ai 0.139 g2
digestion three days
The control experiment (31e) here reported clearly indicates that
the uricolysis observed in the other trials is not attributable to the
reagents used. At precisely what age the uricolytic power is devel-
oped, z. ¢., whether prior to birth or later, was not determined. The
oldest embryos (200 mm. = 88 days) obtained by us showed entire
absence of the enzyme at that age. It is of interest to note that

} Cf. also MITCHELL: Journal of biological chemistry, 1907, iii, p. 145.

Chemical Studies on Growth. 115

Wiechowski ! has lately identified allantoin among the products of
uricolysis by tissues.

Extracts of other embryonic viscera. —The uricolytic enzyme could not be
detected in the extracts of these viscera.

. Size of the Uric acid
PILE embryo. added. recovered.
mm. gm. gm.
33 large 0.15 0.13
34 200 0.25 0:235 (N = 33-4%)
RESUME.

The nucleic acid of the liver of the embryo pig probably contains
only two purine complexes: adenine and guanine.

The liver is capable of undergoing autolytic changes at an early
age. Nucleases are present which liberate purine bases from the
nucleic acid complexes.

The liver of the embryo pig contains adenase, even in its early
stage of development, but no guanase. In this respect it shows the
specific character of the liver of the adult animal.

An extract of embryo viscera, other than the liver, readily gives
indication of the presence of guanase at an early age.

The unlike distribution of these two enzyme reactions under com-
parable conditions of experiment give further evidence in favor of
the existence of two distinct and specific deamidizing enzymes.

It*has not been possible to demonstrate the formation of uric acid
from preformed or added purine bases (adenine or hypoxanthine) by
extracts of embryonic tissues. The preliminary oxidative transfor-
mation of hypoxanthine to xanthine is likewise doubtful. Xantho-
oxidase is not present in the embryo visceral organs of the pig; it is
found, however, in the livers of the full-grown and suckling animals
of the same species.? The latter readily form uric acid from purines
under suitable conditions.

The uricolytic enzyme has not been found in extracts of embryo
pig tissues under conditions in which it is readily identified in the
adult organs. The enzyme appears either shortly before or after
birth. These observations speak in favor of the specific uricolytic

1 WIECHOWSKI: Beitrage zur chemischen Physiologie, 1907, ix. p. 295.

2 Comparable data obtained from a study of the embryo were reported by JONES
and AUSTRIAN at a meeting of the American Society of Biological Chemists in
Washington, May, 1907; Journal of biological chemistry, 1907, iii, p. 227.

116 Lafayette B. Mendel and Phitip H. Mitchell.

power of tissue extracts, and indicate that the destruction of uric
acid in such solutions is not solely due to the alkaline reaction, etc.,
of the digesting medium.

The tardy appearance of the oxidative and katabolic enzymes con-
cerned in the transformation of the purines is suggestive as a char-
acteristic of growing, synthetic organisms.

CHEMICAL STUDIES ON GROWTH.—III. THE OCCUR-
RENCE OF GLYCOGEN IN THE EMBRYO PIG.!

By LAFAYETTE B. MENDEL anp CHARLES S. LEAVENWORTH.
[From the Sheffield Laboratory of Physiological Chemistry, Vale University.]

ie summarizing the results of his earliest classic investigations upon
the occurrence of glycogen in embryonic tissues, Claude Bernard
wrote, in 1859: ‘Il est permis de penser que chez le foetus cette ma-
tiére glycogéne a un réle important 4 remplir dans le développement
organique.” ? Since then a peculiar importance has been assigned
by many writers to the existence of glycogen in growing organs and
tissues. This view has been supported by the apparent richness
of these parts in the distinctive storage carbohydrate. Indeed, it
has been assumed quite generally that the embryonic structures have
a proportionately larger supply of glycogen than is the case in the
fully grown organism. The presence of glycogen in rapidly growing
neoplasms, especially such as partake of an embryonic character,
lends additional support to the suggestion that this carbohydrate
plays a characteristic part in the phenomena of growth. In his
‘Specielle Physiologie des Embryo,” Preyer expressed this belief, in
1885, in these words: “ Vielmehr ist es wahrscheinlich, dass alles junge
Protoplasma Glykogen bildet und dass Leukocyten es dahin bringen,
wo nicht schon die noch nicht differenzirten embryonalen Zellen es
erzeugt haben. . . . Jedenfalls gehort diese stickstofffreie Verbindung
zu denen, welche im Fotus selbst entweder ihrer ganzen Menge nach
oder zum grossen Theil entstehen und vergehen. Das Vogelei enthilt

1 This research was conducted with the aid of a grant from the Carnegie Institu-
tion of Washington.

2? BERNARD, CLAUDE: Journal de la physiologie de l’-homme et des animaux,
1859, li, p. 336. Many years later BERNARD wrote: “From all these facts and
all these examples, especially from those which relate to the embryo, we should
conclude that the amylaceous material, in animals as well as in vegetables, is in-
dispensable to the histological synthesis, and that its presence in certain tissues is
related to the evolution of the cellular elements which compose them.” (Lecons sur
les phénoménes de Ja vie communs aux animaux et aux végétaux, 1879, ii, p. 80).

117

118 Lafayette B. Mendel and Charles S. Leavenworth.

kein Glykogen, der ganz junge Embryo gibt aber bereits die charac-
teristische Jod-Reaktion” (p. 272). In 1898 Schafer wrote in his
Text Book of Physiology: “In the embryo glycogen is much more
widely distributed and occurs in much larger proportion than after
birth, especially in the developing muscles. At this time the liver
may contain very little” (i, p. 918).

So long as the purely biochemical relationships of glycogen were
not clearly appreciated, it was inevitable that its exact chemical
function in the embryonic cells and tissues should be obscure. An
extreme view deprived glycogen of any histogenetic significance
whatever; it was looked upon as ‘a by-product in the splitting of
complex albumens to build up the tissues,’ —a product of disinte-
grative metabolism. In distinct contrast with such an opinion, gly-
cogén has been classed as an “anaplastic ” rather than ‘‘ kataplastic”
substance, stored like fats and used functionally in the growth of the
organism.! One may think of the substance in this connection either
as a stored material, furnishing energy as it is required, or as a com-
pound, entering into the protoplasm in some more intimate way as an
integral part of the differentiating bioplasm. Creighton? has ven-
tured the following interpretation: “ Briefly expressed, the formative
property of glycogen is analogous to or parallel with that of hamo-
globin. . . . I have been led to the conclusion that glycogen plays the
part of a carrier to the tissues, that it contributes somewhat to the
building up without losing its own molecular identity, that it is pres-
ent at the formation of tissues, and employed therein without becom-
ing part of them, and that it acts thus, in some cases as the precursor
or deputy of hemoglobin, and until such time as the vascularity of
the part is sufficiently advanced; in other cases as the substitute of
hemoglobin from first to last, —in those tissues which are built up
in whole or in part without direct access of blood. ... The pla-
cental and amniotic glycogen I believe to have a different meaning
from that of the embryo itself, to mean, in fact, the throwing down of
a formative medium no longer needed.”

The older literature on glycogen has lately been reviewed with
considerable care. A few careful microscopic observations have

" PREYER: Specielle Physiologie des Embryo, 1885, p. 272.

* CREIGHTON : Microscopic researches on the formative property of glycogen,
1896, i, pp. 8-9. :

® Cf CREMER: Ergebnisse der Physiologie, 1902, i, 1, p. 803; PFLUGER:
Archiv fiir die gesammte Physiologie, 1903, xcvi, p. 1; RICHET’s Dictionnaire de

Chemical Studies on Growth. 119

more lately been made on pig embryos of 15 and 50 mm. by Gierke.!
They confirm the finding that not all embryonic cells contain glyco-
gen. The main store of embryonic glycogen is noted microscopically
to be in the muscle and cartilage. The liver is devoid of glycogen,
as is the nervous system. Lubarsch? has also arrived at similar con-
clusions. He points out especially the fact that the organ particularly
prominent in relation to glycogen in extra-uterine life, namely, the
liver, is glycogen-free in embryo pigs at an age when many other tis-
sues show the carbohydrate in abundance. A review of the litera-
ture on the glycogen content of embryonic organs leads Lubarsch to
summarize in part:

1. The glycogen content varies with the age and species of the embryo.

2. Most epidermal epithelia, striated muscles and cartilages uniformly contain
glycogen.

3. Glycogen is uniformly absent in the blood, spleen, connective tissues, bones,
and nervous substance in all embryonic stages.

Gierke and Lubarsch agree in emphasizing the fact that it is not
always the most rapidly developing cells which tend to be richest in
glycogen.

Statements like those just quoted are based in large part upon
microscopic observations made on tissues stained by the iodine
method. The micro-chemical test for glycogen is, however, one
which demands most careful application in order to give a conclusive
answer regarding the exact distribution and identification of glyco-
gen; and its limitations are even more obvious when the quantitative
aspects of the subject are under consideration. A corroborative
study of the distribution of the carbohydrate by adequate quantitative
methods seems almost imperative before any far-reaching conclusions
can be reached. Some of the earlier investigators, notably Claude
Bernard, actually isolated glycogen from the tissues. The inadequacy
of the earlier extraction methods has repeatedly been pointed out by
Pfliiger,* who has found that small quantities may easily be overlooked
unless the tissues are completely dissolved in a hot solution of potas-
physiologie: article Glycogéne,; also GIERKE: ZIEGLER’S Beitraége zur patholo-
gischen Anatomie, 1905, xxxvii, p. 502 (literature on microchemical studies on
glycogen).

1 GIERKE: Loe. cét., p. 512.

2 Luparscu: Archiv fiir pathologische Anatomie, 1906, clxxxiii, p. 192.

8 Cf. also the criticism of ADAMOFF : Zeitschrift fiir Biologie, 1905, xlvi, p. 283.
4 PrLUGerR: Archiv fiir die gesammte Physiologie, 1904, cli, p. 305.

120 Lafayette B. Mendel and Charles S. Leavenworth.

sium hydrate, and glycogen precipitated by means of alcohol aceord-
ing to his method.

Under Leon Asher’s guidance, Adamoff! has investigated the
quantitative occurrence of glycogen in embryonic life, using the
Pfliiger method. The food factor in connection with the nutritive
condition of the mother, —a feature which has repeatedly been dis-
cussed in relation to mammaltan embryos — was eliminated in one
series of experiments by using chicks, which develop independent of
a variable food supply. Newly born rabbits were also analyzed, as
well as the livers of human foetuses prematurely delivered. It was
found that chicks which have just left the shell contain, at most, in-
significant traces of glycogen. After the fourth day, when they have
utilized the ingested egg residues and are fed, the content of glyco-
gen increases. Newly born rabbits contain about 0.4 per cent of
glycogen; compared with well-nourished, full-grown dogs, this quan-
tity is rather small. The glycogen in the human liver at a late foetal
period does not exceed that of an adult unfed animal in quantity.
From these facts the conclusion is reached that abundance of glyco-
gen is not a characteristic feature of embryonic organs. Energy of
growth and content of glycogen bear no direct relation to each
other.

The paucity of glycogen in certain specific embryonic tissues was
pointed out by Claude Bernard.2 He failed to find it in the liver in
early embryonic life, and notes (p. 335) that this exception deserves
special mention, because of the peculiar rdle which the organ plays
in glycogen metabolism in adult life. Not until Jater in the develop-
ment of the embryo, toward the middle of intra-uterine life, did Ber-
nard find evidence of a glycogenic function in the liver. Paschutin 8
is reported to have missed glycogen in the liver of the embryo calf
before the period when a length of 400 mm. is reached. In contrast
with the small quanties of glycogen found in the liver of the newly
born animal by previous investigators, Demant‘ reported as much as
II per cent in the livers of newly born dogs. He stated that the
quantity speedily diminishes immediately after birth. The analyses
were made by Briicke’s method.

' ADAMOFF: Zeitschrift fiir Biologie, 1905, xlvi, p. 281.

* BERNARD, CLAUDE: Journal de la physiologie, 1859, ii, p. 326.

®* PASCHUTIN, quoted by DEMANT: Zeitschrift fiir physiologische Chemie,
1887, xi, p. 142.

* DeMANT: Loc. cit. The conflicting data of others are there reviewed.

Chemical Studies on Growth. 121

Pfliiger! has reinvestigated the occurrence of glycogen in the em-
bryo liver by means of his improved method. He reports the finding
of glycogen in all the embryo livers examined. These were obtained
from the cow, sheep, and pig, the smallest embryo (calf) measuring
130 mm. The proof was in some cases qualitative only, owing to the
mere traces of glycogen present. Pfliiger’s view is best summarized
in his own words: ‘“ Die wichtigste Thatsachenreihe besteht darin,
dass die embryonale Leber ganz ausserordentlich grosse Schwank-
» ungen im Glykogengehalt darbietet, so dass bald reichliche Mengen,
bald nur Spuren gefunden werden, wahrend die Muskeln stets
betrachtliche Vorrathe an Glykogen darbieten. Es ist genau dasselbe
Verhaltniss, wie ich es bei dem erwachsenen Pferde gefunden habe,
und hier liegt die Ursache unzweifelhaft in dem Ernahrungszustande
des Thieres. Die Leber ist eine grosse Vorrathskammer, welche bei
guter Nahrungszufuhr Massen von Glykogen aufspeichert, welche zur
Zeit des Mangels mit grésster Freigebigkeit an die verschiedenen
Organe des KG6rpers abgegeben werden. Es ist etwas ganz Ge-
wohnliches, dass die Muskeln der geschlachteten Pferde 2% Gly-
kogen, die Leber aber nur einige Zehntel Procent enthalt. Es liegt
deshalb nahe, anzunehmen, dass es sich beim Embryo genau ebenso
verhalt, weil auch seine Leber die uneigenniitzige Vorrathskammer
zur Ernahrung des iibrigen K6rpers darstellt.”2_ One must not draw
the conclusion that in the earlier embryonic period the liver has a
behavior different from that of later life, after birth. The low content
of glycogen in the foetal liver is ascribed by Pfliiger to the deficient
feeding of the mother animal; and he suggests that the animals killed
for sale which furnish the embryos for analytical study often fail to
receive food before being slaughtered.

Since our own experiments were completed ? Lochhead and Cramer #
have made a report on the glycogen metabolism of the foetus. They
completed quantitative estimations of the carbohydrate in the placentae,
foetal liver, and remainder of the foetal body in an age series of preg-
nant rabbits from the eighteenth day to the end of gestation. ‘The
results show that at the earlier dates the maternal placenta possesses

1 PFLUGER: Archiv fiir die gesammte Physiologie, 1903, xcv, p. 19; 1904, cil,
P- 395.

2 [bid., 1904, cii, p. 307.

8 Some of the results were briefly presented to the section in physiology, Brit-
ish Medical Association, at Toronto, 1906. Cf MENDEL: British medical journal,
1906, p. 1787 (December).

* LocHHEAD and CRAMER: Journal of physiology, 1907, xxxv, p. xi.

122 Lafayette B. Mendel and Charles Ss Leavenworth.

a considerable store of glycogen, its percentage amount being quite
comparable with that of a normal adult liver. It remains constant until
the twenty-fourth day, and then a distinct and progressive decrease
occurs each day till the end of gestation. . . . The results present a
striking relation to those obtained from the analyses of foetal livers.
In the latter, though definite amounts of glycogen are found at dates
when histological examination proves negative, the percentage is very
low up to the twenty-fifth day, when for the first time it rises above
that of the rest of the fcetal bodies. This, therefore, represents the date
at which the liver assumes its adult glycogenic function. It will be
noted that the time coincides accurately with the beginning of the
decrease in the placental glycogen. After this date the percentage
of liver glycogen rapidly increases, but even at birth it falls short
of the average found in the liver of an adult rabbit.

“Neither the placental store of glycogen nor that of the fcetal
liver is affected by feeding the animals on a diet rich in carbo-
hydrates.

“ There is a marked correspondence between the foetal weight
and the percentage amount of feetal glycogen, suggesting that a
close relation exists between the growth of the foetus and the
glycogen metabolism.”

EXPERIMENTAL PART.

We have made a large number of analyses in order to determine the
distribution of glycogen in the embryo pig, especially at the earlier
ages. The method followed was that described by Pfliiger! for em-
bryonic tissue. We have found the long heating with alkali, recom-
mended by him, necessary to avoid the difficulties presented by
this kind of tissue. The precautions more recently suggested 2 were
also carefully observed. The measurements of the embryos given
below were taken from the crown of the head to the root of the tail.
Embryos larger than 230 mm.are rarely available here. The material
was brought directly from the slaughter-house and examined at once.
(ee Table I, page 128;)

Glycogen in the entire embryo. — Schéndorff? found variations of
0.7-3-7 per cent in the glycogen content of adult dogs. Several
factors may contribute to the comparatively low figures reported

' PFLUGER: Archiv fiir die gesammte physiologie, 1904, cii, pp. 307-308.
? Ibid., 1906, cxi, p. 307.
* SCHONDORFF : Archiv fiir die gesammte Physiologie, 1903, xcix, p. 191.

Chemical Studies on Growth. ie

here for the embryo pig, foremost among which is the higher water
content of the embryonic tissues. There is a progressive increase
in the glycogen, but the quantities in any event give no occasion to
assume a special richness in this carbohydrate.

TABLE I.

Glycogen content.

Estimated Number of Total
age. embryos used. weight.

Total tissue. Per embryo.

days. * per cent. gm.
oH 0.25 0.022

54 0.37 0.107
62 0.33 0.163
71 0.50 0.601
74 0.56 0.771
84 0.57 1.521
90 0.69 3.339

Glycogen in the embryo liver. — In the livers of embryos ranging in
size from 85 to 230 mm. we have uniformly failed to detect glycogen,
2. é., to obtain any reduction of alkaline copper solutions with the
hydrolysis products obtained by Pfliiger’s method, using about 100
gm. of liver in each trial. No color test was applied; nor have we
attempted any microchemical examination. It is noteworthy that
Pfliiger! obtained so little glycogen from 62 gm. of liver from six pig
embryos of an age of two months that the iodine test alone was pos-
itive. Our failure to demonstrate glycogen in any measurable amount
in the embryo liver can scarcely be due to any loss occasioned by
long heating with alkali; for in control trials, carried out with
embryo muscle heated during long and shorter periods, no loss was
observed by prolonged heating when undue concentration of the
alkali was carefully avoided.

Glycogen in embryo nervous tissues. — Claude Bernard failed to
obtain glycogen from the nervous or osseous tissues of the embryo,
and it is not ordinarily included among the constituents of adult
nervous tissues. Contrary to the opinion of Gierke and of Lubarsch,

1 PFLUGER: Archiv fiir die gesammte Physiologie, 1903, xcv, p. 20.

124 Lafayette B. Mendel and Charles S. Leavenworth.

Gage! has recently expressed the conclusion that at some period in
the growth of nervous tissue glycogen plays as important a part as
in the other tissues of the body. “A series of pig embryos from 7
to 70 mm. were sectioned, and plentiful glycogen was found in the
cells of the dorsal ganglia in embryos from 7 to 20 mm. In the
20 mm. embryo glycogen was also present in the developing nerve
trunks, and in older embryos it nearly or quite disappeared from
the ganglion cells, but became exceedingly abundant in the nerve
trunks.”

Our observations were made upon the brain substance of embryos
85 to 230 mm. in length. The material from the different sizes was
examined in lots of about 100 gm. each, with wxzformly negative
results.

Glycogen in the embryo muscle and skeletal structure. — The heads
were removed and the bodies completely eviscerated. The brain and
liver — both deficient in glycogen — were thereby eliminated. Cor-
responding with this fact the percentage of glycogen in the bulk of
the residual tissue structures appears somewhat larger than that of
the entire body reported above.

TABLE II.

Number of Weight of Glycogen
embryos used. | tissues used. found.

per cent.

gm.
12 95 0.46
115 0.48
102 0.47
115 0.46
92 0.73
87 0.33
70 0.44
0.42

0.58

ibaa

1.10

1 GAGE: Science, 1906, xxiv, September 7.

~

Chemical Studies on Growth. 125

These data indicate a fairly uniform content of glycogen in the em-
bryonic muscle mass at definite ages when the inevitable variations
in such material are borne in mind. That it belongs primarily to the
non-osseous structures is shown in the experiments which follow.

Glycogen in embryo skeletal structures. — Creighton! has enun-
ciated the principle that “those cartilages which are destined to
continue as such through life, including the septum varium, the ensi-
form, etc., are, generally speaking, without glycogen in the foetus.
. . . On the other hand, the embryonic cartilages in which one finds
uniformly a large amount of glycogen are those destined to an early
transformation or absorption. ...In the cartilages of the long
bones, vertebre, etc., glycogen first appears exactly in the spots that
afterwards become ‘centres of ossification.’ ”

We have isolated the skeletal structures in small embryos as care-
fully ds the texture of the growing osseous tissues will allow, and
analyzed them. The skull was not included.

TABLE III.

Number of Weight of
embryos used. | material used.

Glycogen.

per cent.
none

0.10
0.54
0.40
0.40

Conclusions. — The preceding experiments cannot be interpreted as
giving evidence that a large glycogen content is a characteristic of
embryonic structures or developing tissues; for the tissues, examined
by adequate methods, show no unusual richness in this carbohydrate.
The distribution is not markedly different from what pertains in the
adult animal, except that the liver does not assume its glycogen-storing
function early, at least in the pig. This conclusion seems as reason-
able, in the absence of direct contradictory evidence, as to attribute
the uniformly noted poverty in hepatic glycogen entirely to the defi-

? CREIGHTON : Microscopic researches on the formative property of glycogen,
1896, i, pp. 80, 81.

126 Lafayette B. Mendel and Charles S. Leavenworth.

cient nutrition of the maternal animal. The experience of Lochhead
and Cramer also favors this view. The metabolism of glycogen in
the embryo is doubtless comparable with its rdle in the nutrition of
the adult; and it seems unnecessary to postulate any special “ for-
mative” property to account for its presence. Glycogen may thus
be regarded simply as a store of nutrient energy rather than as a
peculiar mark of histogenesis. .

SEE INFLUENCE OF ELECTROLYTES AND OF CER-
Dain OLTHEK CONDITIONS ON THE -OSMOTIC
PebsoURE OF COLLOIDAL. SOLUTIONS.

BY Ke 5.) LILLIE.
[From the Physiological Laboratory of the Fohns Hopkins University.|

I. INTRODUCTORY AND CRITICAL.

T is somewhat remarkable that so much disagreement has existed

and still appears to exist in reference to the osmotic properties
of colloidal solutions. Graham! in 1861 made the explicit asser-
tion, “colloid bodies in general are highly osmotic.” The fact is
indeed easily observed that colloidal solutions placed in a dialyser
absorb water often rapidly and in considerable quantity, — an obser-
vation implying the existence of osmotic pressure, and the one on
which Graham’s statement was based. It is true that in his time
no definite theory or even conception of an osmotic pressure existed,
and that the phenomenon of endosmose was destined to remain only
imperfectly intelligible until many years later, when the work of
Pfeffer and van t’ Hoff laid the foundation for an exact theory of
solution. Yet the early recognized fact that colloids have a “high
endosmotic equivalent ” in reality implies the existence of a certain
osmotic pressure. It is actually easier to demonstrate such pressure
in colloidal than in crystalloidal solutions, on account of the ease
with which membranes impermeable to colloids may be prepared.
Why, then, are there still to be found investigators who deny that
subStances in true colloidal solution exert osmotic pressure?

The chief reason for this state of uncertainty has been the diffi-
culty of freeing colloids from adhering crystalloid substances. The
purest preparations of proteids always contain a certain proportion
of inorganic ash; and colloids, in general, absorb and retain with
great tenacity foreign substances present in their solutions. The
observed pressures or depressions of freezing-point are usually very

1 GRAHAM: Philosophical transactions, 1861, cli, p. 183.
127

128 R. S: Laliie

low and most readily accounted for as due, not to the colloid, but
to the associated impurities. Ina recent investigation, conducted
with great care, Reid! found that proteids purified by repeated
crystallization, re-solution, and re-crystallization yielded in many
cases solutions having no appreciable osmotic pressure. The
natural inference — if these results are accepted as correct — is that
the pressures observed in determinations conducted without such
precautions are due, not to the colloid, but to the impurities present
in solution with it; and such, in fact, has been Reid’s inference
from his experiments.

The observed experimental results may, however, be attributed to
other causes; such treatment, as Moore and Roaf? have urged in
criticism of Reid’s conclusions, certainly alters the state of aggre-
gation of the proteid, and conceivably in such a manner as to give
solutions of too low pressure for determination by mercury manom-
eters. The experiments of Moore and Roaf, and my own which
follow, show clearly that the osmotic pressure of a colloid changes
with its state of aggregation; and I shal! give below measurements
in which a great reduction of the normal pressure is produced by
addition of small quantities of salt to a proteid solution. This
reduction of pressure can, of course, be carried much further with-
out producing precipitation; and a state of aggregation gained in
this way tends to be retained, by a kind of hysteresis, after the re-
moval of the conditions that produced it; so that a more or less
permanent condition of reduced osmotic pressure, due to treatment
with concentrated salt solutions, is certainly a possibility.

The other chief series of facts frequently adduced as showing that
colloids have either a negligibly small osmotic pressure or none
at all, consist in demonstrations that the freezing and boiling points
of pure solutions, as determined with the Beckmann apparatus,
are practically identical with those of the pure solvent; this, however,
as both Starling ®and Moore and Roaf have pointed out, merely indi-
cates that the osmotic pressure cannot exceed a certain upper limit;
the method is, in fact, incapable of demonstrating with any certainty
pressures of less than the very considerable value of 40 to 50 mm.
Hg. Hence such determinations merely show that the osmotic

' REID: Journal of physiology, 1904, xxxi, p. 438.

? Moore and RoaF: Biochemical journal, 1906, ii, p. 34.
* STARLING: Journal of physiology, 1899, xxiv, p. 316.
* Moore and RoaF: Loe, cit.

The Lnfluence of Electrolytes. 129

pressure of such solutions is small, and have no bearing in the
question of its existence or non-existence.

Moore and Roaf have urged another objection to the frequently
employed method of reasoning which ascribes directly observed
osmotic pressures to the salts or other crystalloids associated with
the colloid. They point out — what ought to have been self-evident
—that such substarces cannot possibly produce permanent pressures
in osmometers having parchment paper or gelatine membranes, since
all crystalloids readily traverse such membranes. This simple
reasoning seems perfectly conclusive; substances giving permanent
osmotic pressures with such membranes must in fact be in colloidal
solution; it is at least impossible that they should be the simple
crystalloidal substances to which the pressures have usually been
ascribed.

What part, then, if any, do the crystalloid substances present in
colloidal solutions play in the production of the observed osmotic
pressures? Moore and Roaf propound the following hypothesis at
some length: that the associated crystalloid substances, while not
directly responsible for the pressure, are nevertheless necessary to
its production; in brief, that in some unexplained manner they confer
on the colloid the power of exerting osmotic pressure, which in their
absence would be altogether lacking. It is difficult to see the justi-
fication of any such view; the fact that the presence of certain
“crystalloid” substances (the term uniformly employed by Moore and
Roaf) has its influence on the osmotic pressure of the solution by
no means implies that such substances coxdztion the existence of the
pressure, and that with their complete removal the osmotic proper-
ties of the solution would vanish. Prolonged dialysis of albumin
solutions certainly does not produce such a proportionate change in
osmotic pressure as the acceptance of such a view would lead us to
expect, although by far the greater portion of the crystalloids may
be thus removed. The view seems to imply a belief in the existence
of a fundamental difference between the colloidal and the crystal-
loidal states of matter, the latter alone being capable of independ-
ently maintaining a state of solution and of exerting osmotic pressure.
But it will, I think, be generally agreed that the progress of research
affords less and less reason for drawing the sharp distinction between
colloids and crystalloids that it was once the custom to observe —
possibly in pursuance of the example and precept of Graham himself,
with his comparison of the two states to two different worlds of

130 ROS

material, differing as do organized and unorganized beings.! There
are at least a few colloidal solutions that can be prepared free from
admixture with crystalloids, namely, the metal hydrosols, prepared
by fine subdivision of the pure metal in pure water. The hydrosols
of platinum and gold are, in fact, notoriously the more stable the
greater the freedom from electrolytes,” and if such solutions are
stable, they certainly exert osmotic pressure. It seems at times to
be forgotten that the very conception of solution implies the exist-
ence of such pressure, since the primary criterion of this condition
is a persistent and self-conserving homogeneity of composition.
This signifies that inequalities of concentration, however occasioned,
are equalized by diffusion; hence, if a partition permeable to solvent
but not to solute is interposed between two unequally concentrated
portions of a solution, the diffusing particles of solute neces-
sarily exert pressure on this partition. To imply otherwise would
be to deny that the property of diffusion (and hence of the dependent
homogeneity of composition) is essential to a solution. In view of
such considerations, the inference seems clearly unavoidable that
since colloids are capable of independent solution, they must be
capable also of independently exerting osmotic pressure; and we
may conclude that though the presence of electrolytes may alter the
pressure, they are not necessary to its production.

The readiness to attribute the state of solution of colloids to the
action of electrolytes seems to have had its origin in the conception,
due largely to the work of Hardy,? that the stability of such solu-
tions depends on the charge carried by the colloid particles. The
function of the electrolyte is to provide the particles with the neces-
sary charge. Ions precipitate colloids by neutralizing the charge;
the increased surface tension of the particles due to removal of the
potential difference between particle and medium is, according to
Bredig,* the direct cause of the separation of the colloid from solu-
tion. On the whole, subsequent investigation has led much support
to these ideas; the electrical condition of the colloid is undoubtedly
a factor in the stability of its solutions. Billitzer,® however, has

1 GRAHAM: Loc. cit., p. 220.

2 Cf. BrepIG: Anorganische Fermente, Leipzig, tgor, p. 28.

8 Harpy: Proceedings of the Royal Society, 1g00, Ixvi, p. 110; Journal of
physiology, 1899, xxiv, p. 288.

* BREDIG: Loc. cét.

° Cf. BILLirzER: Sitzungsberichte der kaiserlichen Akademie der Wissen-
schaften, Wien, math.-naturw. Klasse, 1904, cxiii, 7te Hft., p. 1159.

The Influence of Electrolytes. 13

shown that isoelectric point and precipitation point need not neces-
sarily coincide; and Pauli! has shown that in stable colloidal solu-
tions of certain proteids after prolonged dialysis the particles
appear uncharged. The charge, therefore, cannot be considered
the essential feature in solutions of colloids any more than in those
of crystalloids, where both electrolytes and non-electrolytes form
equally stable solutions.

Hence also the osmotic pressure of colloidal solutions cannot be
explained on the basis of the mutual repulsion of charged particles.
It is possible that a small portion of the pressure may be thus
accounted for, as pointed out by both Bredig? and Billitzer;# but
the greater portion (apart from that due to free crystalloid impuri-
ties) must be due to some other condition, — possibly, as both the
above authors suggest, to the adsorption of the solvent at the greatly
developed surface of the colloid. Nevertheless, we still find
attempts to explain the pressure as dependent on the electrified
condition of the particles. Thus, in the important and interesting
recent paper of Duclaux,* describing his extensive experiments
with inorganic colloids, we find the state of solution or stability of
the colloid to be ascribed to the so-called “active part” of the col-
loidal complex, that is, the portion of electrolyte associated with
the colloidal substance (as Fe,Cl, with ferric hydroxide) which by
its ionization imparts the charge to the particle and so conditions
its electrical properties. He finds a decrease in stability with
decrease in this “active’’ proportion of electrclyte, —z. ¢, addition
of less salt is then required for precipitation, — and correlatively a
decrease in the maximum osmotic pressure which the solution can
then exert (or its maximum concentration) without solidifying. Thus
a ferric hydroxide solution in which the colloid has an approximate
composition Fe,Cl, + 40 Fe,O, can be brought to a concentration
corresponding to an osmotic pressure of nearly 3 metres of
water (ca. 220 mm. Hg) before solidifying; while, after removal of
the “active part” to a proportion of 1 Fe,Cl, + 800 Fe,O,, this
“limiting osmotic pressure” proved only 10 cm. of water; the
corresponding concentrations were 15 per cent and 2 percent. The

1 PAULI: Beitrage zur chemischen Physiologie und Pathologie, 1906, vii, p. 531.
Cf. also: Naturwissenschaftliche Rundschau, 1906, xxi, p. 3.

2 BREDIG: Loc. cit.

8 BILLITZER: Zeitschrift fiir physikalische Chemie, 1903, xlv., p. 307.

* DucLaAux: Journal de chimie physique, 1907, v, p. 29.

132 Y ems Be be NEE

decrease of both stability and maximum osmotic pressure with
removal of the electrolytic part of the colloid leads him to refer
both properties to the electrical condition of the colloid particles or
“micella.” The osmotic pressure and the movement of the micellz
(Brownian movement) he ascribes tq the electrostatic repulsions of
the charged micelle, “both mutual and with those of the inter-
micellar liquid." He finds that the charge on the particles increases
with the “active part’ of the colloid; hence their electrical repul-
sion and Brownian movement also increase, — thus being explained —
the fact that solutions of high osmotic pressure can be prepared only
from colloids with a large “active part.”

Duclaux’s view is similar to that of Moore and Roaf, in that both
ultimately refer the osmotic pressure of the colloid to the associated
electrolytes. But, as already urged, certain colloidal solutions con-
tain no electrolytes and others appear to have little or no charge;
and on the above view it would be necessary to regard these two
types of solution as quite distinct in nature from the usual charged
type. The most natural means of avoiding this difficulty would
seem to be the conclusion that there is no essential difference be-
tween the conditions of solution of crystalloids and colloids; that
just as the former comprise both electrolytes and non-electrolytes
with all gradations between the two, so also colloids may be
stronger or weaker electrolytes (z. ¢., may liberate ions in varying
degree, and so acquire charges), or in some cases may be practically
uncharged. The general relations between solute and solvent must,
then, be referred to other conditions; and the suggestion that these
relations are of the kind classed broadly under “adsorption phe-
nomena” seems most worthy of consideration.

In view of the present tendency to regard the attraction between
solute and solvent as of the same general nature as true chemical
affinity, it is interesting to note that solutions show certain analo-
cies to the class of combinations known as “adsorption compounds.”
These last may vary continuously in composition, just as do solu-
tions;? moreover, if there is an adsorption between solute and sol-
vent, the distinctive characteristic of solutions becomes intelligible,

1 Cf. the recent work of H. C. Jones and collaborators, ‘“ Hydrates in aqueous
solution,” the Carnegie Institution, Washington, 1907.

? Cf. v. BEMMELEN’S various papers on the hydrates of various metallic oxides
in Zeitschrift fur anorganische Chemie, especially iv Abhandlung, 1899, xx, p. 185,
and succeeding papers.

The Influence of Electrolytes. 33

namely, the fact that work is needed to separate the two, and hence
that dissolved substances exert osmotic pressure and lessen the
concentration of the vapor phase of the solvent. Certain similari-
ties between the phenomena of osmotic pressure and those of simple
swelling of colloids like gelatine or agar in water afford further
indication that the two processes are fundamentally alike. Since
the swelling of solid gelatine sheets in water is quite generally
referred to adsorption or “imbibition"’ of the liquid by the col-
loid particles, it is difficult to imagine why such action should not
be regarded as the effective factor causing absorption of water in
the fluid mixtures of the two substances, 7. ¢., solutions; the typical
absorption-pressure or osmotic pressure of solutions would thus be
accounted for. In point of fact, the respective action of electrolytes
on swelling processes of solid colloids, and on the osmotic pressure
of solutions of the same colloids, are so nearly identical (as I shall
show below) that the conclusion seems unavoidable that only one
fundamental process is concerned in both. Finally, if solution is
of the same nature as swelling action, and if the latter is due to an
adsorption, we are led to infer, since no sharp distinction can be
drawn between colloids and crystalloids, that osmotic effects in

general are due to attractive or adsorptive relations between solute
and solvent.!

II. EXPERIMENTAL METHOD.

The following experiments have been performed chiefly on solu-
tions of egg albumin and gelatine; a few inorganic colloids, as
ferric hydroxide, arsenious sulphide, and shellac suspensions, have
been used in some experiments. I have, throughout, made use of
a very simple type of osmometer, consisting of a flask-shaped sac of
nitro-cellulose (celloidin or gun-cotton) provided with a perforated
rubber cork, through which passes a narrow vertical glass tube, and
immersed in a fluid contained in a battery jar covered with a glass
plate. The flask-shaped membrane is made by coating the interior
of a 50c.c. flask with a 10 per cent solution of nitro-cellulose in
equal parts of alcohol and ether, draining off the superfluous solu-

1 Compare ARMSTRONG: ‘The origin of osmotic effects,’ Proceedings of the
Royal Society, Series A, 1906, Ixxviii, p. 264. ARMSTRONG refers osmotic effects
ultimately to an attraction between the molecules of solute and those of solvent simi-
lar to that which the water molecules themselves have for one another. He quotes
v. LAAR (Amsterdam Royal Academy, 1906) as giving expression to similar views.

134 ORES Le

o

tion, and then removing the solvent by passing a current of air, and
bathing in hot water. The membrane may then easily be withdrawn
from the interior of the flask. It retains the shape of the flask per-
fectly, is tough, only slightly extensible, and, if of the proper thick-
ness, almost impermeable to the proteids employed, while allowing
extremely ready passage to all crystalloids. It thus makes an ex-
cellent dialyser. The procedure 1s briefly as follows: the membrane
is washed in boiling distilled water; the colloidal solution is intro-
duced toa level with the neck, and a well-fitting rubber cork, with
glass manometer tube inserted, is placed in position and tied
securely with a long elastic rubber band which encircles the neck
a large number of times; an absolutely hermetical junction is thus
secured. The band is so adjusted that the fluid within the mem-
brane is forced part way up the manometer tube, preferably to a
height corresponding to the expected pressure reading; the mem-
brane is thus distended from the first, and entrance of water has an
immediate effect in raising the level of the column of fluid. The
membrane with manometer-tube attachment is immersed in a
definite volume of the pure solvent (water or salt-solution, etc.)
contained in a battery jar; the jar is covered with a glass plate
perforated by a small hole through which passes the manometer
tube; the latter is clamped ina vertical position. The purpose of
the glass plate is simply to prevent evaporation.

The advantage of this osmometer is its simplicity and ease of
preparation; by its use it is possible to perform twelve or fifteen
experiments in a day without difficulty; numerous determinations
can thus be made, —a matter of great importance in dealing with
colloidal solutions, as will appear below. Direct reading of the
height of the column of fluid has the advantage of giving more
delicate pressure determinations than those possible with a mercury
manometer; the readings can easily be reduced to millimetres of
mercury when the specific gravity of the solution within the mem-
brane is known. In most instances this may be considered as simple
unity; in any case it can easily be determined. The error intro-
duced by the slight dilution of the solution due to the entrance of a
certain small quantity of solvent is usually so slight as to be negli-
gible; if desired, the necessary correction may be made. The
passage of water through celloidin membranes is very rapid; the
actual upward movement of the column of fluid in the osmometer
tube can easily be seen at first, if the osmotic pressure is high and

The Influence of Electrolytes. 135

the fluid stands low inthe tube. This ready permeability to water
permits equilibrium to be reached within a few hours. The height
of the column then remains stationary with constant temperature, —
at least for several hours. Usually osmometers have been left over
night and readings made the next morning. By the end of the second
day, the height of the column, especially if the pressure is high, may
have fallen a few millimetres; this is due to the fact that the mem-
brane is seldom absolutely impermeable to the prcteid. The column,
however, remains stationary at its maximum height for some hours
before showing perceptible decline.

I have not taken any especial precautions in the matter of stirring;
in a few instances the contents of the osmometer have been stirred
at intervals by slight compression and relaxation of the membrane;
the results after such stirring were in no observable respect different
from usual. With an osmometer having such a relatively large area
of membrane in proportion to the volume of solution, stirring appears
unnecessary, though it may hasten to a certain degree the attainment
of an equilibrium.

Equilibrium, however, is reached very soon in this apparatus ;
and there can be no doubt that a maximum height, recorded after
an immersion of eighteen or twenty-four hours, and remaining sta-
tionary for a number of hours thereafter at constant temperature,
represents a true equilibrium, and the following records have been
made with this understanding. All observations have been made
at room temperature, and no attempt has been made to determine
degree of variation with changes of temperature. Within the range
of room temperatures such variations would be slight, as compared
with those due to other factors considered below (as agitation, rate
of admixture of electrolyte, etc.).

In the following experiments the attempt has been made to de-
termine the comparative influence of various electrolytes and non-
electrolytes on the osmotic pressure of two proteids, gelatine and
egg albumin. The solutions of these colloids have not been es-
pecially purified; the egg albumin was partially freed from globulin
by the dilution with distilled water and filtration; the gelatin
solutions were prepared from commercial sheet-gelatine. So far as
my experience has gone, dialysis does not affect the osmotic proper-
ties of these colloids to any noteworthy degree; if necessary, how-
ever, the following results may be held to apply to solutions of the
above proteids #/ws a certain unknown and small crystalloid content.

136 R: S. Laie

Dialysis, at best, effects only an incomplete removal of crystal-
loids, and it has accordingly been omitted in the present series of
experiments.

The procedure has been as follows: All experiments are performed
in series; in any one series the same colloidal solution (¢ g., 1.5
per cent gelatine) is used. One solution serves as control; to the
others are added definite quantities of the electrolyte (or non-
electrolyte) whose action is to be tested. AI solutions of the series
are brought to equal volumes; the concentration of colloid through-
out a series remains constant, the quantity and nature of the added
electrolyte alone varying. To the distilled water outside the mem-
brane is added the same electrolyte in the same concentration as
within the membrane. The whole system within the osmometer —
contents of membrane and surrounding fluid —thus contains the

given electrolyte in uniform concentration. The pressure observed |

must, under these conditions, be due to the colloid and not to the
substance added (which, moreover, always readily traverses the
membrane). The effect of the substance on the osmotic properties
of the colloid is seen on comparison with the control. The volume
of the colloidal solution within the membrane has been always ap-
proximately 60 c.c.; that of the outer fluid, 420 c.c. ; their relative
volumes have thus been approximately constant throughout.

III. EXPERIMENTAL RESULTS.

A. Action of non-electrolytes on osmotic pressure.— In agreement
with their generally observed negative action in precipitating col-
loids, non-electrolytes are found to exhibit no definite influence on
the osmotic pressure of proteid solutions. The following experi-
ments illustrate (see Table 1).

In each series the pressure readings show a general uniformity,
with certain minor fluctuations. The greatest divergence from the
control pressures is seen in the solutions containing urea; it will
be noted that in the two series of gelatine solutions the pressures
are approximately one fifth higher in the presence of urea than in
the control. It is possible that this variation has some significance;
the osmotic pressure of gelatine is perceptibly increased by the
presence of a very small proportion of alkali (see Table IJ, below),
and the above slight excess of pressure may indicate a partial action
of the urea asa feeble base. In the solutions containing the other

The Influence of Electrolytes. 177

more typical non-electrolytes, the variations from the osmotic pres-
sure of the control are slighter, and no especial significance is to
be attached to them. They are, I believe, to be referred chiefly
to unavoidable variations in the manipulation of the solutions, —
variable changes in the aggregation state being produced by this,

TASB ICEy 1.

Ecc ALBUMIN + NON-ELECTROLYTES.

Series I. Series IT.

if
{|

Solution. Pressure. cp. Solution. Pressure.

|

mm. Hg. | mm. Hg.

1.25 % egg albumin | 22.4 || 1.6 % egg albumin 29.4

“ “ 4% 6sucrose | 21.5 « “ 4 m/6 glycerine 29.5

« « + m/6 dextrose FS. « «© 4 m/6 urea 27.9

GELATINE + NON-ELECTROLYTES.

Series III. Series IV.

Solution. Pressure. Solution. Pressure.

mm. Hg. mm. Hg.

1.25 % gelatine 6.2 1.25 % gelatine 5:5

“ + m 6 sucrose 6.6 « © 1 m/6 dextrose 5.7

“ + m/6 dextrose 5.8 “« 6“ 4+ m/6 glycerine 5.6

“ + m/6 glycerine 5.9 <4. m/6 urea 6.6

“ 4 m/6 urea

and by the process of mixing itself. The number of uncontrolled
variables is always considerable in work with colloidal solutions ;
all influences affecting the aggregation state, such as agitation, rate
of admixture of electrolyte, etc. (see below, page 163), have a certain
influence on the osmotic pressure, so that solutions prepared in as
nearly as possible identical manner will often show slight but un-
mistakable differences in physical properties. This variability is
seen to a still greater degree in the readings, given below, of experi-
ments with electrolytes. The above determinations are sufficient to

138 Ry iS: Late.

show that non-electrolytes have — at least in comparison with elec-
trolytes — little or no influence on the osmotic pressure of the above
colloidal solutions.

B. Action of electrolytes on osmotic pressure. — In contradistinction
to non-electrolytes, all classes of electrolytes produce marked alter-
ation in the osmotic pressure of colloidal solutions, in some cases
depressing, in others increasing, that of the original solution. In
general, the principle holds that any electrolyte which, when added
in sufficient quantity, precipitates the colloidal solution, depresses
the osmotic pressure when added in concentrations less than those
required for precipitation; while an electrolyte that promotes solu-
tion of a colloid increases the latter's osmotic pressure. There is,
in short, a relation between the precipitating or anti-precipitating
action of an electrolyte, and its depressant or augmentative influence
on the pressure of the colloid. This principle, stated thus unequivo-
cally, seems indeed almost self-evident when it is considered that the
osmotic pressure is a direct measure of the work necessary to sepa-
rate solute and solvent, and that the addition of a precipitating salt,
even in quantity insufficient for precipitation, must decrease the work
necessary to effect such separation, since less salt is then required to
complete the process. The first evident effect is naturally a depres-
sion of osmotic pressure. On the other hand, the addition of acid to
a solution of gelatine furthers solution, renders separation more diffi-
cult, and correspondingly increases osmotic pressure; and obviously
the osmotic pressure of a globulin solution is directly dependent on
the presence of salt, — probably increasing up toa certain maximum
as the concentration of salt increases, after which it declines again;
these latter relations, however, have not yet been investigated.

In the following account the effects of addition of acid and alkali
and of various salts will be considered in order.

C. Experiments with acid and alkali. a. Gelatine solutions. —
Addition of small quantities of either acid or alkali to solutions of
gelatine produces a marked increase in osmotic pressure. Two
preliminary experiments gave the following result: 1.5 per cent
gelatine gave a pressure of 7 mm. Hg; the same solution after the
addition of HCl to m/610 concentration gave a pressure of 33.1
mm. Hg; with m/610 KOH the pressure was 22.8 mm. Hg.

The following table gives the record of three series of experi-
ments; 60 c.c. of the colloidal solution was added rapidly and with
constant stirring to the measured quantity of z/10 HCl or KOH

The Lnfiuence of Electrolytes. 139

in a small beaker, and all members of a series were made to equal
volume. The solutions were then introduced into the osmometers.
The following pressures were obtained.

TABLE II.

1.5 % GELATINE + HCl anp KOH.

Series I.
Exp. Electrolyte. mm. Hg. Exp. Electrolyte. mm. Hg.
1 0 (control) 8.2 5 m /1025 HCl 26.5
2 m /3100 HCl 6.8 6 m/770 HCl 32.4
2 m /2050 HC) 1253 7 m/620 HC) 34.9
4 m,1550 HCl 17.9 8 m/412 HCl 39 3
‘ 1.5 % GELATINE + HCl anp KOH.
Series II. Series III.
Exp. Electrolyte. mm. Hg. Exp. Electrolyte. mm. Hg.
1 0 (control) 6.2 1 0 (control) fs
2 m/3100 HC1 6.1 2 8 m/3100 KOH 14.1
3 m /\240 HCl 27.4 3 m/620 KOH 23.7
4 m /620 HCl 33.1 4 m/412 KOH 25.1
5 m /310 HCl 33.2 5 m/310 KOH 29.0

In the presence of either acid or alkali the osmotic pressure of
gelatine thus shows a marked increase, which, within the above
range of concentrations, exhibits a certain proportionality to the
quantity of acid or alkali added. For equivalent concentrations,
acid produces a somewhat greater increase than alkali.

The change in osmotic properties is to be attributed to a finer
subdivision of the colloid particles and a consequent increase in the
surface of interaction between colloid particles and medium. The
change can hardly be attributed to an increase in the molecular
concentration of the solution, due to hydrolytic splitting of the

140 RS Lilhe

gelatine molecule, for in the first place the quantity of acid or
alkali is insufficient to produce, in so short a space of time and at
room temperature, so extensive a chemical transformation as a
fourfold increase in osmotic pressure would imply; again, the
change in osmotic properties occurs ina strictly continuous manner,
and is readily, though gradually, reversible when the acid is re-
moved from the solution by dialysis. The increased pressure may
safely be regarded as due to a simple change in aggregation state.

It is interesting to compare the above results with those described
by Wolfgang Ostwald! on the influence of acids and alkalis on the
swelling of solid gelatine plates immersed in water; acid in low
concentration (7/210) causes a slight diminution in the swelling
capacity; higher concentrations increase the velocity and degree of
swelling to several times the original; with progressively increasing
concentration of acid the degree of swelling increases rapidly up
to a certain optimum concentration (7/26), after which there is a
decline. Alkali produces similar effects, except that there is no
initial decrease. The resemblance to the above results, is striking ;
there is even a slight diminution of the normal osmotic pressure in
presence of low concentrations of acid. I have not yet performed
experiments with higher concentrations of acid; the above figures,
however, show that the increase of pressure is at first rapid, and
then more gradual as the concentration of acid increases. The
absolute concentrations of electrolyte for corresponding effects are
much lower in my own than in Ostwald’s experiments; this is, no
doubt, simply due to the fact that the gelatine is in lower concen-
tration in the solutions than in the solid or swollen form.

In the experiments with salts given below, it will be seen that
in their respective actions on osmotic pressure, alkali salts with
different anions form a series closely similar to that found by
Hofmeister? for their action in swelling gelatine plates. The two
processes — swelling in the solid condition of the colloid, and
absorption of water due to the osmotic pressure of its solutions —
are, in fact, influenced by ions in essentially the same manner; a
clear indication that at bottom both processes are dependent on the
same condition. This agreement confirms the view expressed above,

1 OsTwALp, W.: Archiv fiir die gesammte Physiologie, 1905, cviii, p. 563, and
Cix, p. 277.

2 Cf. HorMEIsTER: Archiv fiir experimentelle Pathologie und Pharmakologie,
1891, Xxviil, p. 210.

The Influence of Electrolytes. 141

that the phenomena of osmotic pressure are due, like those of swell-
ing, to an adsorption of the solvent by the particles of dissolved
substance.

b.- Albumin solutions. —Egg albumin differs from gelatine in
exhibiting no increase of osmotic pressure in the presence of acid
or alkali. The following three series of experiments, made at
different times and with different preparations of albumin, indicate

TABLE III.

Ecc ALBUMIN + ACID AND ALKALI.

Series I. 1.5 % Albumin.

Electrolyte. mm. Hg, Electrolyte.

25.6
20.7
m/1240 HCl W5 | 9 m/620 KOH 20.2

0 (control)

m/3100 HCl

m/3100 KOH

m /1240 KOH

- WwW WD

m/620 HCl 14.1 10 m/412 KOH 180

5 m/412 HCl 20.4 1] m/310 KOH 17.9

m/310 HCl 222

Series II. 1.5 % Albumin.

1 0 (control) 23.7 7 m /3000 KOH 23.8
2 m/3000 HCl Sill 8 m [1200 KOH 23.3
3 m/1200 HCl 16.2 9 m/600 KOH 21.6
4 m/600 HCl 14.1 10 | m/400 KOH 20.7
5 m/400 HCl 20.4 11 m/300 KOH 18.4
6 m/300 HCl 22.6

Series III. 1.25 % Albumin.

1 0 (control) 18.4 5 m/120 KOH 11.5
2 m/120 HC] 1321 6 m/80 KOH 10.1
3 m/80 HCl 10.9 7 m/60 KOH 9.4
4

m/60 HCl 8.6

142 Pen Sz fale.

the general character of the results,— Series I and II with low
concentrations of electrolyte, Series III with higher.

The osmotic pressure of albumin solutions is thus depressed by
both acid and alkali, though to a less degree, within the range of
concentrations employed, than by neutral salts (see below). The
determinations with the lower concentrations of HCl (Series I and
II) show a result that seems curious at first sight, —a maximum of
depression with the two lowest concentrations of HCl (#/3000 and
m/1200); this apparent anomaly is in all likelihood to be explained
on the ground of the presence of a certain variable proportion of
alkali albumin in the egg white employed (obtained from common
shop eggs); solution 2 of Series II showed, in fact, a considerable
clouding on first admixture of the acid; a removal from solution or
approach in this direction, due to partial neutralization, would ac-
count for the initial marked decline in pressure. Addition of
somewhat more acid (7/400 to 7/300) apparently restores the origi-
nal state of solution; further addition of acid produces a steady fall
of pressure, as seen in Series III; such progressive decline would
lead eventually to precipitation. The series with alkali show steady
decline with increase in concentration of alkali.

D. Reversibility of action of acid in gelatine solutions.— As already
mentioned, the withdrawal of the acid from gelatine solutions by
dialysis is followed by a return of the osmotic pressure toward its
original value. There is, however, a marked lagging or “hystere-
sis,” the solution tending to conserve the heightened osmotic
pressure conferred on it by the acid. In other words, as the con-
centration of acid in the colloidal system diminishes, the osmotic
pressure also falls, but at a much slower. rate, so that at any time
the pressure is greater than can be accounted for on the basis of the
acid actually present at that time. The osmotic properties of the
solution can then be explained only by reference to a previous stage
in its history, when the concentration of acid was higher. Thus,
in one experiment, the outer fluid of the osmometer of solution 8 of
Series I (#/412 HCl) was replaced by distilled water. In seven
days, after several changes of water, the pressure had fallen to about
half its original value, z. ¢., to 18.9 mm. Hg, although nearly all
the acid must by this time have been removed. Several days later
the pressure had reached 9.5 mm., which approaches that of the
control. The tendency to retain the aggregation state conferred on
the colloidal particles by the acid, and with it the increased osmotic

The Influence of Electrolytes. 143

pressure, constitutes an interesting case of hysteresis of a kind not
previously described, so far as I am aware. Probably the pressure
would have fallen more rapidly if the solution had been agitated or
the temperature higher. The exact investigation of this class of
phenomena, so interesting to the biologist, is, however, as yet in its
early stages. It is necessary continually to take them into account
in dealing with changes in colloidal solutions, as I shall endeavor
to show in more detail later (see page 163).

£. Experiments with salts.— The addition of salts invariably
produces a fall in the osmotic pressure of the above colloidal solu-
tions. The degree of this depression varies with the concentration
of the added salt, and with the nature of both anion and cation. In
general, neutral salts of the alkali metals produce least depression,
alkali eartks somewhat more, and heavy metals most of all, these
last showing wide variations among themselves. For salts of the
same metal with different anions, chlorides depress to a greater
degree than bromides, and these than iodides, while sulphocyanates
have still less depressant action. The relative action of different
anions will be considered at length below. The degree of hydroly-
sis is also a factor in the action of any salt, especially in the case of
gelatine, which, as already seen, is especially sensitive to changes
in the acidity or alkalinity of its solutions.

In the following determinations all experiments were made as
follows: 50 c.c. of the colloidal solution —e. g., 1.5 per cent egg
albumin — was poured somewhat rapidly and with constant stirring
into 10 c.c. of the salt solution—e. g. m/2 NaCl—contained in
a beaker; the resulting solution contains 1.25 per cent albumin
and m/12 NaCl. Care was taken to conduct this mixing process
with the greatest possible uniformity throughout; the rate of ad-
mixture of the electrolyte appears to be a factor of no smal] im-
portance in determining the physical properties of the colloidal
system,! and part of the variability in the following determinations
is undoubtedly to be ascribed to variations in the rate of mixture
and in the other manipulation of the solutions before their in-
troduction into the osmometer. Mechanical stirring or agitation,
apart from its influence on the rate of mixture, has in itself an
effect on the aggregation state of the colloid, and so on its osmotic
pressure (see below, page 163). Hence, in experiments of the fol-

1 Cf. FREUNDLICH: Zeitschrift fiir physikalische Chemie, 1903, xliv, p. 129.

144

1.25 % Ecc ALBUMIN + SODIUM AND PorassIuM SALTS.

TABLE IV.

Series I.

0
m/12 NaCl
m/\2 NaBr
m/12 Nal
m/12 NaNOg
m/12 NaCNS
m/12 Na2SO,4

m [24 NaCl

m /24 NaBr
m/24 Nal

m [24 NaNO3
m/24 NaCNS
m/24 NagSO,

STAM WD He

m/48 NaCl
m/48 Nabr
m/48 Nal
m/48 NaNOs
m/48 NaCNS

m/48 NagSO4

Series IT.

0.231
0.213
0.226
0.222
0.245
0.185

ie) SS. aie:

m/12 KCl
m/12 KBr
m/12 KI
m/12 KNOg3
m/12 KCNS
m/12 K2SO, |
Aver. pressure

m /24 KCl
m/24 KBr
m/24 KI
m/24 KNO3
m/24 KCNS
m/24 KySO,
Aver. pressure

Series ‘III.

0.264
0.291
0.289
0.289
0.291
0.245

m/48 KCl
m/48 KBr
m/48 KI
m/48 KNOz
m/48 KCNS
m/48 KySO,4
Aver. pressure

Series IV.

m/96 NaCl
m/96 NaBr

m /96 Nal
m/96 NaNOg3
m/96 NaCNS
m /96 Na,SO,4

m/96 KCI
m/96 KBr
m/96 KI
m/96 KNOg
m/96 KCNS
m/96 KeSO,4
Aver. pressure

The Influence of Electrolytes. 145

lowing kind, uniformity in manipulation should be observed as far
as possible.

In Tables IV to VIII I have given, in addition to the actual
pressure in millimetres of mercury observed in each determination,
the proportion which such pressure bears to that of the pure colloidal
solution (the “control” pressure); this facilitates comparison be-
tween solutions belonging to different series. In Table IV
I have also given, at the foot of each series, the average pres-
sure for all the solutions cf that series (exclusive of the control)
and its proportion to the control pressure; the general manner in
which the pressure varies with the concentration of salt is thus
made clearer by neglecting the variations of the individual salts.

It is to be noted that with decrease in the concentration of salt,
there is a decrease in the proportionate lowering of osmotic pres-
sure, but at a much slower rate; there is also seen a decrease in
the variability of the pressures in the different solutions of a
series, the individual pressure determinations in a series showing
progressively greater uniformity (z. ¢., less average deviation from
the average value for each series) as the concentration of salt
decreases.

The following two series were made with egg albumin in more
concentrated solution, using the same salts (with potassium tartrate
substituted for the sulphate) as in Series II in m/24 concentration.

TABLE V.

Ca. 2.5 % Ecc ALBUMIN + m/24 SODIUM AND POTASSIUM SALTS.

Series I. : Series IT.

Salt. : s : Salt.

ras
1
2
3
4
5
6
7

0 (control)
m/24 KCl
m/24 KBr

0 (control)

m/24 NaCl
m/24 NaBr
m /24 Nal m [24 KI
m /24 NaNO, m /24 KNOg
m/24 NaCNS m/24 KCNS

m [24 K,C4H 40,

146 R. S. Lillie.

In these series osmotic pressure is proportionately less lowered by
this concentration of salt than in the more dilute albumin solu-
tions of Table IV, there being less electrolyte relatively to a
given quantity of the colloid; and the readings with different salts
vary less from an average. Otherwise, the general result is the
same, and the salts range themselves in the same order, tartrate
acting similarly to sulphate.

A number of similar determinations have been made with solu-
tions of gelatine, and the following Tables (VI and VII) give the
results of these.

On inspection of the several tables of pressure determinations, the
following chief uniformities are apparent. There is, in all cases, a
depression of the osmotic pressure of the pure colloidal solution,
and the degree of depression increases with increase in the concen-
tration of the added salt, but evidently at a much slower rate. Since
ionization is almost complete in the salts of Table IV at the dilu-
‘tions employed, it is evident that the ionic concentration decreases
in the four series in a geometrical progression with a factor of ap-
proximately one half; taking the average depressions in the four
series of m/12, m/24, m/48, and 72/96 salt content, we find these to
be respectively about 82 per cent, 78 per cent, 72 per cent, and
64 per cent of the control pressures. So far as any deductions can
be drawn from a series of data manifestly incomplete, the indica-
tions seem to be that doubling the ionic concentration produces an
approximately constant absolute decrease in osmotic pressure, 2. @.,
that the effect would be found in general to vary with the logarithm
of the ionic concentration. Measurements with colloidal solutions
are, however, so variable that a very large number of determinations
would be needed to establish definitely the existence of such a rela-
tion, and it is best to defer for the present any attempt at exact
formulation.

Two chief differences are at once noticeable between the two
proteids. The osmotic pressure of gelatine solutions is much lower
than that of egg-albumin solutions of the same concentration, and a
given concentration of salt produces uniformly a much greater pro-
portionate depression with albumin than with gelatine. Otherwise
the general effects observed are the same in the two series of deter-
minations. The high osmotic pressure of egg-albumin solutions is
a fact previously observed; in the experiments of Moore and Parker!

? Moore and PARKER: This journal, 1902, vii, p. 261.

The Influence of Electrolytes. 147
TABLE Vil:
1.25 % GELATINE + PoTassiuM SALTs.
Series I.

Exp Salt. mm. Hg.| Ppn. || Exp. Salt. mm. Hg. | Ppn.
it 0 (control) 7.9 7 m/24 KBrOg 3.6 0.455
2 m/24 KCl 3.3 OAs lines m/24 KCNS S15 0.474
3 m/24 KBr Sh) 0.468 9 m/24 KCOOCHs fae} 0.367
4 m /24 KI 3.7 0.468 || 10 | m/24 K,SO, 3.4 0.430
5 m /24 KNO; 3.5 0.443 || 11 | m/24 K,C,O,4 3.4 0.430
6 | m/24KCIO; 3.7 | 0.468

Series IT.
1 | 0 (control) 14.4 5 m/24 KNOs 78 | 0.541
2 m/24 KCl 5.6 0.389 6 m/24 KBrOg 8.3 0.576
3 m /24 KBr D2, 0.639 7 m/24 KCNS 11.0 0.764
4 m/24 KI 9.2 0.639 || 8 m /24 K,SO,4 8.1 0.562
Series un,
I 0 (control) 46 5 m/48 KNOg3 2.5 0.543
2 m/48 KCl 2°4 0.521 6 m/48 KC1O3 2.2 0.478
3 m/48 KBr 2.9 0.630 || 7 m /48 KBrOg 2.4 0.521
4 m/48 KI 3.0 0.650 | 8 m/48 KCNS 32, 0.693
Series IV.
1 0 (control) 10.4 4 m/48 KC1Og oy 0.490
2 m/48 KCl ai 0.355 5 m/48 KBrO3 3.8 0.365
3 m /48 KNO3 3.9 0.375 || 6 m /48 NagCO31 5.9 0.567

1 For comparison with neutral salts.

148 R. S. Leite.

TABLE VII.

1.25% GELATINE + ALKALI AND ALKALI-EARTH SALTS.

Series I.

Salt.

0 (control)

m /24 NaCl
m/24 NaBr

m /24 Nal

m /24 NaNOg
m /24 NagSO4

0 (control)
m/48 NaF
m/48 NaCL
m/48 NaBr
m /48 Nal

0 (control)
m/96 NaF
m/96 NaCl
m/96 NaBr

m /96 Nal
m/96 NaNOg
m/96 NaCNS

m [24 KNOg
m/24 KCNS
m [24 KoSO4

m/48 NaNOg3
m/48 NACNS

m /48 NaCOOCH;
m/48 Na,SO,

m /48 Nag citrate

m /96 NaCOOCH,
m/96 NagSO,

m /96 Nag citrate
m /96 MgCl

m/96 CaCl,

m/96 SrCl,

m/96 BaCl,

The Infiuence of Electrolytes. 149

TABLE VIII.

1.25 % Ecc ALBUMIN. SHOWING THE PROPORTIONATE LOWERING OF OSMOTIC
PRESSURE FOR SODIUM AND PoTAssiUM SALTS WITH VARIOUS ANIONS. FIGURES
GIVE THE PROPORTION OF THE CONTROL PRESSURE.

|

0.228 | (no deter-

mination)

0.245 0.185

(2) 0.305

(a) 0.377
m [96 |
Av. 0.341

Average for
all conc.’s

0.258 0.255
Order for Na-salts: SO, > Cl and Br > NO; and I > CNS.
Conc. KCl KBr KI -KNO KCNS

m [12 0.144 - : 0.166 0.211
m [24 0.204 } 0.255 0.264

(2) 0.282
<i (6) 0.283 0.245 0.296

(a) 0.377
Av. 0.341

Average for
all conc.’s 0.243

Order for K-salts: SO, > Cl > NO, and Br >I > CNS.

Other salts with plurivalent anions (citrate, tartrate, phosphate, ferrocyanide) all
showed, in the few experiments tried, marked depressions comparable to or greater
than that of sulphate.

1 Supposing m/12 Na,SO, to produce the same depression as m/24.

KR. So Lehtee

150

SND <I <4 < °O1ID < "ON <10 < "Oa 3 10p10

ns ee

£69'0 S90 +8b'0 6S+'0 0s9'0

9¢S°0 (9)
9TS'0 (2)

06¢'0 (2)
SLt'0 (2)

SLE0 (9)
€rS'0 (2)

9(NO)ea'o SNOM ®O10 8O1IOM ®"ONM IM IDM

8zb'0 asvioaAy

+er'0 (A)
L8¢0 (P)
bbb'O (2)
Sse'0 (9)
1Zs0 (”)

Sp / ue

IDX

"SNO <I < 414 < *O1d-< °ON < 8010 <."0S <10 <"HD00D :29p10

Z9r'0 (2)
1+s'0 (9)

LLYv'O (2)
6£9'°0 (9)
89r'0 (2)

Ler (2)

z9S0 (9) 6£9°0 (9)

O£t'0 (»)

L9¢°0

Yos*’u =| ®HOOOOM! SNOM

8010 ®0O1IOM

‘ANNSSAYY IOYINOD AO NOMAOdONG AAIO sAUNDIY
‘SNOINY INAUAIAIG HLIM SLIVS-Y GNV -VN YOd AYASSAY OILOWSQ JO ONIYAMOT ALVNOMYOdONY

“XI HIAVL

aSVIOAW

68£°0 (9)

‘AINILVIGD % Sol

151

The Influence of Electrolytes.

T1090 ‘SNO !ZL9°0 I '€bS'0 “Id SELb'0 WON !ZLb0 FOS !1bF'0 1D FIITA AIeL WU osye pasn szyes ay} soy sanjea aaoqe ay} [le Jo aBe19AY

‘810000 <I < SND <1 < °ON pure ’os < 19 pue a < og ‘10pi9

LS+'0 805°0 Le9°0 6S5°0 80$°0 9LS°0 $20 FL+'0 bLr'0

ayeayofen | FOS*eN | *HOOOORN | § SNON SONPN IPN gen IOUN AeN

‘SND < °HDOOD < I < 4g pue FO¢g ‘9}eI319 FOI SID < A <"ON :4°PIO

80S'0 26+'0 Sor'0 09+'0 aSVIIAY

€9F'0 (P)

61r'0 (9) Ist'0 (2)
: $2S'0 gss'0 +bt'0 80S'0 26+'0

soso. (2) Ist'0 (9)

9L+'0 (»)

*OdH7EN | VIIDFeN | FOS’eN | *HOOODRN | SNOVN | 9ONEN IVN IO®N

I <1ig < "ON <9 pue fos ‘19pio

Ter'0 c6r'0 L6S'0 osS'0 I¢t0

’Od H°eN ayeIID PEN Fos*en =| §HOOOORN]| SNOPN ®ONRN I®N gen IOeN

152 RK. S: Lethe.

its pressure for a given concentration proved from three to five times
greater than that of the serum proteids. The relatively slight
depressions produced by addition of salt to gelatine solutions indi-
cate that its state of solution is a more stable one than that of egg
albumin.. This is in accord with the general observation of the
relative difficulty of separating gelatine from solution by addition
of salts, or otherwise.

Specific action of different tons.— There is also evident a certain
specificity in the action of each salt of a given metal, indicating in-
equality of influence on the part of the different anions. Thus, in
general,.chlorides depress more than bromides or nitrates, and these
to a somewhat greater degree than iodides; sulphocyanates have.
relatively slight depressant action, while sulphates have more than
any of the salts just named.

In order to gain a more exact and objective estimate of the
relative depressant powers of the various anions, I have tabulated
in Tables VIII and IX the proportionate pressures observed
in all the determinations with salts of sodium and potassium in
various concentrations, and 1.25 per cent solutions of both albumin
and gelatine. Individual variability is so strongly marked in col-
loidal solutions that little reliance can be placed on the order of
action observed in a single series of determinations; thus occa-
sionally a determination with a bromide may show a distinctly
greater depression than the corresponding one with a chloride (as in
Series I of Table IV), although the reverse is the true order. The
errors introduced by uncontrolled fluctuations of this kind are best
avoided by taking the average of a number of determinations.

From the above tables it appears that when the salts of any given
series are arranged in the order of their depressant power, a certain
constancy is seen, which becomes better defined on taking the aver-
ages of aconsiderable number of independent determinations. Thus,
in the series of experiments with gelatine, the averages of all values
for salts with the following anions (for which the greatest number
of determinations were made) give the following order:

Cl (0.441) > SO, (0.472) > NO; (0.473) > Br (0.543) > I (0.572) >
CNS (0.601).

The similar series of averages for the albumin series is for Na- and
K-salts taken together:

The Influence of Electrolytes. 153

SO, (0.224) > Cl(0.250) > Br (0.260) > NO, (0.268) > I (0.277) >
CNS (0.286).

The chief difference between the two proteids is that the deter-
minations with gelatine give sulphate a somewhat less depressant
action than chloride.!_ The mean of the results with the two colloids
gives the following order:

SO, (0.348) and Cl (0.345) > NO; (0.370) > Br (0.401) > I (0.424) >
CNS (0.443).

Sodium fluoride showed, in the two experiments with gelatine solu-
tion, a depression somewhat greater than that of the chloride.
Chlorate appears similar to nitrate. In the case of albumin solu-
tions other salts with plurivalent anions — tartrate, phosphate,
citrate, and ferrocyanide — agreed in the few experiments tried, in
showing well-marked depressant action similar to that of sulphate.
In the case of gelatine the influence of such salts on the neutrality
of the solution is apparently a factor, for phosphate and ferrocyanide
produced less depression than would otherwise have been expected.
Potassium oxalate, in the one experiment tried, gave the same depres-
sion as the corresponding solution of potassium sulphate.

So far as can be judged from data at present insufficient for pur-
poses of very accurate comparison, the anions range themselves,
according to their relative influence on the osmotic pressure of
proteid solutions, in somewhat the following order: sulphate (with
other plurivalent salts: tartrate, citrate, phosphate, ferrocyanide) >
(fluoride and) chloride > nitrate (and chlorate) > bromide > iodide >
sulphocyanate. This order is essentially identical with that found
by Hofmeister and Pauli for the general influence of anions on the
aggregation state of colloids; and the above results afford, in this
sense, a decided confirmation of theirs. They also indicate the
closeness of the relation existing between osmotic eta and ag-
gregation state in colloidal solutions.

Several series of determinations were made with the aim of ascer-
taining the relative influences of lithium, sodium, potassium, and
ammonium on osmotic pressure, the chlorides of these four metals
and in addition the sulphate and sulphocyanate of ammonium being

1 In the experiments of Table XI below with ammonium salts the sulphate

showed uniformly a decidedly greater depressant action than the chloride for solu-
tions of albumin; in gelatine solutions the difference was less decided.

154 RS. Lille.

used. The result shows an approximate equality of action for all
four cations; the anions showed the typical order of depressant
action seen above: SO, > Cl > CNS. _ For purposes of comparison,

TABLE X.

1.25 % GELATINE + SALTS OF ALKALI AND BIVALENT METALS.

Series IL.

Salt.

0 (control) ee m/48 NH,CNS

m/48 LiCl : ) m /48 (NH4)2SO4
m /48 NaCl 4 L m /192 CaCl
m /48 KCl : L m /192 CoCle

m/48 NH,Cl : k m /192 CuCl,

Series II.

0 (control) a m/48 NH,CNS
m/48 LiCl 4 | 0.387 m/48 (NH4),SO,
m/48 NaCl : 0.451 m/48 NagHPO,
m /48 KCl 4 10387 m [48 K,Fe(Cn)<
m/48 NH,Cl 4 | 0.387

Series IIT.

0 (control) : Sievate m/48 NH,CNS
m/48 LiCl ; m/48 (NH4).SO4
m/48 NaCl ; m/48 Na,gHPO,
m/48 KCl : m/48 KgFe(CN)¢
m /48 NH,Cl

several experiments with salts of bivalent metals, and alkali metals
with plurivalent anions were introduced into this series.

The results of these determinations show no noteworthy differ-
ences between the four alkali chlorides; lithium chloride shows on

The Influence of Electrolytes. 155

the average a somewhat less depressant action than the other three,
but the difference is not sufficiently marked to be decisive. Lithium,
according to its general precipitating powers, should depress more

TABLE XI.

Ecc ALBUMIN + SALTS OF ALKALI METALS.

Series I.
1.6 % Albumin.

0 (control)

m/48 LiCl

m/48 NaCl
m/48 KCl

m/48 NH,Cl
m/48 NH,CNS
m /48 (NH4).SO,

Series III.
1.25 % Albumin.

Series II.
1.25 % Albumin.

Salt.

; Salt. a Ppn. ;

0 (control)
m/48 LiCl
m /48 NaCl
m /48 KCl

m/48 NH,Cl
m/48 NH,CNS
m /48(NH4).SO4

Series IV.
1.25 % Albumin.

0 (control) 21.0 asec 0 (control) 19.9

m/% LiCl 6.9 m/48 NaCl 5.7
m/96 NaCl 6.4
m [96 KCl 6.4
m [96 NH,C1l 6.0
m [96 NH,CNS 6.0
m/96(NH4)2SO, | 4.6

m /48 Na,Citrate 4.0
m /48 NagHPO, 3.7
Mm /48 K,4Fe(CN)., Sell

strongly than sodium. With respect to sodium and potassium, the
result has appeared, on averaging all the determinations made with
salts of the two metals under like conditions, that sodium has a
slightly greater depressant action than potassium ; the former is some-
what more active as precipitant, and this result should be expected.
The average depressant power of ammonium, as estimated from the
above relatively few determinations, appears almost the same as

156 RS, Lphheg:

that of potassium. The three salts of this base show the usual
order of relative activity:

(NH,).SO, > NH,Cl > NH,CNS.

The three experiments with bivalent metals in Series I of Table X
show, in the case of calcium and cobalt, a strongly marked depressant
action on gelatine solutions; copper, on the contrary, depresses

TABLE XII.

1.25 % GELATINE + ALKALI EARTH CHLORIDES.
Series I. Series II.
Salt. mm. Hg.} Ppn. || Exp. Salt. mm. Hg.| Ppn.

0 (control) 5.9 0 (control) 4.4
m/96 MgCl, 3.2 0.542 m/192 MgCl, 2.0
m/96 CaCl, ri 0.457 m /192 CaCl, 2.0
m /96 SrCly 3:1 0.525 m /192 SrCly 2.0
m/96 BaCl,1 21 0.457 m /192 BaCl,1 1.8

1 Considerable precipitate of BaSO,. 1 Precipitate of BaSO,.

relatively slightly, a peculiarity to be ascribed in all probability to
the acidity of its solutions, which counteracts to a certain degree
the depressant action of the Cu-cations. The relatively slight
depressant action, as compared with other alkali salts, shown by
phosphate and ferrocyanide in the same table, may also be referred
to the hydrolysis of the salts, and the influence of the slight alka-
linity in checking the depression. The case of Na,CO, (Series IV
of Table VI) is similar. Salts like citrate and oxalate, which form
more nearly neutral solutions, have a similar depressant action to
sulphates (cf Table IX).

With albumin solutions, the alkali salts of plurivalent acids show
in all cases marked depressant action similar to, or greater than,
that of sulphates. This is distinctly seen in the determinations
with citrate, phosphate, and ferrocyanide in Table XI, and tartrate
in Table V. Similarly with salts of plurivalent metals, these all
act as powerful depressants, and in the case of many heavy metals
begin in very low concentrations to precipitate the albumin.

The Influence of Electrolytes. 57

The results of a number of determinations with salts of the alkali
earths and heavy metals are given in Tables XII and XIII.

TABLE XIII.

Ecc ALBUMIN + SALTS OF BIVALENT METALS.

Series I. 1.6 % egg albumin.

0 (control) ; so0c m /192 MnCl,
m/192CaCl, 5.6 |0.189|) 4 m/192 CoCl,1

Series II. 1.25 % albumin.

Salt. mm. Hg.| Ppn. || Exp. Salt. mm. Hg.| Ppn.

m/960 MnCl, 6.9 0.321
m [960 CoCl,? 5.6 0.260

0 (control) 21.5
m /960 MgCl, 73 | 0.339

6
7

m/960 CaCl, 76 |0.353|| 8 | m/960 CdCl,8 41 | 0.190
9

m/960 SrClz 72. \\O335 m/960 Pb(NOg).2| 2.8 ‘| 0.130
m /960 BaCl, 76 |0.353|| 10 | m/960 CuCly4 16 | 0.074

1 A little precipitation of albumin. Other salts — CuCly, Pb(NO3)2, CdCle, AICI;
— produced marked precipitation when added in the above proportion, and the pres-
sure readings obtained were uncertain. 2 Solution cloudy.

3 Considerable precipitation. 4 Largely precipitated.

It is evident from these tables that the above salts have more
energetic depressant action than those of the alkali metals. The
series with egg albumin show that heavy metal cations act more
energetically than those of alkali earths; and also that the salts
most effective as precipitants, are those having greatest depressant
action. It is interesting to note that Series II of Table XIII shows
a close parallelism between the decomposition tensions of the
various cations and the action of the respective salts, the degree
of depression increasing with decrease of decomposition tension.?

1 Compare MATHEWS’ experiments on comparative precipitating powers of
the heavy metal cations: This journal, 1905, xiv, p. 203.

158 Rs Sp Lithia.

There is again seen a decided difference between the two proteids,
with respect to the degree of depression produced by a given con-
centration of any salt; alkali-earth salts, when added in 22/960
concentration to albumin solutions, produce fully as great depres-
sions as they do when added in five times this concentration to the
equivalent gelatine solutions, — another indication of how much
more readily the state of solution is altered in the former than in
the latter of these two colloids.

IV. GENERAL INFLUENCE OF SALTS ON THE STATE OF
SOLUTION OF PROTEIDS.

Moore and Roaf! describe several experiments in which sodium
chloride and magnesium sulphate were added to serum with the
effect of lowering the osmotic pressure; the addition of 1 per cent
NaCl brought the pressure from 28.2 mm Hg (the value with dis-
tilled water as outer fluid) to 19.3 mm. Hg; using 20 per cent
MgSO, as outer fluid, the pressure first rose and gradually declined
to 18 mm. Hg. The depressions obtained are relatively slight as
compared with those described above. The authors’ aim is given
as follows: “Since a large amount of neutral salt added to a
proteid-containing solution has the effect of salting out the proteid
in many cases, the present experiment was designed to determine
whether, at a point short of precipitation, any increased aggregation
occurred as shown by diminished osmotic pressure.” The experi-
ments described in the present paper appear to justify the conclu-
sion that precipitation by addition of salt to a colloidal solution
represents merely the end stage of a process which in itself is
perfectly continuous. The state of solution of the colloid, so far
from remaining unchanged until a certain quantity of salt has been
added, changes continually as the concentration of salt increases;
the osmotic pressure appears steadily to decline, with increasing
concentration of salt, until a certain critical point is reached; this
is probably the point at which the surface energy of the colloidal
particles begins to exceed in value the energy keeping the colloid
in solution (or solution energy), of which the osmotic pressure is a
measure. At this stage equilibrium can no longer persist, and the
colloid begins to separate from solution; until this point is reached
no visible change may occur in the solution, though its relative

1 Moore and RoAF: Loe. cit.

The Influence of Electrolytes. 159

instability may be made evident in a variety of ways. The salts most
effective in- “salting out” are naturally those combining a well-
marked specific depressant action on osmotic pressure with a high
solubility — like ammonium sulphate.

V. VARIABILITY IN OSMOTIC PRESSURE OF COLLOIDAL
SOLUTIONS.

It is evident, at a glance, that the above proportionate pressure
values vary widely, even for solutions of the same composition, and
the impression might hence arise that the experimental error is
unduly large, and the method untrustworthy for exact determina-
tions. I believe that the true explanation of this variability is of
quite a different nature, and that the variable values observed corre-
spond more or less accurately to an actual variability in the solutions
themselves. It is noteworthy that all the hitherto published deter-
minations of the osmotic pressure of proteids show the same wide
variation in values, even in cases where the most careful and exact
methods have been employed. This is exemplified especially by the
work of Reid. Taking his determinations of the osmotic pressure

_ of hemoglobin solutions,! — which in this instance he attributes to

the proteid itself and not to the associated crystalloid impurities, —
we find, for solutions of crystallized haemoglobin varying in concen-
tration from 2.76 per cent to 6.05 per cent, a widely varying osmotic
pressure per unit concentration of colloid, ranging from 3.51 to
4.35 mm. Hg. for each 1 per cent of hemoglobin. It is noticeable
that the lowest relative pressures are found in the more concen-
trated solutions, —a circumstance suggesting that in these the colloid
has a somewhat coarser state of aggregation than in the more dilute
solutions. The measurements of Starling,? and especially of Moore
and Parker,* and Moore and Roaf,‘ also show marked variations of
pressure in solutions of the same composition.

It appears necessary to recognize that the conditions influencing
the osmotic pressure of a colloidal solution are more numerous than
in the case of a crystalloid solution, where concentration and tem-
perature together suffice to fix the value definitely. A great variety
of conditions enter in determining the osmotic pressure of a col-
loidal solution, of which temperature and concentration are only two.

? REID: Journal of physiology, 1905, xxxiii, p. 12.

2 STARLING: Loc. cit. 8 Moore and PARKER: Loc. cit.

* Moore and Roar: Loc. cit.

160 RR. S.. Lathe:

The state of aggregation of the colloid is of primary importance in
this relation; and this is influenced by numerous, conditions —
thermal, mechanical, and those dependent on the previous indi-
vidual history of the colloid. The individual particles of solute
are no longer monomolecular (as is typical, though not universal,
in solutions of crystalloids), but consist of more or less complex
aggregates of molecules; obviously the variability of these aggre-
gates in dimensions and physical properties introduces a further
complexity, which must be considered in dealing with these solu-
tions. To the two variables, concentration and temperature, must
be added a third, the aggregation state, with whatever other con-
ditions are involved by this; this third variable, which is largely
independent of the other two (though it is influenced by both) must
have a known value before the properties of the solution can be
predicted.

It is natural to assume that each “solution aggregate” or colloidal
particle has the same osmotic effect as a molecule or ion in a crys-
talloidal solution, though this has not been proved. The osmotic
pressure, on this assumption, depends on the number of particles in
unit volume and onthe temperature. It seems also a gv7ori probable
that the colloidal particles must offer a certain resistance to sub-
division or fusion, and hence that a given aggregation state, once
attained, will be only slowly altered by a change of conditions.
Hence, if in the preparation of a particular colloidal solution a cer-
tain state of aggregation is produced, this will tend to be conserved,
and the csmotic properties appropriate to this state will tend to
persist, for some time after the original conditions have been
changed. We have, in fact, evidence that this is the case in the
characteristic “hysteresis” of colloidal solutions, —a peculiarity to
which I shall now refer more particularly, since it accounts for much
of the otherwise apparently anomalous behavior of these solutions.

Given a colloidal system in equilibrium under a given set of con-
ditions, it is usually found that when the external conditions are
altered, the resultant change in the physical properties of the
system proceeds at a much slower rate, the change in properties
tending to dag behind the change in conditions, sometimes to a very
considerable degree. This peculiarity has been especially studied
by van Bemmelen.! As one of its consequences, he found that the

1 VAN BEMMELEN: Zeitschrift fiir anorganische Chemie, 1896, xiii, p. 232, 1898,
xvill, 1898, e¢ doc. ctt. above.

The Influence of Electrolytes. 161

hydration and dehydration curves of inorganic gels (silicic acid,
metallic oxides) obtained by plotting the relation between vapor
tension and water content at various temperatures, showed quite
different courses, so that a gel has different properties for the same
water content according to whether it has reached the particular
condition by dehydration from a more swollen condition or by
addition of water from one less swollen. In general, owing to this
peculiarity, the physical properties of the colloid system are de-
pendent not only on the conditions of temperature, concentration,
and chemical composition existing at the time of the determination,
but also, to a variable and often considerable degree, on the special
conditions to which the system has previously been expcsed.
Hence two gels of colloidal silicic acid, identical in composition,
may exhibit different vapor tensions; also, under certain conditions,
a more dilute solution may have a lower vapor tension than one more
concentrated, so that water may pass from the more concentrated to
the more dilute solution, —a result that seems paradoxical and
would be impossible in crystalloid solutions.

Similar conditions exist with respect to the osmotic pressure of
colloidal solutions; the pressure for a given composition may vary
according to the previous history of the particular solution, and
under certain conditions a less concentrated solution may have a
greater osmotic pressure than one more concentrated, involving a
possible transfer of water from the more concentrated to the more
dilute solution. Such possibilities should be taken into account in
considering the conditions of absorption, secretion, and other
processes involving transfer of fluid in the organism.

In considering the factors of the above variability, we must also
take into account the tendency to “spontaneous” change in aggre-
gation state so frequently observed in colloids. Van Bemmelen
concludes that in a colloidal substance no state of rest exists,! but
that there is a continual and apparently spontaneous alteration in
the molecular condition, and hence in the physical properties of the
colloidal substance. This alteration is very slow at normal tempera-
ture and is accelerated by heat, —a peculiarity which would seem to
indicate that the underlying condition is simply extreme slowness
in reaching a state of true chemical equilibrium. For instance, the
vapor tension —and with it the correlated power of absorbing or
giving off water — of a gel of given percentage composition changes

1 Cf. VAN BEMMELEN: Loc. cit., 1899, Xx, p. 206.

162 RR SLi.

slowly under the influence apparently of time alone. In conse-
quence, the water content of a hydrogel of silicic acid or ferric
hydrate depends not only on the concentration of the gas phase
and the temperature (as do the properties of a typical crystalloid
solution), but also on the simple lapse of time since its formation.

In general, according to van Bemmelen, the properties of such a
colloidal system depend not only on its percentage composition, but
also on the molecular structure acquired in its preparation,! together
with the modifications it has undergone in the process of dehydra-
tion, or through time, action of water, heating, mechanical treat-
ment, etc., —that is, on special peculiarities due to influences acting
on the colloid at various times during its previous history; these
have produced modifications that remain more or less permanently
fixed in its structure by virtue of its typical hysteresis. Similar
considerations apply to solutions of proteids, and must be taken into
account in considering the conditions of the above variability. It
is impossible to subject any two solutions to exactly the same
treatment, and variations in the aggregation state due to inequalities
of treatment, combined with inertia in changing an aggregation
state once acquired, are, in all probability, the chief reasons for the
above variability.

It may not be out of place here briefly to suggest that what applies
to colloids in simple solution must also apply a /fostzor to the col-
loids composing protoplasmic structures. The possibilities of such
differentiation are here necessarily much greater, and modifications
of the above kind must play an important part in determining the
properties of “ living’’ systems at given periods of their history. The
suggestion has been made by Pauli and others that it is to this prop-
erty of hysteresis that we must look as possibly affording a physical
substratum for one of the most. distinctively adaptive of the proper-
ties of organisms; namely, their possession of peculiarities of action
and reaction that can be explained only by reference to conditions
and stimuli that have acted at previous periods in their individual
history, —in a word, those processes collectively designated in con-
scious organisms as memory. The chief value of such a hypothesis
consists in its suggestions of exact research, and it ought not to be

' GIOLITTI, F.: Gazetta chimica Italiana, 1906, xxxvi, pp. 157 and 433, gives a
number of striking instances of how the properties of colloidal solutions of ferric
hydrate and tungstic acid can vary according to their method of preparation. See
also the paper Of DucLAUXx, Zoc. cit. above.

The Influence of Electrolytes. 163

impracticable to subject it to the direct experimental test.) At all
events, the colloids composing living substance demonstrably pos-
sess this property; and the inference can scarcely be avoided that it
must play some 7é/e, and possibly the above all-important one, in
the processes of the organism. This remains for future experiment
to determine.

VI. INFLUENCE OF MECHANICAL AGITATION AND OF CHANGES
OF TEMPERATURE ON OSMOTIC PRESSURE.

Inequalities in the rate of admixture of the electrolyte and in the
mechanical treatment of the solutions have been already suggested
as possible sources of the observed variability of pressure. Differ-
ences in temperature conditions in different series of experiments
also probably play some part. I have as yet made no study of the
influence of the first named of these factors; but the results of
several experiments in which solutions were subjected, respectively,
to violent shaking and to temporary warming before measuring tke
pressures, seem significant and are accordingly presented here.

In the first of the two series of the following table (with 1.25 per
cent gelatine) the solutions were mixed with as little agitation as
possible; one solution of each of the six pairs was then shaken vio-
lently in a smail flask for a short time; the other remained unshaken.
The solutions were then introduced into the osmometers under as
nearly as possible identical conditions. In the series with 1.6 per
cent albumin soJution the procedure was similar; one solution of each
pair was shaken by hand as violently as possible for thirty seconds.

In the gelatine series there is relatively slight difference between
the shaken and the unshaken solutions; the former, however, in
every case show a slightly higher osmotic pressure. In the albumin
series little difference is seen between Solutions 1 and Ia; in all
the others the shaken solution shows a distinctly lower pressure.
Several conditions must be considered in accounting for tlis effect;
the shaking process transforms a considerable part of the solution
into the condition of a finely divided foam, and it is probable that

' Possibly by determining the temperature coefficient for the establishment of
associations in animals, ¢. g., the relative number of trials required to fix an asso-
ciation or habit in any given cold-blooded animal at different temperatures. Agree-
ment with the temperature coefficient of hysteretic changes would confirm the

hypothesis. So far as I am aware, no exact investigation has hitherto been made
of the influence of temperature conditions on either one of these processes.

164 Te. SS. Lhe:

in the foamwork of thin films forming this portion of the solution
the concentration of proteid is somewhat higher than elsewhere,
since, according to Gibbs and J. J. Thomson,! substances that lower
the surface tension of the solvent, tend to accumulate in the surface

TABLE XIV.

Series I. 1.25 % Gelatine. Series II. 1.6% Egg Albumin.

Solution.

Pure gelatine
a << (shaken)

Gelatine + 2/48 NaCl

ss © “ (shaken)
Gelatine + m/48 NaBr

a = “« (shaken)
Gelatine + m/48 Nal

s of “ (shaken)

Solution.

Pure albumin
# s (shaken)
Albumin + 2/48 NaCl
e < “ (shaken)
Albumin + m/48 NaBr
es “ “ (shaken)

Albumin + m/48 Nal
as ie “ (shaken)

Gelatine + m/48 NaCNS Albumin + 2/48 NaCNS

“ of “ (shaken) 5 sf 4 “ (shaken)

Gelatine + m/48 Na,SO,

* “ (shaken)

film in somewhat greater concentration than in the other regions of
the solution. On account of the large surface extent of the above
foam, this effect would be multiplied in that part of the solution,
and the concentration of the remainder would be somewhat lowered.
It seems hardly possible, however, that this effect is sufficient to
account for the differences of pressure observed in the series of
albumin selutions. From the 60 c.c. of solution, it is easy after
shaking to pour off 50 c.c. of a clear fluid; the volume of liquid
contained in the foam is thus certainly less than Io c.c., and it
would be necessary to suppose that the concentration of albumin in
this portion of the solution was approximately doubled to account
for the average difference between shaken and unshaken in the

* Cf. HOEBER: Physikalische Chemie der Zelle und der Gewebe, 2te Auflage,
Leipzig, 1907, p. 209.

The Influence of Electrolytes. 165

series of solutions containing salt. The true explanation is, in all
probability, that a certain aggregation of colloid particles occurs in
consequence of the shaking, and that this condition of increased
aggregation tends to persist, and is responsible for the lowering of
pressure. Ramsden’s! work has shown that mechanical shaking
may produce coagulative changes in colloidal solutions. The fact
that the change is well marked only in those solutions to which
salt was added, and is seen in all of these, tends to support this
view; it is to be expected that agitation will more readily induce
agsregation changes in solutions whose stability has already been
lowered by the addition of salt, than in those where the state of
subdivision is relatively fine, and the osmotic pressure correspond-
ingly high. It may be concluded that the osmotic pressure and the
other physical properties of a colloidal solution are affected to no
slight degree by the mechanical conditions acting on it during or
after its formation. The change in the gelatine solutions is less
marked, and is in the opposite direction. What the meaning of
this difference may be is difficult to say; it is probably related to
the above-mentioned difference between the two colloids in regard
to the ease with which their respective states of solution are altered.

Temperature conditions also influence the osmotic pressure of the
solution. Moore and Roaf found in the case of gelatine solutions
that the pressure increased with rise of temperature at a more rapid
rate than the absolute temperature; and that the increased osmotic
pressure persisted, by a kind of hysteresis, for several days after
normal temperature was restored, gradually returning to its original
value. They ascribed the effect to a finer subdivision of colloidal
particles at high temperatures. In the following four experiments
a similar condition was found. A I per cent gelatine solution was
used; this was kept in a flask surrounded by chopped ice for a
certain period (twenty-four to forty-eight hours) ; a portion was then
transferred to a small, tightly corked flask which was kept immersed
in a water bath at 65° to 70° for a period of three to four hours.
The contents of the flask were then cooled; 60 c.c. of the solution
was then placed in a beaker, an equal quantity of the still cold solu-
tion was taken; the temperatures of both were then equalized by
bringing to room temperature, and each was placed in an osmometer.

The next morning (ca. eighteen hours later) readings were taken.
The following results were gained in four experiments.

1 RAMSDEN : Zeitschrift fiir physikalische Chemie, 1904, xlvii, p. 336.

166 RS. Lellee,

Warming the solution thus produces a considerable increase in its
osmotic pressure, and this increase persists for some time after the
solution has cooled. The difference between the warmed and the
unwarmed portions of the solution gradually diminishes; in Experi-
ment 1, four days later, the cooled portion still gave ca. 5.0 mm. Hg
pressure; that of the warmed had déclined to 5.3 mm. Hg. The
heightened pressure is thus only temporary.

TABLE XV.

Pressure in mm. Hg.

Time of experiment.

A. Unwarmed. | B. Warmed.

May 2-3 5.0 64
May 2-41 4.9 6.0
May 7-8 57, 6.2

May 13-14 5.6 6.0

1 Same solution as in Experiment J, but cooled twice as long.

I hope before long to make a more extended study of this and
similar changes, and of the rate at which they proceed under different
‘conditions of temperature and concentration, the influence of me-
chanical factors, etc. The slight permeability of the above pure
celloidin membranes to gelatine introduces a difficulty; it will be
necessary in experiments extending over prolonged periods of time,
such as would be necessary in an investigation of this kind, to use
membranes that are not only unaltered by the action of the solution,
but are in addition absolutely impermeable to the colloid. Several
other questions remain for examination, among them the rate of
change of osmotic pressure under influence of time, the influence
of the rate of addition of the electrolyte to the colloidal solution,
the influence of withdrawal of the electrolyte after having allowed
it to act for varying periods, the general part played by purely
mechanical factors, etc. It will be fruitless to form any detailed
theory regarding the physico-chemical condition of the colloids in
living cells, and the nature of the changes undergone by the colloidal
constituents of protoplasm, until we are in possession of a more
exact and extended series of data bearing on the above questions.

The Influence of Electrolytes. 167

VII. INFLUENCE OF THE OSMOTIC PRESSURE OF THE COLLOIDS
ON THE ABSORPTION PHENOMENA OF CELLS.

The absorption of water by cells has hitherto been referred
rather to the osmotic pressure of the crystalloids than to that of the
colloids of protoplasm. In plant cells, where the phenomena admit
of readiest study, the osmotic pressure or turgor is usually ascribed
to the crystalloid substances contained in the central vacuole or
vacuoles, the protoplasmic portion of the cell being regarded as merely
a semi-permeable membrane. In the appearance of such vacuoles,!
however, osmotic relations between the colloid-rich portions or
phases of the protoplasmic system and the more fluid portions must
play a part. In brief, the osmotic properties of the protoplasmic
colloids must also be considered. The possibility that this factor
may be of importance in determining the water content of the cell
appears to have been little considered by biologists, although, as
Duclaux expresses it,? “it is evident that if the colloids of living
cells can develop pressures of the same order” (as those observed in
artificial colloidal solutions) “we have no right to neglect them, and
to attribute to crystalloids alone the observed osmotic phenomena.”

The most distinctive characteristic of the osmotic pressure of the
cell colloids is that it is not a constant varying with concentration
and temperature alone, but that it changes with the state of aggrega-
tion, which must constantly vary according to the conditions affect-
ingthe cell. Thus the osmotic pressure may at one pericd increase,
with resultant absorption of fluid from the surroundings; a change
may then follow in the reverse direction, due to the colloids’ enter-
ing a state of increased aggregation, with a consequent fall of
osmotic pressure and separation or secretion of fluid.

The facts which I have recently described in considering the
influence of various salts in promoting water absorption or swelling
in ciliated epithelial cells,? receive a ready explanation on the basis
of the above theory. It is observed that cells swell more or less
rapidly in pure isotonic solutions of sodium or potassium salts; the
rapidity of swelling with salts of different anions increases in the
order: plurivalent anions < Cl < NO, < Br< I< CNS, an order

1 VAN BEMMELEN describes the formation of hollow spaces in various inorganic
_ gels, a phenomenon apparently analogous to vacuole formation in protoplasm.

2 DucLaux: Loe. cit., p. 41.

3 R. LILuiE: This journal, 1906, xvii, p. 91. Cf especially p. 122 e¢ seg.

168 We. S02 Laie.

identical with that found above. The explanation of these facts is
simple: the electrolyte content of the cell in its normal medium,
sea water, imparts to the colloids a state of aggregation, and hence
of osmotic pressure, such that the water content remains — so far as
can be seen — approximately constant. When such a cell is im-
mersed in (e. g.) a pure isotonic NaCl solution, a tendency arises
for the ions of the protoplasm other than Na and Cl to diffuse out-
ward and to be replaced by sodium and chlorine, leaving the proto-
plasm with an excess of these ions and a relative deficiency in
Mg, Ca, and SO, ions; in the presence of these last-named ions the
colloids have a lower pressure than in their absence; hence their
removal results in a heightening of osmotic pressure and absorpticn
of water. If isotonic NaBr is used, the swelling is more rapid in
consequence of the greater osmotic pressure imparted by the Br
ions. Similarly, iodides induce greater swelling than bromides,
and so on with the other salts. The above order of effectiveness
is thus readily accounted for; and the dependence of the rate of
absorption on the osmotic pressure of the cell colloids is clearly
indicated.

In conclusion, it affords me much pleasure to express to Dr.
Howell and the members of the physiological staff of Johns
Hopkins University my best thanks for the many courtesies I en-
joyed during my stay in their laboratory. To Dr. Abel and the
members of the department of Physiological Chemistry I am also
under many obligations — including the use of a room in the Pharma-
cological Laboratory — which it is a pleasure for me to acknowledge.
My thanks are due also to Dr. Mall for the use of certain materials
from the Anatomical Laboratory.

SUMMARY.

1. The osmotic pressure of solutions of gelatine and of egg
albumin were measured directly, using osmometers with membranes
of nitro-cellulose (celloidin). The following pressure determinations
were made: (1) of the approximately pure colloidal solution; (2) of
the same with the addition of various electrolytes and non-electrolytes
in definite concentrations; and (3) of the colloidal solution (or
mixtures of colloidai solution and electrolyte) after subjection to
mechanical agitation or temporary elevation of temperature.

2. The osmotic pressure of the colloids remains unaffected after

The Influence of Electrolytes. 169

the addition of non-electrolytes (sucrose, dextrose, glycerine, urea)
to the solution. All electrolytes alter the osmotic pressure, in
some cases decreasing, in others diminishing, the pressure of the
original solution.

3. Acid and alkali increase the osmotic pressure of gelatine solu-
tions; in general these substances affect the osmotic pressure of
gelatine solutions in the same manner as they do the rate of swelling
of solid gelatine plates immersed in water.

4. Addition of salts depresses the osmotic pressure of both col-
loids; the degree of depression is a function of the nature of both
the anion and the cation of the salt. It increases in the order:
alkali metals < alkali earths < heavy metals (for cations); and CNS
< I< Br< NO,< Cl< F < plurivalent anions — SQ,, tartrate, cit-
rate, phosphate (for anions).

5. The hysteresis of the colloidal solution plays an important part
in determining its osmotic properties; temporary elevation of tem-
perature and mechanical shaking are shown to produce more or less
persistent changes in the osmotic pressure of the solutions. Simi-
larly with other conditions affecting the aggregation state of the
colloid (mode of preparation, age, rate of admixture of electrolyte,
etc.). The individual history of the particular solution is thus a
factor; hence solutions of the same composition may show quite
different osmotic properties.

6. In their respective actions in promoting the absorption of
water by cells, the series of alkali metal salts show the same general
order of effectiveness as in their action on the osmotic pressure of
the above colloids. The absorption of water by cells under these
conditions is thus to be referred to the osmotic pressure of the cell
colloids. This last is not a constant quantity, but varies accordirg
to the state of aggregation of the colloids.

THE ACTION OF NORMAL FATIGUE SUBSTANGES
ON. MUSCLE,

BY EREDERIC S. EEE.

[From the Department of Physiology of Columbia University, at the College of
Physicians and Surgeons, New York.]

HERE exists in the abundant literature of fatigue, strangely

enough, no account of an adequate investigation of the physio-
logical action of fatigue substances. The most comprehensive study
is that which was chronologically the first, namely, Ranke’s, whose
account was published in 1865; and to it we must still refer most of
our knowledge of this subject. Although invading a new field, it was,
and continues to be, a most valuable contribution. But at the time
of its appearance there was wanting the most important single
method of investigating muscular activity, the graphic method with
its modern refinements, and without this there were necessary limi-
tations to Ranke’s research. He did, however, demonstrate the
depressing or fatiguing action on skeletal muscle, first, of an extract
of muscle supposedly containing all the products of muscular activ-
ity; and then of two of those products, lactic acid and kreatin; and
he gave to these two the name “ fatigue substances,” —a term which
has ever since remained in good usage, though with a varying
domain. The depressing action of these substances when injected
into a muscle was exhibited in a lowering of the irritability of the
muscle, as measured by the intensity of the electrical stimulus
required to elicit a minimal contraction; and in a lessened distance.
over which a pointer attached to the muscle was swung when the
latter was stimulated bya tetanizing series of stimuli constant in inten-
sity. Decreased irritability and working power are two of the promi-
nent, but not the only, characteristics of the fatigued muscle. Of the
materials originally designated fatigue substances, Ranke himself
later rejected kreatin, because of an error discovered in his earlier

experiments, and furthermore accepted carbon dioxide and acid
170

The Action of Normal Fatigue Substances on Muscle. 171

potassium phosphate. In recent years there has been a pretty general
acceptance of Ranke’s later classification. Weichardt is, I think,
the only one who has attempted to add specifically to the list of
substances whose presence is responsible for normal fatigue. He
believes that he has demonstrated the existence in fatigued muscles
of a distinct toxin allied to bacterial toxins, and capable, when injected
into fresh animals, of producing of itself the phenomena of fatigue.
He has gone even further, and obtained, he believes, by accepted
bacteriological methods, an antitoxin, which possesses the power of
nullifying the effects of the toxin and restoring the fatigued organism
to its previous condition. So sweeping a claim, and one of such
interest, demands careful examination, and I hope at a future date to
review Weichardt’s work experimentally. It is not improbable that
future research will discover other substances that are causative of nor-
mal fatigue. It seems scarcely credible that the few named shall prove
to be, of all the links in the long metabolic chain, the only substances
that are depressant to protoplasmic activity. It is especially the in-
termediary metabolic products that must be examined from this
standpoint.

But whether any of these prove to be efficacious, and whether a
fatigue toxin be a reality or not, one fact must be accepted, namely,
that of acidification of the muscle in fatigue. That the working mus-
cle produces acid has been known since du Bois Reymond’s demon-
stration in 1859 of an acid reaction in this tissue after activity.
Many and conflicting attempts have been made to determine the
nature of the substance that is responsible for this acid reaction. By
various investigators it has been ascribed to one or all of the three
recognized fatigued substances, and it seems not improbable that
they all share in it. Carbon dioxide is apparently a factor; mono-
potassium phosphate is probably so; paralactic acid is surely present,
as is now clearly demonstrated by the excellent work of Fletcher and
Hopkins. Whether the last-named substance occurs free or in com-
bination with potassium or other elements as a neutral salt, is not
quite clear.

Before proceeding further in our search for fatigue substances,
however, it is desirable to know how the three that are recognized
affect the muscle in which they occur. The present research deals
with this phase of the general problem, and leaves untouched the
question of possible fatigue substances other than these.

It may be stated at the outset that I find these substances to act

172 Frederic S. Lee.

on skeletal muscle in ways essentially identical. Such action is,
however, of two directly opposite modes, the appearance of the one or
the other mode being dependent upon the quantity of the substance
that is used, and the duration of its activity. If used in what we
may term moderate or large quantity, or in smaller quantity over a
longer time, each substance is distinctly fatiguing, its action being
characterized by a decrease in the irritability and the working
power of the muscle; a lessened height to which the load is lifted ;
a decrease in the total amount of work that is performed; in warm-
blooded animals a diminished duration of the single contraction; and
in cold-blooded animals a slowing or increased duration of the single
contraction, a slowing which affects chiefly the phase of relaxation
and is followed by a quickening. The opposite action of the same
substances is seen when they are employed in small quantity, or in
moderate quantity for a brief time. Instead of a diminution of activ-
ity there is an augmentation, which is characterized by an increase
in irritability and working power, an increase in the height to which
the load is lifted, and an increase in the total amount of work per-
formed. This latter mode of action is, I believe, responsible for the
progressive augmentation of activity that is customarily present in
the early stages of a series of successive muscular contractions, and is
known as the treppe. I have already considered this in a separate
paper. The present contribution is confined to the depressing
action of the fatigue substances in question.

My experiments were performed between the months of October
and June inclusive, on both frogs and cats. The methods that were
used are the same that were employed in the study of the cause of
the treppe, and for an account of them the reader is referred to the
former article. It may be stated here, briefly, that in the majority of
cases, after killing the animal, irrigation of two corresponding muscles
of opposite legs was employed, physiological salt solution or whipped
blood being used for the one, or control, muscle, and for the other
the same liquid containing a certain quantity of the particular fatigue
substance to be tested. The muscles were then stimulated at regular
intervals by break induction shocks applied directly until fatigue was
pronounced, and graphic records of the contractions were made. A
delicate isotonic system was employed, and great care was exercised
to eliminate errors of technique. A very large number of experi-
ments have been made, and the results have been desirably uniform.

The Action of Normal Fatigue Substances on Muscle. 173

RESULTS.

Some of the results that have been obtained have been presented
in various preliminary papers, especially in a lecture before the
Harvey Society. The reader is also referred to the article on the
cause of the treppe, in some of the illustrations of which may be
found graphic records of typical experiments demonstrating not only
the augmenting, but also the fatiguing, action of fatigue substances.
In considering the results it should be clearly borne in mind that
all three fatigue substances act in ways that qualitatively are essen-
tially identical. The action of any one substance should not there-
fore be considered as specific for that substance in distinction from
the others. Results obtained from each of them will, however, be
presented.

Their fatiguing action on frog’s muscle is illustrated by Figs. 1
and 3 of the article on the treppe and by Fig. 1 of the present paper,
which is the graphic record of a typical experiment on the action of
mono-potassium phosphate.

December 22,1905. Frog, weight 42 gm., pithed. One thigh was ligated
temporarily. 25 c.c. of 0.75 per cent solution of sodium chloride were
injected into the bulbus arteriosus during a period of eight minutes. After
thirty minutes more had elapsed the irrigated gastrocnemius was excised
and prepared for direct stimulation. Maximal stimuli, 30 per minute.
Weight actually lifted 5 gm. Every fiftieth contraction was recorded.
After the record was completed the temporary ligature on the opposite
leg was removed, and 25 c.c. of 0.75 per cent solution of sodium chloride
containing 7 KH,PO, were injected through the bulbus into the oppo-
site leg. After a wait of thirty minutes the gastrocnemius was removed,
and a record as above was made, the muscle curves rising from the same
abscissas as those of the first muscle. The tracing shows every fiftieth
curve of each muscle from the rst to the 651st inclusive. The longer,
or, in the later contractions, the lower, curves are those of the muscle
under the influence of mono-potassium phosphate.

Contrary to what appears in many experiments with mono-potas-
sium phosphate, no preliminary augmentation is here present. From
the 51st contraction onward, however, the fatiguing action of the
drug is manifest. This is seen first in a greater lengthening of the
muscle curves as compared with the normal curves — in other words,
a greater slowing of the contraction process, which is confined almost

174 Frederic S. Lee.

wholly to the phase of relaxation. This excessive slowing in duration
of the contraction is apparent during the first 400 contractions, after
which time, in harmony with the diminished lifting power, the dura-
tion becomes relatively lessened. The diminished lifting power is
exhibited from contraction I5f onward in a constantly progressive
degree. When the experiment ends with contraction 651, the poisoned
muscle is performing about one half the work of.the other. Increased
length and diminished height are thus the striking features of the
contraction curves of the poisoned, as compared with those of the
normal, muscle. Curve 51 of the poisoned muscle is almost the exact
counterpart, in both height and length, of curve 151 of the normal
muscle. In other words, the poisoned muscle at its 51st contraction
is already fatigued to the same extent as the normal muscle at its
151st. The same comparison may be made between curves 101 of
the poisoned and 201 of the normal muscle. The pronounced fa-
tiguing action of mono-potassium phosphate is evident. In the pres-
ent experiment a fairly strong solution of the drug was employed.
A similar effect may be obtained when a weaker solution acts for a
longer time.

Figure 2 demonstrates the fatiguing action of mono-potassium
phosphate on cat’s muscle.

March 18, 1907. A cat was killed by decapitation at 2.20 p.m. A cannula
was placed at once in each femoral artery, and at 2.27 irrigation of the
two legs was begun simultaneously under a constant pressure of 150 mm.
Hg and a temperature of 4o° C. The right leg was irrigated with the
whipped blood of a bullock, the left leg with similar blood to which had
been added pure mono-potassium phosphate in quantity to equal 1/25 of
a grammolecular solution. ‘The irrigation continued for fifteen minutes.
After its cessation the corresponding extensor longus digitorum muscles
were rapidly excised, placed in moist chambers at room temperature,
attached to exactly similar levers, weighted so as to lift 5 gm. each, and
stimulated simultaneously by the same series of maximal break induction
shocks at the rate of 29 in the minute, every fiftieth contraction being re-
corded on a rapidly moving drum. The record began at 2.48. The left-
hand figure shows the tracing of the normal muscle, the right-hand figure
that of the muscle under the influence of the mono-potassium phosphate.

The normal muscle of this experiment becomes fatigued in the
manner customary to the muscles of warm-blooded animals. There
is no lengthening of the duration of the contraction, as fatigue pro-
ceeds, but there is a progressive diminution in the extent of the lift.

5 4
201
s

a

ii

ial

NO

The Action of Normal Fatigue Substances on Muscle. 175

(For a discussion of this subject the reader is referred to the author’s
articles, ‘‘ The fatigue of cold-blooded compared with that of warm-
blooded muscle,” and “ Temperatur und Muskelermiidung.”) The
muscle that is under the influence of the added fatigue substance
shows this phenomenon to a more marked degree than the normal
muscle. Beginning with the first contraction, the curves of the
former are much lower than those of the latter. The 4oIst contrac-
tion reveals pronounced exhaustion, while the normal muscle at this
time is still performing nearly one half the amount of work that it
performed at first. Beginning with the first contraction also, the
poisoned muscle exhibits the shortness of curve that is characteristic
of warm-blooded muscle in the later stages of fatigue. The physio-
logical action on warm-blooded muscle of KH,PQO, in such quantity
as is here employed — and the same may be said of the other fatigue
substances — thus differs in one sense qualitatively from its action
on cold-blooded muscle. This is a striking and important fact. In
each case, however, the substance serves simply to hasten the appear-
ance of, and thus accentuate the phenomena of, normal fatigue, which
are characteristic of the type of muscle involved. In this sense the
action is qualitatively the same with the two types. The qualitative
difference may be cited in support of the accepted theory of fatigue
as due to products of metabolism. In such an experiment as the
above we artificially introduce into the muscle a store of such prod-
ucts, and in consequence the muscle, when put into action, exhibits
already in its customary manner a certain degree of fatigue.

Figure 3 shows the characteristic result of an experiment in which
CO, is employed in considerable quantity.

October 18, 1906. <A cat was killed by decapitation at 10.10 A.M. A cannula
was placed at once in each femoral artery, and at 10.15 irrigation of the
two legs was begun simultaneously under a constant pressure, and a tem-
perature of 36.5° C. The right leg was irrigated with o.g per cent solu-
tion of NaCl, the left leg with similar solution which had been charged
with CO,. The irrigation continued for three minutes. After its cessa-
tion the corresponding extensor longus digitorum muscles were rapidly
excised, placed in moist chambers at room temperature, attached to ex-
actly similar levers, weighted so as to lift 5 gm. each, and stimulated
simultaneously by the same series of maximal break induction shocks at
the rate of 29 in the minute, the contractions being recorded on a slowly
moving drum. The record began at 10.23. In the figure the upper
tracing represents the normal muscle, the lower the muscle under the
influence of CQOg.

176 Frederic S. Lee.

Here we have a typical case of fatigue superinduced by an exces-
sive amount of a fatigue substance. In the muscle charged with CO,
there was no preliminary augmentation. There was lessened lifting
power from the beginning, and an earlier onset of exhaustion, than in
the companion muscle. Figure 1 of the article on the cause of the
treppe illustrates the same phenomena together with the additional
slowing of relaxation peculiar to cold-blooded muscle. These phe-
nomena which are exhibited by the striated muscle of the frog under
the influence of CO,, have been previously observed by several inves-
tigators, especially Waller and Sowton, Lahousse, Lhotak de Lhota
and Spada.

Lastly, Fig. 3 of the article on the treppe represents the fatiguing
action on cold-blooded muscle of sarcolactic acid, and Fig. 4 of the
present article that of lactic acid on warm-blooded muscle. In the
latter experiment I employed the ordinary lactic acid of the chemist
(Merck) instead of sarcolactic acid, because of the fact that my sup-
ply of the latter substance had become exhausted. This experiment
was performed in the customary manner as follows:

April 22, 1907. — A cat was killed by decapitation at 12.26 p.m. A cannula
was placed in each popliteal artery, and at 12.33 irrigation of the two legs
was begun simultaneously under a constant pressure, and a temperature
of 39° C. The right leg was irrigated with the whipped blood of a bul-
lock, the left leg with similar blood to which had been added lactic acid
in quantity to equal 1/20 of a grammolecular solution. The irrigation
continued for ten minutes. After its cessation the corresponding extensor
longus digitorum muscles were rapidly excised, placed in moist chambers
at room temperature, attached to exactly similar levers, weighted so as to
lift 5 gm. each, and stimulated simultaneously by the same series of
maximal break induction shocks at the rate of 29 in the minute, the con-
tractions being recorded on a slowly moving drum. The record began at,
12.49. In the figure the upper tracing represents the normal muscle, the
lower the muscle under the influence of lactic acid.

There is here the customary picture of preponderating fatigue in
the muscle placed artificially under the influence of a fatigue
substance.

In dealing with the cause of the treppe the question was raised
as to whether the fatigue substances act through their acid character
solely. The same problem presents itself here. The physiological
actions of the three substances are essentially identical qualitatively,
and free hydrogen ions are present in the solutions of all. The same

The Action of Normal Fatigue Substances on Muscle. 177

facts are true of other substances with which I have experimented,
namely, oxybutyric acid, hydrochloric acid, sulphuric acid, and mono-
sodium phosphate. It thus appears to be a fair inference that the
fatiguing action of all of these substances is associated, in part at
least, with free hydrogen ions. But our knowledge of the physiolog-
ical action of potassium salts makes it certain that in solutions of
KH, PO, K ions are also responsible for a part of the results. With
regard to lactic acid I have tried to obtain light by comparing the ac-
tion of the free acid with that of its salts. I have employed sodium,
potassium, and ammonium lactates in solutions that are neutral to
litmus, alkaline to Congo red, and only faintly acid to phenolphtha-
lein, and in each case I have obtained qualitatively the same fatiguing
effect as with the free acid. In equimolecular solutions (75), acting
for the same number of minutes, potassium lactate proved to be the
most fatiguing ; sodium lactate seemed to rank next; while ammo-
nium lactate exerted at first an augmenting action, which was followed
by depression. These facts seem to justify the conclusion that both
anions and kations share in the physiological action of lactic acid.
The actual state of the latter in muscle, whether free or combined,
has not yet been elucidated. We seem to be warranted, however, in
deducing the general conclusion that normal fatigue substances exert
their fatiguing action partly through their acid character, that is,
their hydrogen ions, and partly through other portions of their
molecules.

Another obvious question is whether fatigue substances act di-
rectly on the living substance of the muscle cells or on their nerve
supply. I find that the effects occur similarly with both curarized
and non-curarized frog’s muscle, and with cat’s muscle long after its
excision, when its nerves have long ceased to respond to stimuli and
presumably all nerve substance is dead. There is no doubt, there-
fore, that fatigue substances act directly on the protoplasm of the
muscle cells. This, however, leaves still open the question as to
whether they may not also depress the substance of both nerve
fibres and intramuscular nerve-endings. Probably they do this, but
we have no direct experimental evidence regarding it. Neverthe-
less, it is probable that the fatigue substances of muscle are general
fatigue substances, depressing not only the cells in which they arise,
but distant ones belonging to both similar and dissimilar tissues.

To what extent in the intact body fatigue of muscle shares in gen-
eral bodily fatigue is not yet fully known. Following Wundt, James,

178 Frederic S. Lee.

Miinsterberg, and Baldwin, we must believe, however, that both the
feeling of effort which accompanies muscular action, and the sensa-
tion of fatigue, are chiefly of peripheral origin. Moreover, the long
current belief that in continued labor the central nervous system be-
comes fatigued before the peripheral tissues is now rendered ex-
tremely doubtful by the work of Kraepelin, G. E. Miiller, Henri,
R. Miller, Hough, Woodworth, Storey, and Joteyko. Their results
make it even probable that the brain. and spinal cord are, like the
nerve fibre, resistant, and they throw a certain measure of doubt on
all supposed proofs of central fatigue. For a more specific discussion
of this subject and references to the literature, the reader is referred
to my Harvey Lecture. I hope also to be able to take up the matter
experimentally at an early date.

CONCLUSIONS.

1. The physiological action on skeletal muscle of each of the
commonly recognized fatigue substances, namely, carbon dioxide,
paralactic acid, and mono-potassium phosphate, is of two opposite
modes, the appearance of the one or the other mode being depend-
ent upon the quantity of the substance that is present and the
duration of its activity.

If present in small quantity, or moderate quantity for a brief time,
each substance causes an augmentation of activity of the muscle,
which is characterized by an increase in irritability and working
power, an increase in the height to which the load is lifted, and an
increase in the total amount of work performed.

If present in moderate or large quantity, or in smaller quantity
for a longer time, each substance causes a depression of activity, or -
fatigue, of the muscle, which is characterized by a decrease in irri-
tability and working power; a lessened height to which the load is
lifted; a decrease in the total amount of work performed; in warm-
blooded animals a diminished duration of the single contraction; and
in cold-blooded animals an increased duration of the single contrac-
tion, which affects chiefly the phase of relaxation and is followed by
a quickening.

2. The depressing action of fatigue substances is due in part, but
not wholly, to free acid, since it is shown by neutral solutions of
sodium, potassium, and ammonium lactate.

3. The depressing action of fatigue substances occurs in both

The Action of Normal Fatigue Substances on Muscle. 179

curarized and non-curarized frog’s muscle and in cat’s muscle in
which the nerve has long since died. Hence the action is exerted
upon the muscle protoplasm itself.

BIBLIOGRAPHY.

®

DU Bols REYMOND: Monatsberichte des kénigl. Akademie der Wissenschaften
zu Berlin, 1859, p. 288.

FLETCHER and HopkINs: Journal of physiology, 1907, xxxv, p. 247.

LAHOUSSE: Archives internationales de physiologie, 1905-1906, iii, p. 453; Bulletin
de l’academie royale de medecine de Belgique, 1898.

LEE: The fatigue of cold-blooded compared with that of warm-blooded muscle,
Journal of American Medical Association, 1905, xlv, p. 1776. Temperatur und
Muskelermiidung, Archiv fiir die gesammte Physiologie, 1905, cx, p. 400.
Some of the chemical phenomena of muscle fatigue, Proceedings of the Ameri-
can Physiological Society, This journal, 1906, xv, p. xxxi. Acid and fatigue,
Proceedings of the New York Academy of Sciences, Section of Biology,
Science, 1906, N. S. xxiii, p. 504. Fatigue, Journal of the Americal Medical
Association, 1906, xlvi, p. 1491; Harvey Lectures, i, p. 169, New York, 1906.
Ueber die Ursache der Entstehung der Treppe, Zentralblatt fiir Physiologie,
1906, xx, p. 869. The cause of the treppe, This journal, 1907, xviii, p. 267.

LHOTAK DE LuoTA: Archiv fiir Physiologie, 1902, Supplement-Band, p. 43;
Journal de physiologie et de pathologie générale, 1902, iv, p. 976.

RANKE: Tetanus, Leipzig, 1865.

SPADA: Archivio di farmacologia e terapeutica, tgor, ix, p. 131; Archives itali-
ennes de biologie, 1902, xxxvii, p. 129.

WALLER and SowTon: Proceedings of the Physiological Society, Journal of
physiology, 1896, xx, p. xvi.

WEICHARDT: Miinchener medizinische Wochenschrift, 1904, li, p. 12; Jézd., 1904,
li, p. 2121; Jbéd., 1905, lii, p. 1234; Zbzd., 1906, liii, p. 7; Lbid., 1906, iii,
p- 1701.

THE RELATION BETWEEN THE BLOOD SUPPLY TO
THE SUBMAXILLARY GLAND AND THE CHARACTER
OF THE CHORDA AND THE SYMPATHETIC SAEIVE
IN THE DOG AND GE CAT.

By A. J. CARLSON; J. R: GREER, AND F.C. SECHA
[From the Hull Physiological Laboratory of the University of Chicago.]

I. INTRODUCTORY.

alee. literature. —In the dog stimulation of the chorda tympani
nerve produces vaso-dilation in the submaxillary gland pavz passu
with secretion of saliva, and stimulation of the cervical sympathetic
causes vaso-constriction in the gland simultaneously with a scanty se-
cretion of relatively concentrated saliva. The natural inference is that
the difference in the character of the chorda and the sympathetic saliva
is due to this difference in the blood supply of the gland. Heiden-
hain! investigated this question in the dog and the rabbit, reaching
the conclusion that the percentage composition of the saliva does not
vary with the vascular condition of the gland. Heidenhain reports
one experiment on the dog’s parotid, in which he tied both subcla-
vians and both carotids and stimulated Jacobson’s nerve. By this
procedure the blood supply to the gland is not completely shut off,
but it is greatly diminished. He found no increase in the per-
centage composition of the chorda saliva collected during the
diminished blood supply. Heidenhain also reports two experi-
ments on the rabbit’s parotid. In the rabbit the ligation of both
subclavians and both carotids shuts off all the blood from the parotid
gland, according to Heidenhain. He found no increase in the solids
of the saliva following the stimulation of Jacobson’s nerve with the
gland completely deprived of blood. Heidenhain concludes, there-
fore, that the difference in the vascular conditions of the salivary
glands on stimulation of the sympathetic and the cranial secretory

1 HEIDENHAIN: Archiv fiir die gesammte Physiologie, 1878, xvii, p. 1; HER-

MANN’S Handbuch, vy, p. 46.
180

Blood Supply to the Submaxillary Gland. 181

nerves is not a factor in determining the difference between chorda
and sympathetic saliva. This conclusion is generally accepted by
physiologists.

Even prior to Heidenhain’s observations Eckhard! reports that
compression of the submaxillary vein in the dog does not increase the
solids in the chorda saliva. It does not appear from the published
accounts that Eckhard actually made quantitative determinations. If
such determinations were made, the figures are not reported.

Zerner? diminished the blood supply to the dog’s submaxillary
gland by lowering the general blood pressure by transsection of the
spinal cord in the cervical region. He reports two such experiments,
finding in each case the chorda saliva richer in organic constituents
than normally.

Langley and Fletcher? found that diminution of the oxygen
supply by dyspnoea, as well as diminution of the arterial blood
supply by partial bleeding, increases the percentage of solids in the
dog’s chorda saliva from the submaxillary gland. The bleeding ex-
periments consisted in two trials on one dog only, and the secre-
tion was obtained by: pilocarpin injection. No statement is made
as to whether the increase in the percentage composition of the
chorda saliva under these conditions approximates the usual concen-
tration of sympathetic saliva. Langley and Fletcher point out that
their results suggest that the dog’s sympathetic saliva may be more
concentrated than the chorda saliva because of the diminished blood
supply to the gland onthe sympathetic stimulation. But they do not
conclude that such is actually the case, especially in view of the fact
that in the cat the submaxillary sympathetic saliva is normally more
dilute than the chorda saliva, — a fact discovered by Langley himself-
In his article on the salivary glands in Schafer’s Textbook, Langley
does not emphasize these results in accounting for the difference
between chorda and sympathetic saliva. In fact, he practically ignores
them by stating (p. 539) that “it has not been shown that the dimi-
nution of the blood supply causes any marked increase in the
organic constituents of the saliva.”

The usual difference in the percentage composition of chorda and
sympathetic saliva constitutes the main support of Heidenhain’s

1 ECKHARD: Beitrage, 1860, ii, p. 212.
2 ZERNER: Medizinische Jahrbicher, 1887, p. 534.
8 LANGLEY and FLETCHER: Philosophical transactions, 1888, clxxx, p. 109.

182 A. ¥. Carlson, ¥. R. Greer, and F. C. Becht.

theory of trophic secretory fibres to the salivary glands. If this
difference can be shown to be due to the difference in the blood and
oxygen supply to the gland, the theory becomes superfluous, if not
untenable; and in view of the conflicting and meagre data on this
point in the literature a re-examination of the whole subject seemed
to us very desirable.

2. Experimental methods. — Our experiments were confined to the
submaxillary salivary gland of the cat and the dog, most of the work
being done on the latter. The arterial blood supply to the gland was
diminished to any degree desired by compression of the submaxillary
artery. This method has the disadvantage that after isolation of the
artery in the dog the stimulation of the cervical sympathetic is usually
without any effect on the submaxillary gland, as it is difficult to
isolate the gland artery without severing the sympathetic branches
that pass from the superior cervical ganglion to the gland. When
sympathetic saliva is desired for comparisons, this must, therefore, be
obtained either before isolation of the gland artery or by stimula-
tion of the sympathetic fibres passing along the artery to the gland.
In some experiments the blood supply to the gland was controlled by
tying off both the vertebrals and one carotid and compressing the
other carotid to the degree desired. But in the ddg it is sometimes
not possible by this method to diminish the blood supply to the sub-
maxillary gland to the same extent that takes place on stimulation of
the sympathetic.

In all of our experiments on the effect of diminished blood supply
on the percentage composition of the chorda saliva, our aim was to
reduce the volume of blood passing through the active gland to that
which passes through the gland on stimulation of the sympathetic.
This can be done only approximately, as the degree of dimi-
nution of the blood supply to the gland on sympathetic stimu-
lation is variable. In general, the stronger stimulation causes the
greater vaso-constriction in the gland, and consequently the greater
diminution of arterial blood. But the same strength and character of
stimulus does not produce the same degree of vaso-constriction in all
animals, nor in the same animals at different periods of the same ex-
periment. These differences are, no doubt, due to the varying
excitability of the sympathetic fibres, the conditions of the gland
itself, the general blood pressure, and the degree of anzesthesia, — fac-
tors that cannot be made to run a uniformly parallel course. In our
experience stimulation of the sympathetic in the dog reduces the out-

Blood Supply to the Submaxillary Gland. 183

put of blood from the submaxillary veins to from one quarter to
one eighth of that in the resting gland.

In order to be sure of the relative amount of blood passing through
the gland at any period, we invariably measured the outflow of blood
from the main gland veins that empty into the external maxillary,
the facial, or, in the dog, directly into the external jugular vein. In
the dog a small vein leaves the hilus of the submaxillary gland along-
side the salivary duct. This vein was not taken into account in
our experiments. It can readily be tied off without injury to the
chorda tympani, but this was not necessary for our purpose, as we
only desired to reduce the arterial blood supply to that on stimulation
of the cervical sympathetic, and in all of our determinations of the
latter that hilus vein was not taken into account in measuring the
blood output. The output of blood from the gland veins was recorded
in drops on the kymograph bya magnetic signal. In the experiments
on the effects of tying the gland veins, the small vein leaving the
gland alongside the salivary duct was also usually left unobstructed.

The saliva was measured, dried, weighed, and finally incinerated in
porcelain crucibles. Py this method some of the volatile salts are
lost, so that the percentage of organic: material probably appears as
slightly greater than the actual. But this error is of no consequence
in our work, as only comparative figures are sought, and it would
seem probable that approximately the same relative amounts of vola-
tile salts should be lost in each sample of saliva.

Since the great work of Heidenhain on salivary secretion, it is well
known that the percentage of inorganic constituents in the saliva is in
part dependent on the rate of secretion and the strength of the stim-
ulus applied to the secretory nerves. Because of the nature of our
experiments these two factors could not be controlled so as to render
them uniform, and for that reason but little significance can be at-
tached to the variations in percentage composition of the inorganic
constituents that appear in our results.

Ether anesthesia was used exclusively in all of our experiments.

II. THe CERVICAL SYMPATHETIC IN THE DOG AND THE CAT CON-
TAINS SECRETORY FIBRES TO SUBMAXILLARY SALIVARY GLAND.

1. It is well established that the vaso-dilation following the stim-
ulation of the chorda tympani has no causal connection with the
secretion of saliva by the gland in consequence of the stimulation.

184 Alf Carlson, F..R- Greer, and £2 CBee.

The secretion is caused by the action of the secretory fibres on the
gland cells. The vaso-dilation is necessary for the secretion, to be
sure, but only in the way of supplying oxygen, water, and organic
pabulum to the gland cells. The old conception that the flow of
saliva from the submaxillary gland following the stimulation of the
cervical sympathetic is not a true secretory process, but simply
forcing out of the gland all or part of the saliva remaining from pre-
vious activity by the contraction of muscular elements in the
gland, has also been quite generally abandoned by physiologists in,
favor of the view that the sympathetic contains true secretory fibres.
But the existence in the cervical sympathetic of true secretory fibres
to the salivary glands is not so conclusively established as that of the.
cranial secretory fibres. Mathews,! in particular, has within recent
years urged the contraction hypothesis of the formation of sympa-
thetic saliva, however, without bringing any decisive experiment to
bear on it. Thecontraction hypothesis, which may be partly right,
is certainly wrong in its main contentions as far as regards the
submaxillary gland of the dog and the cat, because in these ani-
mals sympathetic secretory fibres to the gland can be conclusively
demonstrated.

The main evidence so far advanced for the existence of sympathetic
secretory fibres to the salivary glands is (1) the changes produced in
the gland cells on sympathetic stimulation; (2) the difference in the
character of chorda and sympathetic saliva; (3) the effect on the
composition of chorda saliva of a previous sympathetic stimulation ;
and (4) the paralysis of the secretory action of the cranial fibres by
atropin before the paralysis of the sympathetic secretory action.
This last evidence is, of course, indirect. If all the actions of the
sympathetic on the submaxillary gland can be shown not to be due
to secretory nerves, the supposed greater resistance of the sympa-
thetic secretory fibres to the action of atropin means simply that
atropin paralyzes the chorda secretory fibres or the gland cells before
it paralyzes the sympathetic fibres innervating the muscular elements
in the gland. It will be shown in a subsequent part of this paper
that the effect of sympathetic stimulation on the subsequent chorda
saliva in the dog, as well as the effect on the composition of the saliva
secured by simultaneous stimulation of the chorda and the sympa-
thetic, can be accounted for by the vaso-constrictor action of the

* MATHEWs: Annals of the New York Academy of Science, 1898, xi, p. 293 ;)
this journal, 1got, iv, p. 482.

Blood Supply to the Submaxillary Gland. 185

sympathetic. The argument based on the changes in the gland cells
on sympathetic stimulation remains so far unanswered. No one has
shown that similar changes can be produced by the supposed com-
pression of the cells on contraction of muscular elements in the
gland. The gland cells can, of course, be subjected to any degree
and persistence of pressure by massage of the gland.

That mere mechanical tension or pressure on the gland cells is
the cause of the histological changes in the cells produced by the
stimulation of the sympathetic seems so highly improbable that it
is hardly worth the while to stop to determine experimentally.
However, this has now been done. Mr. F. C. McLean has carried
through two series of experiments on the cat’s submaxillary gland.
In each experiment the gland on one side was left quiescent, while
the other gland was massaged or kneaded directly for twenty-five
to thirty minutes. The resting and the massaged gland were then
carried through the same fixing and staining processes. The results
were negative. The cells of the massaged gland exhibited the
same appearance as those of the resting gland. Pressure or ten-
sion on the cells does not bring about the cell changes caused by
sympathetic stimulation. The massage did not start any secretion
in the gland.

The main arguments advanced in support of the contraction theory
are: (1) the presence of muscle cells and myo-epithelial cells in the
wall of the salivary ducts and about the secreting tubules in some
salivary glands; (2) the small amount and relatively rapid cessation
of the saliva flow on stimulation of the sympathetic; and (3) the
back flow of saliva into the glands in some animals on cessation of
the sympathetic stimulation.

It is well known that the musculature of the gland arterioles takes
no part in the supposed squeezing process, as we may have pro-
nounced vaso-constriction in the glands without any flow of saliva
from the ducts. Furthermore, Kolossow,! who described myo-epithelial
cells in many glands, including some of the salivary glands, states
specifically that he did not find these elements in the submaxillary
salivarf gland. Nothing is known of the innervation of the muscle
cells in the walls of the salivary ducts.

The back flow of saliva into the gland on cessation of the sympa-
thetic stimulation is said to be very marked in the sheep’s parotid.
There is no back flow into the dog’s and the cat’s submaxillary on

1 Kotossow : Archiv fiir mikroscopische Anatomie, 1898, lii, p. 25.

186 A. ¥. Carlson, F. R. Greer, and F. C. Becht.

discontinuing the sympathetic stimulation, according to our obser-
vations, even when the tension in the duct amounts to from 10
to 20 mm. of saliva. The only explanation of a back flow into the
gland, on the secretory nerve theory, is this, that it is brought about
by the reabsorption of water from the saliva by the secreting
cells, and possibly the walls of the salivary ducts. If such a back
flow is very rapid and takes place in the absence of pressure, it is
probably not entirely due to the reabsorption of water, although we
have as yet no means of knowing how rapidly the secreting cells may
take up water from a hypotonic solution like saliva when the stimuli
enabling them to eliminate this dilute solution cease. Nothing can
be concluded from the fact that liquid may be forced into the gland
through the salivary duct under greater or less pressure. Such
pressure will necessarily distend the tissues in the gland, and perhaps
by ruptures cause some of the liquid to escape by way of the lymph
spaces and efferent lymph ducts.

2. The difference in the percentage composition of chorda and sym-
pathetic saliva cannot be explained by the contraction hypothesis. —
(1) What are the conditions in the cat? Langley showed long ago
that the submaxillary sympathetic saliva in the cat is more dilute
than the chorda saliva, and that the greater dilution is mainly due to
the smaller amount of organic constituents present. We have con-
firmed Langley on this point, as will be shown in a later section of
this paper. The sympathetic submaxillary saliva in the cat is usually
more dilute than the chorda saliva. On the contraction hypothesis
the sympathetic saliva is, of course, nothing but chorda saliva forced
out of the gland. This renders necessary the improbable assumption
that the normal chorda saliva has been rendered more dilute on stand-
ing in the gland by the reabsorption of some of the organic constitu-
ents by the gland cells or by the cells of the salivary ducts. But, as
a matter of fact, this dilute sympathetic saliva in the cat is obtained
even when the chorda stimulation is immediately followed by stimu-
lation of the sympathetic, so that there is not even a cessation of the
saliva flow while the electrodes are being shifted from the chorda to
the sympathetic. Consequently there is no time for reabs@rption.
Nor could such reabsorption go on during the stimulation of the
sympathetic, as the sympathetic saliva flow from the cat’s submaxillary
gland is relatively rapid.

(2) How does the contraction hypothesis account for the concen-
trated saliva in the case of the dog’s submaxillary gland? It is well

Blood Supply to the Submaxillary Gland. 187

known that under certain conditions the submaxillary sympathetic
saliva in the dog may be as dilute as the chorda saliva, but ordinarily
‘the former is much richer in organic constituents than the latter.
The contraction hypothesis demands that sympathetic saliva follow-
ing immediately after chorda stimulation be of the same character
and concentration as the chorda saliva. A more concentrated saliva
could only be obtained after a period of rest of the gland by assum-
ing that during rest water and salts are taken up from the saliva
by the gland cells. Neither of these demands stands the test of
experiment.

It is true, however, that after fifteen to thirty minutes’ quiescence
of the dog’s submaxillary gland the first 0.5 c.c. of saliva following
chorda stimulation is more concentrated than the subsequent portion
of saliva secured during the same period of stimulation. The greater
concentration of the initial portion is practically wholly due ta the
greater amount of organic matter present. We made six experiments
on this point. Our results are recorded in Table I. The data of only
five of these experiments are given, because, through an accident in
manipulation, one series yielded reliable figures for the organic con-
stituents only. But so far as these figures go they are in complete
agreement with those given in Table I. And inspection of this table
shows that out of the seventeen comparisons made fourteen cases
show a much greater concentration of the organic solids in the initial
portion of the saliva, in two cases this relation is reversed, while in
one case there is practically no difference. The four comparisons of
the experiment not tabulated all show a greater concentration of the
organic constituents.

There is no constant or preponderant difference in the inorganic
salts. The variations in the concentrations of the salts are probably
due mainly to the fact that their percentage content depends on the
rate of the secretion, and the rate is not constant during any one
period of stimulation.

These figures demonstrate the further fact that the first 0.5 c.c. of
chorda saliva from the dog’s submaxillary gland after a period of rest
is in most instances as rich in organic constituents as the sympa-
thetic saliva of the same gland.

What is the cause of this greater concentration of the organic
solids in the initial portion of chorda saliva after gland rest ? Noth-
ing seems more natural than to suppose that the greater part of the
initial 0.5 c.c. of saliva represents that portion remaining in the gland

188 ~4.\ 7. Carlson, 7. .R. Greer, and LC. Bach.

from the previous period of activity, and that this has suffered con-
centration by the reabsorption of water and salts to the exclusion of
the organic constituents. If this is the correct explanation, there’
should be a back flow of chorda saliva into the gland during the
period of rest, but so far as we know no such back flow has been re-
corded for the chorda saliva of the dog’s submaxillary gland, and we
failed to find any indication of it in our experiments.

But even assuming that the concentration is due to the reabsorp-
tion of water and salts in the manner just stated, our experiments
show conclusively that this does not account for the concentrated
sympathetic saliva so as to render the secretory nerve theory super-
fluous. If the sympathetic saliva is simply chorda saliva forced out
of the gland by muscular contraction, the sympathetic saliva taken
immediately after cessation of chorda stimulation, so as to give no
time for the assumed reabsorption, should be as dilute as the chorda
saliva immediately preceding it. This is not the case. If on cessa-
tion of the chorda stimulation the cannula is quickly emptied of
chorda saliva, the stimulation of the sympathetic immediately begun,
and the first two or three drops of saliva flowing from the cannula
discarded as being undoubtedly chorda saliva, the sample obtained is
typical concentrated sympathetic saliva. The figures proving this
point do not stand out so clearly in Table I, as most of the data
in that table bear ona different question. They are therefore restated
in Table II. The sample of sympathetic saliva is in every case taken
immediately after cessation of the chorda stimulation so as not to
allow time for the assumed reabsorption.

In the absence of evidence of any considerable reabsorption of
water and salts from the chorda saliva, the course of the concentra-
tion of the initial chorda saliva after a period of gland rest is probably
the following. It is well known that the dog’s submaxillary chorda
saliva becomes gradually poorer in organic constituents on repeated
stimulations. The same relation probably holds good for even single
periods of activity, the first part of the saliva being more concentrated
than the later portion. Itis also probable that on chorda stimulation
in the dog the first 0.5 c.c. of saliva is secreted under condition
of insufficient amount of oxygen. The transfer of oxygen from the
capillary walls to the gland cells must take place through the lymph of
the tissue spaces, and is probably a process of diffusion. The increased
oxygen supply to the cells will therefore lag behind the time of action
of the secretory nervous impulses on the gland cells. And it will be

Blood Supply to the Submaxillary Gland. 189

shown in a subsequent part of this paper that deficiency in the oxygen
supply to the active gland increases the organic solids in the saliva,
or rather decreases the rate of secretion of water and salts.

3. The contraction hypothesis is rendered untenable by the amount
of saliva that may be obtained from the dog’s submaxillary gland in a
single period of uninterrupted stimulation of the cervical sympathetic.—
Muscular contraction cannot force out any more saliva than what is
stationary in the secretory tubules and in the salivary ducts, and
it is practically certain that it could not even force out all of that.
It is obvious, from the anatomical structure of the submaxillary gland,
that this stationary saliva can never reach the amount of one fourth
the total volume of the gland. In the dog stimulation of the sympa-
thetic usually yields only a small quantity of saliva from the submax-
illary gland, and the advocates of the contraction hypothesis point to
this as an evidence in support of their contention. It will be shown
later that one cause, and probably the main one, of the usual quick
cessation of the flow of sympathetic saliva in the dog is the vaso-
constriction accompanying the secretion. If the arterial blood sup-
ply to the gland is greatly diminished during chorda stimulation, the
chorda saliva is scarcely any more abundant than sympathetic saliva.

In some cases, however, a single period of continuous stimulation
of the cervical sympathetic with the weak interrupted current will
yield from 1 to 3 c.c. of saliva. The flow of the saliva is slow, to be
sure, but constant, and in order to obtain the stated amounts the
stimulation must be kept up for from ten to fifteen minutes. The
greatest quantity of sympathetic saliva: obtained during a single
period of stimulation was 5.5 c.c. The period of stimulation was
eighteen minutes. The stimulation was discontinued, not because
the flow of saliva had stopped, but because we thought we had
already secured a quantity of saliva equal to the volume of the
gland. But that was not the case. The gland, freed from connec-
tive tissue, had a volume of 6.5 c.c. But even so, this experiment
amounts to a reductio ad absurdum of the contraction hypothesis,
for by a single period of stimulation more water was ‘“‘ squeezed” out
of the gland than was contained in the whole gland tissue, and at the
end of the stimulation the same quantity of water was still left in the
gland! The gland was quiescent at the beginning of the stimula-
tion, and the saliva flow ceased immediately on discontinuing the
stimulation.

On the contraction hypothesis we must assume that the contractile

190 As 3. Carlson,, 7.\R? Greer,.and i. CiMecks.

TABLE TI.

Submaxillary saliva of dog. Percentage composition of the first 0.5 c.c. of chorda
following stimulation of the cervical sympathetic. The numbers

Num- ; Solids per 100 cc.
eld Saliva.
ment. Total. Inorganic. | Organic.
I 1. Chorda, reflex secretion .. . 1.14 0.32 0.72
2. Chorda, after 18 minutes’ rest . 1.51 0.45. - 1.06
3. Chorda, immediately following 2 1.40 0.21 1.19
4. Chorda, after 27 minutes’ rest . 1.68 0.56 a
5. Chorda, immediately following 4 1.16 0.22 0.94
6. Sympathetic, after 5 minutes’ rest 2.11 0.31 1.80
7. Chorda, immediately following 6 ivy 0.26 0.91
8. Chorda, after 18 minutes’ rest . 1.43 0.23 1.20
9. Chorda, immediately following 8 1.42 0.34 1.08
10. Chorda, after 24 minutes’ rest. 1.82 0.26 1.56
11. Chorda, immediately following 10 1.34 0.20 1.14
12. Sympathetic, immediately after 11 Loz 0.34 1.58
II 1. Chorda, reflex secretion .. . 0.85 0.21 0.64
2. Chorda, after 1§ minutes’ rest . 1.65 0.42 1.23
3. Chorda, immediately after2. . 1.24 0.33 0.91
4. Chorda, after 25 minutes’ rest . 1.51 0.40 1.11
5. Chorda, immediately after4 . . 1.46 0.22 1.24
6. SympatBetic. ss =) x) 220 hse, =, Ge 1.92 0.28 1.64
7. Chorda, immediately after6. . 1.30 0.27 1.03
8. Chorda, after 20 minutes’ rest . eS 0.26 1.47
9. Chorda, immediately after8 . . 1.14 0.41 0.73
10. Sympathetic, immediately after 9 1.92 0.34 1.58
JOE 1. Chorda,atter@nest fe. en eee 1.44 0.28 1.16
2. Chorda, immediately after]. . 1.16 0.28 0.88 _
3. Chorda, after 15 minutes’ rest . 1.80 0.26 1.54
4. Chorda, immediately after3 . . 1.40 0.33 1.07

Blood Supply to the Submaxillary Gland. IgI

ABCn Is

saliva after periods of rest of the gland with that of succeeding saliva, and with the saliva
indicate the sequence of taking the samples in each experiment.

Num- Solids per 100 c.c.
ae et Saliva.
ment. Total. Inorganic. | Organic.
s0l 5. Chorda, after 15 minutes’ rest . 2.18 0.25 1.93
6. Chorda, immediately after5 . . E72 0.30 AZ,
7. Chorda, after 15 minutes’ rest . 1.58 0.26 1.32
8. Chorda, immediately after7. . 1.31 0.26 1.05
9. Chorda, after 15 minutes’ rest . a7 0.23 1.49
10. Chorda, immediately after9. . 1.08 0.24 0.84
11. Sympathetic, immediately after 10 1.84 0.28 1.56
12. Chorda, immediately after 11 . 1.15 0.26 0.89
13. Sympathetic, immediately after 12 1.70 0.21 1.49
IV i hoe after 20 minutes’ rest ~. 1.95 0.41 1.54
2. Chorda, immediately after] . . 2.09 } 0.41 1.68
3. Chorda, after 20 minutes’ rest . 3.51 0.46 2.05
4. Chorda, immediately after3 . . 1.54 0.51 1.00
5. Sympathetic, immediately after 4 2.52 0.56 1.96
6. Chorda, immediately after5 . . 1.67 | 0.43 1.24
7. Sympathetic, immediately after 6 1.88 0.59 1.29
8. Chorda, immediately after7. . TELZ 0.53 0.59
V 1. Chorda, after 20 minutes’ rest . 3.26 0.52 2.74
2. Chorda, immediately after] . . 2.33 0.33 2.00
3. Chorda, after 15 minutes’ rest . 2.39 0.49 1.90
4. Chorda, immediately after3. . 1.66 0.55 if alal
5. Sympathetic, immediately after 4 3.06 0.66 2.40
6. Chorda, immediately after5 . . (1.45 0.74
7. Sympathetic, immediately after 6 1.82 0.39
8. Chorda, after 20 minutes’ rest . hey (AL 0.49
9. Chorda, immediately after8 . . 1.36 0.50

192 «= Af. Garlson, FR. Greer, and F Gy Berni,

elements in the gland may require from ten to twenty minutes to
reach their maximum degree of shortening on stimulation of their mo-
tor nerves, for the saliva may flow from the duct for that length of time
on continuous stimulation of the sympathetic. And with the salivary
duct open the flow of saliva will cease practically with the end of the
muscular contraction, that is, when the muscles have reached the maxi-
mum degree of contraction. We know of no muscular elements that
require that length of time to reach the maximal degree of shortening.

The existence of sympathetic secretory fibres to the submaxillary
gland of the dog and the cat is therefore certain. Contraction of
muscular elements in the gland may play a part at the beginning of
the expulsion both of chorda and sympathetic saliva, but the only real
evidence for this is the existence of muscular elements in the walls
of the salivary ducts and between the glandular tissues.

III. DimInuTION OF THE BLoop SUPPLY TO THE SUBMAXILLARY
GLAND DURING ACTIVITY BY COMPRESSION OF THE GLAND
ARTERY GREATLY INCREASES THE PERCENTAGE OF ORGANIC
CONSTITUENTS IN THE SALIVA.

1. The results of our nine experiments, seven on dogs and two on
cats, are recorded in Table III. The two experiments on the cat
were made for us by Mr. G. Ryan and Mr. F. C. McLean. These
results show that when the arterial blood supply to the gland during
chorda stimulation is reduced approximately to that on stimulation
of the sympathetic, the chorda saliva becomes just as concentrated as
sympathetic saliva. Within limits, the greater the diminution of the
blood supply, the greater the percentage increase in the solids. These
figures show, furthermore, that the greater concentration of this
chorda saliva is due practically entirely to the organic constituents.
Even to the naked eye this chorda saliva appears thick and viscid
and in every way similar to the dogs’ sympathetic saliva, and the
quantitative determinations show that this is actually the case.

No conclusions can be drawn touching the concentration of the
salts in these different salivas, as our results show that these may
be practically the same (Nos. 2, 7), a little greater (Nos. I, 3, 6),
ora little less than in the normal chorda saliva. These differences
in the salt content of the saliva obtained under these conditions are
probably in the main due to differences in the rate of secretion
and in the strength of the stimulus, but we have no data by
which to decide the point at present.

Blood Supply to the Submaxillary Gland. 193

2. The chorda saliva secreted under greatly diminished blood
supply to the gland also resembles sympathetic saliva in the quantity
and the rate of secretion. It was noted by Langley and Fletcher
that on partial bleeding of the dog the quantity of saliva obtained on
chorda stimulation was less than the normal.
on this point confirm those of Langley and Fletcher.

Our observations
With the

TABLE II.
Organic constituents of dog’s submaxillary chorda saliva compared with that of the

sympathetic saliva immediately following it.

Organic solids
per 100 c.c.

Exper-
iment.

Organic solids
per 100 c.c.

Chorda 1.14 Chorda 1.00

2 Sympathetic

Chorda

2 Sympathetic
Chorda
Sympathetic

‘Chorda

2 Sympathetic

Sympathetic
Chorda
Sympathetic
Chorda
Sympathetic
Chorda

Sympathetic

1.96

arterial blood supply to the gland almost shut off in the dog, stimula-
tion of the chorda produces only a few drops of saliva, and after
periods of rest subsequent stimulation of the nerve yields gradually
less saliva. We have here, apparently, a duplication of the usual
rapid failure of the secretory nervous mechanism typical for the
sympathetic secretory fibres in the dog.

3. On greatly diminishing the blood supply to the dog’s submax-
illary gland, stimulation of the chorda diminishes for a time the
output of blood from the gland veins. In some of our experiments
the output at the beginning of the stimulation was reduced to one
third that preceding the stimulation. Two factors contribute to this
anomalous condition, namely, the quantity of water that leaves the
blood to enter the lymph and ultimately the saliva, and the active
dilation of the arterioles due to the stimulation of the vaso-dilators in
the chorda. The first factor plays in all probability only a small part,
inasmuch as the chorda stimulation produces only a scanty secretion

194 AF. Carlson;. J: R. Greer, and Fase Beene.

when the blood supply is greatly diminished. The main factor is
the vaso-dilation, as shown by the fact that the diminution in the
venous output from the gland may be nearly as great after paralysis

FicurE 1.— About two thirds the original size.. Dog. Record of blood flow through
submaxillary gland and of salivary secretion on stimulation of the chorda tympani
nerve during partial occlusion of the gland artery. a, stimulation of chorda. 4, record
of saliva flow in drops. c, record of blood flow (in drops) from the submaxillary
vein, showing diminution of flow of blood from the gland on stimulation of the
chorda. Time, seconds.

of the secretory mechanism by atropin as before. When the bore of
the gland artery is greatly diminished, the greater quantity of blood
entering the gland on chorda stimulation, owing to diminished resist-
ance in the arterioles, is not great enough to counterbalance the space

Gre tt

QQ] ——_______—_$§ eer 35 $$

B

Figure 2.— About one half the original size. Dog. Record of blood flow from submax-
illary gland vein on stimulation of the chorda tympani nerve during partial occlusion
of the gland artery. A, before, 2, after, paralysis of the secretory fibres by atropin.
a, stimulation of chorda. 4, record of blood flow in drops from submaxillary vein.
c, record of saliva flowin drops. Showing the diminution in the venous output on
chorda stimulation after paralysis of the secretion by atropin. Time, seconds.

volume yielded by the dilation of the arterioles; and until that
space is filled up there will necessarily be a lessened flow through the
capillaries and the vein. When the dilated arterioles are filled, we
have a return to the original rate of output from the gland veins, or
it may even be slightly increased. Records illustrating these condi-
tions are reproduced in Figs. 1 and 2.

When the stimulation of the chorda reduces the blood flow from
the gland vein, this condition is necessarily accompanied by dimin-

Blood Supply to the Submaxillary Gland. 195

TABLE, Ti:

The effect on the percentage composition of the dog’s submaxillary chorda saliva of

greatly reducing the arterial blood supply to the gland by partial occlusion of the
gland artery.

Solids, per 100 c.c.

2 etl Saliva, submaxillary gland.
Total. Inorganic. Organic.
I. Dog eSympathetier. 2 )-) ts ee 0 40 1.51
DeGhordat sos) silt; cats 1.16 0.40 0.76
3:) Chorda, art, oeels =. 5+... 2.28 0.50 1.78
4 Chaorday,. soca la tint sso 1.13 0.31 0.82
II. Dog Peoympathetioe % > 3). «. 2.26 0.59 1.67
PS ACU RE CE Te ae th Ae ROG te 1.10 0.44 0.66
ee HordanaLtsOCel.a eae 2.31 0.43 1.88
quChorda® Se Be ses ks Ye 1.20 0.31 0.89
Ill. Dog le Tees itere (ee) 4 4 ol Be 0.89 0.24 0.65
ZLoyimipathene.. “ys 3) % 1.23 . 0:33 0.90
25 (Chive es, 6-80 ae pee 132 0.61 O71
4. Chorda, partial occl. art. . se 0.77 0.95
5. Chorda, greater occl. art. . 2.03 0.64 1.38
IV. Dog il; INSU .<(GWiy)) we eo a 3 0.89 0 24 0.65
2a SYMP APetlG, herent) 3! tke 1.88 0.40 1.48
3h: ACINVONGE NS Geo. eeg Lee pee 1.73 0.79 0.93
4. Chorda, partial occl. art. . 1.67 0.69 0.98
5. Chorda, greater occl. art. . 2.05 0.64 1.45
V. Dog PChordabg) y saree ty ou letere 1.62 0.56 1.05
27 SymMpathetics ss 2s = 1.80 0.49 1.30
Se GnROrdasant-OCGle cnn) ver ZA 0.51 1.60
Aon GONG aw-et weaket Pores, s,s 1.61 0.43 1.21
VI. Dog Exroyiapatheties., i) ms) 6 193 0.40 1.52
Po iCiventdey pa ie oJao: Moles aa oe 1.36 0.44 0.91
3: Ghordayartsocel. | . . ; Ptigl 0.59 2.18
ASIGHOLOAS -wasW Sodaes vs cone 1.84 0.49 1.34
Nite Mor eal rGhorda.. 4 caso 2) & \° 0.94 0.19 Os.
Ze GChorda, artaoccly 3.) .. : 1.68 0.21 1.48
SmCUGKday.: Vee Yes «os eee 0.26 0.96
PpeilGat, |) 8 Chtorda 2 5. Liesl
2 Chordah ant joeclia) & < % 2.25
ord COLCA RS rs a tee stat hts 1.61
IX. Cat IP KGCGKGE: ose eas We eee 1.76
2 Ghordayart..0Ccclio. .. 7. |. 241
Si. Kh oh eh Bis ae 1.39

196 A. 4: Cartson, Ff. Ri Greer, aa Te C. Becht.

ished tension and rate of flow of the blood through the capillaries.
Consequently we have here a condition of glandular secretion parz
passu, with reduced rate of flow and pressure of the blood in the gland
capillaries. |
4. The effect of a period of diminished blood supply on the
percentage composition of the chorda saliva obtained after restora-
tion of the normal circulation is not conclusively determined by our
experiments, as will be seen on examination of Table III. In four ex-
periments (2, 6, 7, 8) the chorda saliva obtained after re-establishing
the normal circulation is more concentrated than the normal, and
the main factor is the greater amount of the organic constituents.
In two experiments (1, 5) there is practically no difference, while in
one case (9) the chorda saliva subsequent to readmitting the blood
is more dilute. When the blood is readmitted to the dog’s submax-
illary gland, after a few minutes’ complete occlusion the gland secretes
spontaneously for a time, as noted by Mathews,! and this spontaneous
saliva is very dilute, as compared to normal chorda saliva. Quan-
titative determinations in two experiments gave the following figures
per 100 c.c. of saliva:
1. Chorda saliva before occlusion— solids: 1.16 | Oueee are
: = Inorganic: 0.20
2
o4 I.
Bs

2. Spontaneous saliva after occlusion—solids: 0.69 Organic: 0.49

Inorganic: 0.44
Organic: 0.66

1. Chorda saliva before occlusion— solids: 1.1
2 . :
2. Spontaneous saliva after occlusion —solids : 0.63 / = On O38
The spontaneous saliva following the re-establishment of the gland
circulation after anemia is therefore very poor in both organic and
inorganic constituents. The mechanism of this saliva formation is
not known. It is not reflex. It may be due to substances accumu-
lated in the gland cells or the surrounding lymph during the anemia,
in which case the secretion would be analogous to the transient
heightened excitability and stimulation of muscle cells and nerve
centres on readmitting the blood after anemia. The ganglion in the
hilus of the gland as well as the secretory nerve fibres themselves
may also be involved in this secretion.
The work of Heidenhain and Langley seems to show that stimula-
tion of the sympathetic increases the organic constituent in the sub-

1 MATHEWS: Loc. cét.

Blood Supply to the Submaxillary Gland. 197

sequent chorda saliva. In our experiments this is not invariably
the case. Before all the differences between chorda and sympathetic
stimulations as regards the saliva can be conclusively ascribed to the
differences in the blood supply, it must be shown that diminished
blood supply to the gland concentrates the subsequent chorda saliva.
Our data indicate that such is the case in some of the experi-
ments. There is only an apparent discrepancy between this fact
and the fact that the spontaneous saliva following anemia is so
abnormally dilute, as the condition of the gland is necessarily very
different in these cases, and the mechanism of the secretion may be
entirely different.

IV. DimInuTION oF THE OxYGEN SUPPLY TO THE GLAND BY OCc-
CLUSION OF THE GLAND VEINS INCREASES THE ORGANIC Con-
STITUENTS IN THE CHORDA SALIVA.

1. The data of our seven experiments on this point are recorded
in Table IV. The vein occlusion increases the percentage of solids
in the chorda saliva, and this increase is mainly due to the organic
constituents. In only one case (No. 3) does this chorda saliva attain
a concentration equal or exceeding that of the sympathetic saliva
from the same gland.

In the dog and the cat the chorda tympani remains functional for
over an hour with the main gland veins occluded. Such occlusion
does not completely stop the gland circulation. Some oxygen
reaches the gland by mere diffusion. And the fact that venous oc-
clusion greatly increases the output of lymph also increases the arte-
rial blood supply tothe gland. In the dog, moreover, the small
vein alongside the salivary duct was not occluded, and owing to the
greater pressure in the gland veins, the output of blood from this vein
was probably as great as the total venous output from the gland
on sympathetic stimulation. The condition in the cat was not
closely investigated, but on tying the main gland veins the chorda
remains functional for a relatively longer time than in the dog.

There can be no question, however, but that with the main gland
veins occluded the oxygen supply on chorda stimulation is much less
than on chorda stimulation with the circulation unimpeded. It is
probably even less than that of the resting gland. The capillary
pressure is at the same time greatly increased. The gland cells have
a supply of water, salts, and organic pabulum even greater than under

198 A. F Carlson, F. R. Greer, and F. C. Becht.

normal conditions, the only deficiency being oxygen. And this de-
ficiency makes the chorda saliva alter its percentage composition
in the direction of that of sympathetic saliva.

2. Occlusion of the gland veins diminishes the quantity of the
chorda saliva, and hastens the fatigue or failure of the chorda secre-
tory fibres, just as we found in the case of compression of the
gland artery. The longer the period of venous occlusion, the greater
the diminution in the quantity of chorda saliva obtainable. This
relation appears also to hold for the concentration of the saliva; that
is, the longer the period of Occlusion, the greater the percentage of
organic constituents (Table IV, No. 7).

3. Our experiments were not directed with the view of determining
whether the chorda saliva obtained after a period of venous occlusion
is more concentrated than the chorda saliva prior to the occlusion.
Our data are indecisive on this point. The first three samples of
chorda saliva in Experiment 1, Table IV, are not complicated by
sympathetic stimulation, and in that particular case the chorda saliva
following the venous occlusion is much richer in organic solids than
that obtained prior to the occlusion, in fact, even more so than the
saliva collected during the occlusion. This point needs further
investigation.

V. Doers THE SYMPATHETIC SALIVA IN THE DOG BECOME AS DILUTE
AS THE CHORDA SALIVA ON CONDITION OF INCREASE IN THE
OxYGEN SUPPLY TO THE GLAND?

1. We have shown that diminution in the oxygen supply to the
submaxillary gland by arterial compression and venous occlusion ren-
ders chorda saliva of the same composition as that of the sympathetic
saliva. This furnishes strong presumptive evidence that the cause of
the greater concentration of the submaxillary sympathetic saliva in
the dog is the attendant vaso-constriction. Our data do not consti-
tute a proof of this point, however. This hypothesis cannot be con-
sidered proved until it has been shown that, given an ample blood
supply, the sympathetic saliva becomes as dilute as normal chorda
saliva. So far as we know, no experiments have been reported on this
phase of the question. :

We attempted, in the first place, to separate the secretory and the
vaso-motor fibres in the cervical sympathetic by the degeneration
method. Dogs were used for these experiments. The cervical sym-

Blood Supply to the Submaxillary Gland.

TABLE IV.

199

The effect on the composition of the submaxillary chorda saliva of the occlusion of the
submaxillary veins.

No. of
experiment.

_ I. Dog

II. Dog

III. Dog

IV. Dog

V. Dog

Wiis Cat

VII. Cat

nn FF WO DW

_

->- WwW Nw S&H W NY“ WD HY &

ea ae ets

Saliva, submaxillary gland.

. Chorda .

. Chorda, veins occl.
. Chorda .

. Sympathetic .

. Chorda .

. Chorda, veins occl.
. Chorda .

. Chorda, veins occl.
. Sympathetic .

. Chorda .

. Chorda .

Sympathetic .
Chorda .

Chorda, veins occl.
Chorda .

Chorda, veins occl.

. Chorda .

. Chorda, veins occl.

. Sympathetic .

. Chorda .

. Sympathetic, slight vaso-constr.
. Chorda, veins occl.

. Sympathetic, slight vaso-constr.
. Chorda .

. Chorda, veins occl.

. Chorda, after 60 minutes’ occl.

Solids, per 100 c.c.

Inorganic, | Organic.
0.38 0.64
0.45 129
0 43 1.42
0.53 EOS
0.40 1.23
0.38 1.36
0.32 0.78
0.35 1.12
0.53 1.93
0.39 1.30
0.52 1.69
0.55 ese,
0.55 1.45
O40 Z.59
0.25 0.93
0.28 1.33
0 26 0.96_
0.36 Aalt
0.29 1.83
0.10 1.00
0.12 0.96
0.15 1.15
0.10 0.80
0.18 0.72
0.12 1.10
0.26 139

200 «A. F. Carlson, 7. R. Greer, and F.C. Becht.

pathetic was transsected in the neck, and its secretory and vaso-motor
action on the submaxillary gland investigated two to four days after
the operation. The cut nerve ceases to act on the submaxillary gland
on the third day after the section. Furthermore, the vaso-motor and
the secretory fibres lose their function practically at the same time.
This experiment was repeated on four animals with the same results,
and the method, not promising to yield any result, was abandoned.
If the secretory fibres should have ceased to function sooner than
the vaso-motor, the method would have been equally useless for our
purpose.

We next endeavored to overcome the vaso-constrictor action of the
sympathetic by simultaneous stimulation of the chorda after paralysis
of the chorda secretory fibres by atropin. It is well known that the
constrictor action of the sympathetic can be partly overcome in this
way. In the three experiments made according to this method, the
saliva obtained was less concentrated than normal sympathetic saliva.
But the atropin diminishes the activity of the sympathetic secretory
fibres also, so that the total quantity of saliva obtained in each case
was small. It has been shown by Bancroft,! moreover, that while
atropin paralyzes the secretory activity of the chorda, it does not stop
all of its action on the gland cells, because the stimulation of the chorda
‘ after atropin paralysis greatly increases the CO, output of the gland.
It is obvious that the chorda influences the chemical processes in the
gland, although it does not produce any secretion. After atropin in-
jection, the gland is not under normal physiological conditions. The
data obtained by this method would therefore not be conclusive.

A method of more promise than the foregoing ones is to perfuse
defibrinated blood through the gland artery under sufficient pressure
to overcome the vaso-constrictor action of the sympathetic. This is
easily accomplished; defibrinated blood may be forced through the
gland artery under mercury or air pressure, regulated to completely
counteract the sympathetic action. But under this degree of pres-
sure the gland becomes cedematous very quickly.

As perfusing liquids, we made use of defibrinated dog blood, de-
fibrinated ox blood, Locke’s solution under oxygen pressure, and
Locke’s solution plus oxygenated dog corpuscles. The artificial so-
lutions are useless for this experiment, because, when forced through
the gland under sufficient pressure to overcome the vaso-motor action
of the sympathetic, they produce cedema in the gland almost imme-

1 BANCROFT: Journal of physiology, 1901, xxvii, p. 37.

Blood Supply to the Submaxillary Gland. 201

diately. The gland is therefore in a very abnormal condition, and
the results obtained from such a gland cannot be made use of in ex-
plaining the mechanisms of the processes of the normal gland. Both
the chorda and the sympathetic secretory fibres remain functional
for some minutes with the gland in maximum cedema from perfusion
of a corpuscle suspension in Locke’s solution, but the quantity of
both the chorda and the sympathetic saliva is greatly reduced.

Defibrinated and oxygenated dog and ox blood is the most satis-
factory perfusing medium. But even the dog blood perfused under
this pressure renders the gland cedematous in a few minutes. The
normal mechanism of the capillary endothelium is evidently disar-
ranged under this pressure, so that more fluid gets into the lymph
spaces than can escape through the lymphatics leading from the
gland. In all, three trials were made, using defibrinated dog blood
as perfusing medium. Small quantities of saliva were obtained by
sympathetic stimulation, but this saliva was not identical with normal
chorda saliva. It was more concentrated and viscid, thus resembling
ordinary sympathetic saliva. Actual quantitative determinations
were not made, for the reason that the data would not be conclusive
in any event.

It might seem that these results indicate that the sympathetic sub-
maxillary saliva in the dog does not change to the composition of
chorda saliva on the blood and oxygen supply to the gland being
ample. But they do not constitute a proof of this, for the following
reasons. Although the gland is amply supplied with blood and oxygen,
it cannot be said to work under normal conditions, because of the
cedema rapidly developed by the excessive pressure in the capillaries.
The gland is, furthermore, rendered anemic for the time required to
make the necessary connections with the gland artery for the
perfusion. And we have seen that anemia of the gland may increase
the concentration of the chorda saliva obtained on subsequent stimula-
tion. The method was abandoned, as there seemed to be no way of
eliminating these two objections.

Some experiments were tried with the circulation in the dog’s sub-
maxillary gland reversed. The carotid was united with the external
_ jugular vein by means of a cannula, and all the vein branches tied
off, save the main submaxillary vein. A cannula was also placed in
the submaxillary artery so as to measure the rate of the blood flow
through the gland in the reversed direction. We reasoned that if the
circulatory bed in the gland offers no resistance to the reversed flow

202 A. ¥ Carlson, F R. Greer, and F. C. Becht.

in the way of valves or other obstructions, the constriction of the
arterioles by sympathetic stimulation would not cut down the blood
supply as much as under normal conditions, because of the wider bed
and therefore less resistance on the venous side of the capillaries.
The tension in the capillaries on chorda and sympathetic stimulation
with the circulation reversed is, of course, the reverse of the nor-
mal, the sympathetic stimulation increasing, the chorda stimulation
decreasing it.

The net result of these experiments is this, that only a very small
amount of blood can be made to flow through the dog’s submaxillary
gland in the reversed direction. This is true for the first two hours
of reversal. We did not continue any of our experiments beyond that
time. Although the full pressure of the carotid blood is exerted on
the blood in the submaxillary vein, only a few drops per minute flow
from the cannula in the submaxillary artery. In fact, the gland
becomes rapidly asphyxiated.

The cause of this great resistance to the reversed blood flow is not
obvious. Thrombi in the gland vessels are probably a factor in some
cases, as a slightly greater flow was secured in the two experiments
in which we rendered the blood non-coagulable by intravenous in-
jection of hirudin. But even with the coagulation factor removed
the best that we could secure was 5 to 6 drops per minute. And
this was in relatively large dogs. “his amount of blood is even less
than that normally passing through the gland on stimulation of the
sympathetic.

Both the chorda and the sympathetic remain functional for some
time after reversing the circulation, but the saliva obtained is very
scanty and very concentrated. This is to be expected, as the rever-
sal gradually asphyxiates the gland, and we have shown that diminu-
tion in the oxygen supply to the gland greatly increases the organic
solids in the saliva. The following is the analysis of the saliva
samples from one of the experiments in which hirudin was used.
The percentage of the organic solids was not determined, but it was
evident on mere inspection of the saliva that the great concentration
was due to the organic constituents.

Normal sympathetic saliva : 2.49
Normal chorda saliva : 1.80 | Total solids per
Sympathetic saliva, circulation reversed: 3.24{ 100 c.c. saliva.
Chorda saliva, circulation reversed : 5:60

bh WN HW

Blood Supply to the Submaxillary Gland. 203

The chorda saliva on reversal of the circulation is nearly twice as
concentrated as the sympathetic saliva. The reason for this is
obvious. The chorda saliva is collected after the sympathetic sample.
The condition of diminished blood supply to the gland has conse-
quently lasted longer, and we have shown that the greater the
degree of asphyxia, the greater the concentration of chorda saliva.

TABLE V.

Comparison of the composition of chorda and sympathetic submaxillary saliva in the
cat, and the effect of diminished blood supply to the gland on the percentage of
solids in the sympathetic saliva. j

Solids per 100 c.c.

-
ee Saliva from submaxillary gland.

experiment.

Inorganic. | Organic.

. Chorda .
. Sympathetic (in neck)
. Chorda .
. Sympathetic (in neck)
. Chorda .

. Sympathetic, compression of
subm. artery

. Sympathetic, compression of
subm. artery

. Sympathetic, great vaso-con-
striction

. Chorda .

1. Chorda .

. Sympathetic (on gland artery)
vaso-constriction

. Chorda.

2. None of the foregoing methods yielded conclusive results. But
in the cat nature has apparently furnished conditions for deciding this
question without resorting to any artificial contrivance. It was dis-
covered by Langley! that the submaxillary sympathetic saliva in the
cat is usually even more dilute than the chorda saliva. This we can
confirm. When the sympathetic is stimulated in the neck with a
weak interrupted current, the saliva obtained is usually more dilute
than that obtained on chorda stimulation. At times, however, the

, | LANGLEY: Journal of physiology, 1879, i, p. 96.

204 A. F. Carlson, F. R. Greer, and F. C. Becht.

sympathetic saliva obtained in this way is of the same concentration
or of a slightly greater concentration than the chorda saliva. Lang-
ley finds that the deficiency in solids is in the organic constituents.
We did not make determinations beyond the total solids. One typical
experiment is recorded in Table V, No. 1. In this case the gland
vein was not isolated, so that we had no means of knowing the
vascular condition in the gland on stimulation of the sympathetic.

Langley apparently assumes that this dilute sympathetic saliva
from the cat’s submaxillary gland is secreted under the same vascu-
lar conditions as is the concentrated sympathetic saliva in the dog.
It has been shown, however, that the cervical sympathetic in the
cat contains both vaso-dilator and vaso-constrictor fibres to the sub-
maxillary gland, and that on simultaneous stimulation of both with
the weak interrupted current, the dilators usually overpower the con-
strictors, so that we have an increased instead of a decreased flow
of blood through the gland. This fact naturally suggests that one
of the causes of the cat’s submaxillary sympathetic saliva being
usually so dilute is the absence of vaso-constriction and the
presence of vaso-dilation during the secretion. That the vascular
condition of the gland is not the sole cause of the sympathetic saliva
being dilute is shown by the fact that it is usually more dilute than
the chorda saliva, although the vaso-dilatation on sympathetic stim-
ulation rarely equals that on chorda stimulation.

When the submaxillary sympathetic saliva in the cat is secreted
under conditions of diminished blood supply, it becomes more concen-
trated in organic constituents than the chorda saliva. The blood
supply may be diminished by compression of the artery, or the
determinations may be made on the, individuals in which strong stim-
ulation of the sympathetic branch from the superior cervical ganglion
to the gland gives primary vaso-constriction, The data from two
experiments demonstrating this point are given in Table V, Nos. 2
and 3. Given the same vascular conditions in the submaxillary gland
during the secretion the cat’s sympathetic saliva exhibits the same
difference, in comparison with the chorda saliva, as is normally shown
by the sympathetic saliva in the dog. But because of the difference
in the submaxillary gland cells of the dog and the cat, the sympa-
thetic saliva of the latter will never reach thé absolute concentration
of that of the former.

1 CARLSON: This journal, 1907, xix, p. 408.

Blood Supply to the Submaxillary Gland. 205

SUMMARY.

1. The cervical sympathetic contains secretory fibres to the sub-
maxillary salivary gland.

2. Diminishing the oxygen supply to the active gland by occlusion
of the gland veins or compression of the gland arteries diminishes the
quantity of chorda saliva and increases its percentage of organic solids.
The organic constituents of this chorda salivain the dog may equal
or exceed that of the sympathetic saliva.

3. Atleast in some cases the chorda saliva secreted after a period
of diminished oxygen supply to the gland is richer in organic solids
than the normal chorda saliva.

4. The normal oxygen supply to the gland on chorda stimulation
must be reduced considerably before any marked change in the rate
of secretion and the character of the saliva appears, but, in general,
the greater the oxygen deficiency, the greater the decrease in rate
of secretion and the total quantity of saliva, and the greater the
increase in organic solids.

5. In the cat the sympathetic submaxillary saliva secreted under
diminished oxygen supply is more concentrated as regards the
organic constituents than normal sympathetic or chorda saliva.

6. There is probably no actual increase in the rate and quantity
of secretion of the organic salivary constituents in gland anemia.
This concentration is the result of diminution in the rate of secretion
and the quantity of water and salts. The processes of secretion of
water and inorganic salts are therefore more directly dependent on
free oxygen than is the secretion of the organic matter in the
saliva.

7. The differences between sympathetic and chorda submaxillary
saliva can be accounted for by the difference in the distribution of
the two sets of fibres in the gland and by the difference in the
oxygen supply to the gland on chorda and sympathetic stimulation.
Heidenhain’s theory of trophic secretory nerve fibres is therefore su-
perfluous, at least for the submaxillary gland of the cat and the
dog.

THE INNERVATION OF THE CEREBRAL VESSELS A$
INDICATED BY YHE- ACTION OF DRUGS

By CARL J. .WIGGERS.
[From the Physiological Laboratory of the University of Michigan.]}

CONTENTS.
Page
I: Inixoduction. Methodsjof Research <j. “2 2) 1.) o-) ds) (ep) to) oe
I, . The Action of Drugs’on‘the Cerebral Vessels). <525 0-50) ele) ee
Ti Adrenalin... ec. ue. iss ee, See te ceegnc vc et ane goin) a
2. Barium Chloride: ¢) <9. 9% te heehee isi se oem ome oe cones mee
3. ‘Chioroform .. 6 6s Ge ew be ee he) ee ay co oo
4. Digitalis... 2-6 6) Jape se es spe: erence eet sl ce Wen eben ent a
Sx Chloretone€ <9. ve PAR ome) i) ieutgien ihe sitet Wo) ito us eh cats fod a=
6. Apocodein .. . oa Sat eee)
(z) The Action tice Bievions Araitesen a Digitalis 27 be Gay Aes
7. The, Action of Drugs/after Apocodem = = 2... 25 =). 0+ eee
(a). Adrenalin: 6s. fs Ve oe) Cement) 3eu slag So) 4c) en tn eel
(6) Barium Chloride .¢6.5 5 co 2-2 = «= 5 + © 4) nn
(¢): Chloroform’ (2 fa Sa Soe) fe a Set te 21, 1. SI
(2) Digitalis. . . . 0: gt os) e, enea
(e) The General Relation of Ren re to potter Dies: «co: Je ene
III. Summary of Results .. . fo Si) is iw, S\0" 6 Lwin im © Win gel et» ioe enn
1, Tabulation of Drug Action SRR ee eee ce eS
2. General Conclusions. . . o aya

IV. The Physiological Significance of a eee a Deeks on iGeceas Veseie non oie,

I. INTRODUCTION.

Ros some years past there has existed among physiologists a
feeling of uncertainty in regard to active vaso-motor changes in
the cerebral vessels. This was created by the histological discovery
of fibres to these vessels! combined with an absolute lack of corrobo-
rative physiological evidence. For this reason Professor Lombard,

1 GULLAND: British medical journal, Sept. 17, 1897, p. 78. OBERSTEINER:
Arbeiten aus dem Institut fiir Anatomie und Physiologie des Centralnervensystem
an der Wiener Universitat, 1897, v, p- 215. HuBER: Journal of comparative neu-
rology, 1899, ix, p. I. HUNTER: Journal of physiology, 1902, xxvi, p. 465.

206

The Innervation of the Cerebral Vessels. 207

not willing to accept as final the negative results of active vaso-motor
changes, suggested, in 1903, that the cerebral circulation might
be studied by simpler and more direct physiological means than had
been employed. In 1905 I published a method of studying the cere-
bral circulation in a direct manner.! The chief
advantage claimed for the method was that it
allowed a study of the brain circulation uninflu-
enced by the general blood pressure, —a factor
which has given rise to no little misinterpretation
in the study of brain reactions. The method con-
sisted, in brief, in decapitating a dog, ligating
intracranially all communications with the Circle
of Willis, and then inserting a cannula into the
basilar artery. The isolated brain was thus alone
perfused at a constant pressure and temperature
with a, pulsating stream of Locke’s solution.
Thus prepared, the brain, still within the skull,
was suspended in a warm chamber above a drip-
registering apparatus, into which fluid dropped
from the most dependent portion of the prepa-

; : op - FIGURE].— Drip register-
ration. Thus the entire outflow with changes in ~ . ; Pest ee
2 ; ing apparatus. — Fluid
its rate was constantly recorded. Sincethat re- from preparation drips
port the apparatus, continuously recording the into funnel 4. With
outflow, has been much improved upon. It is Stopcock 4 closed, fluid

J : : rises in both limbs, car-
shown in Fig. rying up a bell float (C)

The curve written by this apparatus on an __ with vaselined steel rod.
evenly moving drum is a straight oblique line if “4nd 2’aretwo guides;
th tf . t t it b 4 the latter is of platinum.

e outflow be constant. It becomes curved aS ty. yapidity of rise is
soon as the outflow rate is changed. Fig. 2 recorded on the drum
shows this diagrammatically. This means of _ by a straw pointer (2).
showing slight and temporary changes in out- /,'S © overflow tube.

2 : A The apparatus dis-joints
flow immediately and accurately is not shared by forcleaning at Gand G’.
a large number of devices used to study the rate
of flow from veins. The crude and often fallacious method of count-
ing or recording the number of drops flowing from a vein cannot do
this, and is, moreover, not applicable to continuous outflow streams.
Comparing by direct measurement or graphic registration the number
of cubic centimetres flowing through an organ in equal time intervals,

1 For a more detailed description of the apparatus used, my former report may
be consulted (This journal, 1905, XIV, p. 452).

208 Carl F. Wiggers.

or determining the time required for definite quantities to flow through
an organ, are methods which give no indication of the nature of the
flow between measurements. In Fig. 2 let the vertical distances AB,
B'C, and C'D each represent an outflow of I0 c.c., and the equal
distances XA, AY, and YZ on the abscissz one-minute intervals.
Measurements of the distances AB, B’'C, and C’D at the end of each
minute will then show that Io c.c. have
passed through the organ in that time.
It gives no indication of what occurred |
during that minute. The outflow may
have been constant, as shown .by the
straight line BC; it may have been
slowed at first and accelerated later, as
shown by the curved line 1/7, or it may
x i, TE Sy have been accelerated at first to be
slowed later, as shown by the curved
line V. Jimportant changes are thus often lost tf a continuous outflow
record is not taken in perfusion experiments.

As constriction between the arterial and venous systems in an
organ ought certainly to diminish the outflow, it was thought that
such continuous records would give evidence of the slightest vascular
change in the organ. The results of some eighty perfusion experi-
ments on the brain and kidneys, however, have shown that changes
in the vessels may occur without causing changes in the outflow
record which differ appreciably from those frequently occurring in
normal outflow records. This occurs:

1. In cases in which the vaso-motor reaction is very slight or tem-
porary and the flow thrqugh the organ is large. The time is then too
short to cause an appreciable change in the outflow rate.

2. In cases in which opposite vaso-motor changes follow each other
very rapidly. In these cases the one reaction tends to neutralize the
other. I have obtained curves in which a dilation of a few seconds
duration, followed by a temporary constriction, showed no change in
outflow ; also those showing an increase, others a decrease. The pre-
dominant changes alone were recorded.

For these reasons changes in the rate of outflow cannot be utilized
in these cases, and a more delicate criterion must be selected. Such
a one is found in the characteristic pressure variations in the tubes
supplying the organ. The pressure for the perfusion fluid was created
by allowing the water, kept at a constant level by the apparatus pic-

The Innervation of the Cerebral Vessels.

209

tured in Fig. 3, to flow into an oxygen-containing bottle, thus displac-
ing the oxygen into the bottle containing the Locke’s solution, the

latter being driven into the cannula by the pressure
created. ‘ Between the oxygen bottle and cannula a
stopcock, regularly opened and closed by a cam and
motor, is interposed, and the oscillating lateral pres-
sure between the cannula and stopcock is recorded by
a mercury manometer.
drum resembles a blood-pressure curve with its car-
diac oscillations. In some experiments these oscilla-
tions were further magnified by an aluminum lever
actuated by the manometer float.

The changes which take place in this curve on
increasing the peripheral resistance are quite differ-
ent from those that we are familiar with in a blood-
pressure record, for the lowest or diastolic portion
of the oscillations feels the peripheral change much
more than the upper or systolic portion. This is
shown in Fig. 4, in which an increased peripheral re-
sistance was artificially produced by gradually tight-
ening a clamp on the outflow tube. The record
shows that, while the systolic portion of the oscilla-
tions always rises, the change may be so slight as to
be almost unmeasurable, for example, from A to B or
from C to D. The diastolic change, on the other hand,
is always evident. The reason that the systolic por-
tion of the oscillations should be so little affected
by slight peripheral changes is found in tHe arrange-
ment of the apparatus. The manometer as arranged
would only record the pressure created by the pressure-
regulating device when the outflow from the cannula
had entirely stopped. As a portion of this pressure
is converted into velocity pressure when fluid leaves
the system, it follows that the highest pressure re-
corded by this apparatus, when an organ is perfused,
depends, not only on the pressure created, but also on

The curve written on the CSS

FIGURE 3. L7es-
sure-regulating
apparatus. — A,
cylindrical lamp
chimney with cork
in bottom. Water
from tap flows in -
through tube 34,
and reaching level
C, flows off by
thistle tube D.

This insures a
constant pressure
level. Tube 2,

bent in U-form to
prevent entrance
of air, communi-
cates with oxygen
bottle. The whole
apparatus is raised
and lowered by
pulley arrange-
ment.l

the size of the outflow tube, as compared with the sizeof the tube

carrying water under pressure into the oxygen bottle.

The recorded

1 The principle of this apparatus was copied after a similar apparatus used by

Professor Cushny.

210 Carl F. Wiggers.

pressure, which is referred to as systolic in this paper, is altered only
by such marked changes in the vessels as cause a pronounced differ-
ence between the rate of outflow from the cannula and the rate of
inflow into the oxygen bottle. Slight constriction causes little rise,
because the slight change in outflow is evenly counteracted by an
opposite change in inflow, so that their relative proportion is not
disturbed.

The diastolic pressure, on the other hand, is being recorded when
the moving stopcock entirely closes off the manometer from the cen-
tral pressure. Whatever changes of pressure occur during this in-
terval can only be attributed to changes in the calibre of the perfused
vessels. As these drain the tubes distal to the stopcock, the extent
to which the pressure falls, as well as the rate of fall, depends entirely
on the degree of constriction or dilation present.

In this research both the changes in the rate of outflow and the
changes in the diastolic pressure were utilized to study vaso-motor
changes in the organs, and the one method served as a check on
the other.

II. THe ActTIon oF DRUGS ON THE CEREBRAL VESSELS.

The use of drugs constitutes at present a much employed and con-
venient means of studying the physiological reactions of blood vessels.
In this investigation their use was at first restricted to the determi-
nation of the existence of an active vaso-motor response in the cere-
bral vessels. Such a response was proved by the experiments reported
in my former paper! When, through the experiments of Lewan-
dowsky,? Langley,? Brodie and Dixon,‘ Elliott,® and Dale,® the idea
became prevalent that the presence of a vaso-motor innervation could
be proved by the use of certain drugs, further experiments were per-
formed to determine whether the reaction of the cerebral vessels to
these drugs necessitated the assumption that they possessed such an
innervation. To this end the actions of those drugs to which a neural
or nerve end action has been assigned were compared with others
supposed to affect the calibre of vessels by a direct muscular action.

1 WIGGERS: This journal, 1905, xiv, p. 452.

2 LEWANDOWSKY: Archiv fiir Anatomie und Physiologie, 1899, p. 360.
3 LANGLEY: Journal of physiology, 1go1, xxvi, p. 237.

* BRopDIE and Dixon: Journal of physiology, 1904, xxx, p. 476.

5 ELLIoTT: Journal of physiology, 1905, xxxii, p. 401.

® DALE: Journal of physiology, 1906, xxxiv, p. 163.

' The Innervation of the Cerebral Vessels. 211

The ability of one drug to modify the action of another was also re-
tested, and all the results were then compared with controls on the
renal vessels.

During the course of this investigation a number of points irrel-
evant to the conclusions sought were brought out, and these are
included for their general pharmacological interest.

I. The action of adrenalin Solution used. — In order to obtain a
solution of adrenalin of calculated strength, small quantities of the
crystalline product (P. D. & Co.) were weighed out intocapsules. Just
before use one of these
capsules was dissolved
by heating with a small
amount of Locke’s
solution, and then im-
mediately cooled by
the addition of cold

Locke’s solution to the
required dilution. This Ficure +.— Shows the effect of increasing the resistance
Ped a rela) Se Sune tom he eal a ae
tively little oxidized the systolic, and the descending limb of each wave be-
solution. comes more gradual as the resistance is increased.
Delicacy of response. —From twenty-five experiments on the cere-
bral vessels, controlled by a like number on the kidney, it was found
that the cerebral vessels constricted when doses of adrenalin ranging
from 3 c.c. of aI : 250 solution to 3 c.c. of a 1 : 20,000 solution were
introduced.2 Greater dilutions proved negative. In some cases, how-
ever, the vessels failed to react to such weak solutions. Thus, in
one experiment, 2 c.c. of a 1: 5000 solution failed to cause a re-
action. Two and one-half cubic centimetres caused a slight, and 3 c.c.
a marked, reaction (Table I). The kidney vessels were constricted
with I : 100,000 solutions, but not by weaker ones, as discernible by
this method. These results demonstrate that, although the kidney
vessels reacted to far weaker solutions than those of the brain, the
latter, far from having no reaction, reacted even beyond expectation
when the paucity of muscular elements in their wall is considered.

1 The literature on the action of adrenalin was reviewed in my former paper
(oc. cit.), and discussion of the relative value of these researches was. made there.

2 The drugs in all these experiments were introduced without changing the
pressure, temperature, or composition of the perfusion fluid, by the method de-
scribed in my former paper (oc. cét.).

“ut $0 = Y og T+ O'sT 9€ bb OT =
“wu 60 = ¥ og rE STZ 9 re OT 0
‘WU $0 = U 7 Or $ Ve Ob OT 0
‘uoljovar | ON ‘
“WU $0 = Y ee T+ 0°02 ce I'v Lv0 a
‘wu [ = y 0% bz O'br 12 sg S10 0
‘wu [ = ¥ 0% Te 00g Le rE om 0
‘mu $9 = YF OZ aj0;dur09 sie Sso[ Ot ae
e ae a mt 001 (Aa Lv i sol 29
S “wu T= 0% L's $2 bb L's OT 0
800 pw ot =a os Zs OU tb zs 990 e0U ‘sh
~~ (LONbOC tS oe oe oe oe oe oe 0
a Hae si we O01 KL €9 S10 fe
ne ixinis auou auou OLT roared O'0T 0% nie
“a euoU SUOM 00E SA OF Sal 18€
=~ "uu ¢ = Y SLs 97 O9L Ul Lt se 0
S outer 7p Sy auou auoU 0'ZS Tal a Canyg OT 0
“Wu € = 1 Ol £% O'8L (et o's Oo t8€
a - etoU 089 60 8% Sell 0
afta) a} “a
‘aSueyo oans | ‘saynurwu ur| ‘Aroa0o01 | 924°1°HIP eTeI3% tout “Uru UT ‘998 Ul
-said otoqseiqy | £19a000y : jo quaIxy akon ‘995 OF sod | ‘90s O¢ rad "191498109 potiad
posvjuac1ag| MONO MOWING, | [BWIXe, quae]

22

‘STUSSHA ITVUddNa) NO NITVNAYAGY HO NOILOVY

oly A Leh Val

00002: T “99 ¢€
000'S: T “99 ¢
000'S? T “99 GZ
OOO'S: T “99 Z
000%? T “99 ¢
00'S: T “99 T
000%: T “99 T
OOOT: T “99 ¢

OOS ha
OSL: T “99 ¢
OSL: T “9 ¢
OSL+T “20° ¢€
OOS aL” “e0' 1g
OOS: T “9 ¢
OOS: T “9 ¢
OOS ale eal. S
OOS {. “90 ¢
OOS Tt IRE
OSZ?1T “oD ¢€

*pasn uornyos
jo yysuUa.4S

*yUaUIT
-1adxa jo
JoquinN

ou

The Innervation of the Cerebral Vessels.

EL

(ASL ,, SUBALU UUUN]OO JSLT UT YY 19}}9] OY Lg

001: X =
Snap dojoq MOL + 9dUEIOPIP MOYING ‘uonszodoid oy) Suryeut Aq poureyqo sj}[Nser oy} syUdsoidas ooUdsIFfIp MOYINO asezUIIIOg 5
*‘poyjaut 10,09 Aq poure;qo yNSexY ¢ *poye[no[eo y[NSaY x
es ‘JUOWIOdXS SITY} UL UT]BUdIpPY 9LOJOq poqso} o}LAIU WINIpOS |
Pie! sa! ajojduioo eh Sele $02 3 ny OOOOT? T “99 ¢
uOoTJIVIA | ON Deu 00 ove 000‘00Z : T &9"9 ¢
uol}OveL | ON “19 Of oes (ALONOOs Ayr Make) Cs
“WU oS = Y S0 a9; dw09 OLT 09L 4°02 be coe OOO;OOT: T “99 ¢
“UU SQ = W 99:0 ayo;dwos oe oe oan .. oe 000‘08 : I arg ¢
So bee eh at OST 0:92 Wa O'sT a, 8 000';09 :T “99 ¢
Saece $0'T 0:02 SE Vel 0°02 y a O00'OF: T “OD €
pu SoS = ual O'ST 00 Sol OST cos oa OOO';OT * T “99D EF
“mu €¢ = a eT O91 0'9E XOT 99T a a OOO ZT tf “S9—E
‘WU 6'7Z = WT +T 7ST O'S+ O01 ost i oa 000'8: T “99 ¢
“WU OZ = Y +T O'6T 0 ¢b L01 O'6l a a 000'r: T “9 ¢ £s
‘uu 90 = a rll ss 98 cael ss $c'0 0 000%? T “9
a - t+ 208 ula +P a 0 O00? T “OD ¢ Ob
‘WSs = 7 ST OL OLT i) O'L OT - OOO;OT? T “99 ¢
‘wu T = U OT 08 +e (tS 18 OT (0) OOO'T: T “99 ¢€ Le
Sey ered 1 ayayduioo OE se 8s OT 0 Osh TE “SSE
pee =e 09 6S O'SE ob GL S20 0 OSE 1 “ae 1g
ww +1 = U OT OT S Saul O'eT £f0 0 OOS: T “99 ¢
wu 3 = a 0% OL Oe oes) OL sO 0 ‘Nora Verer 45 ce

‘STASSaA ‘IVNAY NO NITVNAYGY AO NOLLOY

214 Carl 7. Wiggers.

It is to be noticed that the degree of dilution to which the vessels
reacted was never found to approximate the figure given by Meyer.
He showed that a strip of artery, if immersed in adrenalin solution,
would react to dilutions as great as I : 1,000,000. Attention should
be directed to the fact that a greater quantity of adrenalin was avail-
able for his immersed strip than would be available for the vessels of
an organ perfused with the same quantity of solution; for in the lat-
ter case it must be distributed to a much larger surface of vessel
wall, and remains in contact but temporarily. Moreover, the magni-
fication of the contraction used by Meyer undoubtedly enabled the
recognition of slighter reactions than could be hoped for by ordinary
methods.

The latent period. — Oliver and Schafer,” in their study of supra-
renal extract, reported that the rise of pressure always occurred
after an interval of latency. Regarding it, they say, ‘‘ No doubt this
interval is mainly occupied by the passage of the extract injected
from the peripheral vein toward the heart, and thence to the arterial
system.” The “latent period” results of most investigators do not
deduct the time required for the injected drug to pass through the
animal or apparatus to its point of intended action. Such results are
variable and misleading in comparisons. I fell into this error in my
former paper when I showed that my latent periods with adrenalin
were in harmony with those of S. J. and C. Meltzer.2 These authors
had observed that the cutting of nerves to the ear vessels appreciably
lengthened the latent period of adrenalin, — an observation from which
it must be inferred, either that the cutting of nerves affected the
ability of vessels to react as readily, or that the first constriction
resulting is of central origin. My substantiation of their long Jatent-
period results must be retracted, for the fact is that the latent-period
results of one worker cannot be compared with those of another if
the latent period of the animal or apparatus be included, or if an
average latent period for the apparatus be deducted. The reason is
that, even were the quantity of fluid ahead of the drug always con-
stant, the time taken to displace it would depend on the rate of flow
through the organ, and must therefore vary in and during each ex-
periment. For this reason the term “latent period ” is employed in
this paper to denote the time elapsing between the entrance of the

1 Meyer: Zeitschrift fiir Biologie, 1906, xlviii, p. 3.
? OLIVER and SCHAFER: Journal of physiology, 1896, xviii, p. 230.
8 MELTZER, S. J. and C.: This journal, 1903, ix, p. 147.

The Innervation of the Cerebral Vessels. 215

drug into the blood vessels and the time a reaction occurs. Knowing
the total latent period, the total amount of fluid passing through the
organ during this time, and the amount that must be displaced in the
apparatus before fluid reached the vessels, a little mathematical cal-
culation can approximately determine this actual latent period. The
results obtained by this calculation are not infrequently slightly neg-
ative, due, probably, to unequal onward diffusion of the drug in the
different experiments. For this reason this method is less exact than
another procedure employed, namely, to add to the injected drug, if
this was not itself colored, some inert color, and mark with a time
signal the time of its appearance in the end of the cannula.

Results obtained from such procedures can lay claim to no minute
_ degree of accuracy, but are sufficiently correct to warrant the con-
clusion that adrenalin produces its reaction very shortly after the
drug is in contact with the vessels, and that separation from the cen-
tral nervous system does not actually lengthen the latent period.!
These results are confirmed by those reported by Meyer. He showed
that a strip of artery removed from the body reacted a few seconds
after it was in contact with the drug.

Course of the reaction. — The results recorded in Table I show that
the maximal degree of constriction, measured from the time a reac-
tion began, was somewhat more rapidly reached in the kidney than
in the brain vessels. In neither organ could any relation between
the size of the dose and the rapidity of onset be determined. As re-
gards the degree of reaction, the results show that a great variation
exists in differences in outflow before and after the injection of the
drug. To compare degree of outflow in, the various experiments,
however, such actual differences in the outflow, measured for thirty-
second intervals before and after the use of the drug, must first be
converted into percentage figures of the total flow through the organ
at that time. These figures may conveniently be termed the fercezt-
age outflow differences. A glance at this column of figures shows that
the degree to which the vessels reacted was but little dependent on
the strength of the adrenalin solution used. In the cerebral vessels,
for example, a I: 5000 solution could cause either a greater or smaller
constriction than a 1: 500 solution.2, The same was true in the ves-

1 In the two experiments shown in Table I in which longer latent periods were
obtained, sodium nitrite had previously been used. For this reason they cannot
be looked on as exceptions to this statement.

2 Cf. Experiments 2, 36, and 46.

216 Carl F. Weggers.

sels of the kidney. We may conclude, then, that a weak dose of
adrenalin, the actual size of which must vary with the number and
size of the vessels in the organ as well as their state of irritability at
the time of application, supplies a stimulus to the vessel walls suffi-
cient to cause a maximal response. The 1 : 40,000 solution in Exper-
iment 53 may be looked on as such a threshold dose, just capable of
producing a maximal constriction. Any smaller dose (1 : 60,000)
caused a slighter degree of constriction.. Any larger one, on the other
hand, had no appreciably greater effect.

It cannot be said, however, that strong doses of adrenalin react in
every respect like such a threshold dose, for the table shows that,
though the degree may not have been affected, the duration of the
reaction, as well as the completeness of recovery, vary directly with
the strength of the dose. This is shown not only by successive appli-
cations of weaker solutions to the same preparation, in which the
duration became less and less, as in Experiment 53; but also by ap-
plications of varying strengths to different preparations.

As regards the completeness and rapidity of recovery the cerebral
vessels differed from those of the kidney, in that strong solutions
caused a much more prolonged reaction, the recovery was slower and
more imperfect, or at times failed altogether. This, however, cannot
be regarded as a property peculiar to the cerebral vessels. The ex-
planation is a mechanical one. The less spacious vascular channels
of the brain allow fluid to pass through them less rapidly than do
those of the kidney, so that the brain vessels, arteriole for arteriole,
are exposed more thoroughly to the adrenalin than those of the kid-
ney, which allow a greater portion to pass through unused. This
explanation is based on the easily demonstrated principle that, if the
time that the vessels of the perfused organ are exposed to a drug is
altered, the duration of the reaction is changed. This was experi-
mentally shown in several ways:

1. By augmenting the flow through an organ by raising the per-
fusion pressure. This was done in Experiment 35 on the kidney.
Two successive doses of a 1: 500 solution of adrenalin were applied.
After the first dose the reaction passed off in two minutes. Before
the second dose was applied the pressure was raised so as to almost
double the flow through the organ, and the action of the second
passed off in one minute.

2. By perfusing a part of an organ. If only one half of the kid-
ney was perfused by ligating one of the two main branches of the

The Innervation of the Cerebral Vessels. 217

renal artery, the kidney vessels recovered as slowly and imperfectly
as those of the brain.!

3. By utilizing the development of oedema which probably dimin-
ished the calibre of the vessels. It was found that when such cedema
occurred, the adrenalin caused a much more prolonged reaction than
when it was not present. In Experiment 53, for example, a I: 10,000
solution at the end of the experiment, when cedema was present,
caused a more prolonged reaction than a 1: 4000 solution at the
beginning.

2. The action of barium chloride. JLzterature. There seems to
be but little literature on the peripheral action of barium chloride,
and, as far as the cerebral circulation is concerned, no reference has
come to my notice. In 1875 Boehm and Mickwitz? observed that
an intravenous injection of barium chloride caused a constriction
of the blood vessels in many parts of the body. Ringer and Sains-
bury,’ in 1883, came to the conclusion that the cause of the constric-
tion was due to a direct muscular action. Ringer,* in 1886, showed
that a 2 per cent solution caused a contraction of perfused vessels.
Brodie and Dixon,® in 1904, described the course of action of barium
chloride. They found that after a long latent period a constriction
of slow onset but of persistent character was obtained with moderate
doses. Larger doses caused a more rapid constriction, and this was
preceded by a preliminary dilation.

Solution used. — Barium chloride was dissolved in Locke’s solution
to the per cent desired, and the slight precipitate forming allowed to
settle. Only the supernatant fluid, which still contained barium as
tested, though in somewhat smaller proportion than calculated, was
used. The solutions used varied in dilution from 1:1000 to 1:10.

Course of reaction. — Barium chloride was applied to the normal
vessels of the brain in only three experiments, and two controls on
the vessels of the normal kidney were made; but it was applied
in a dozen other experiments after other drugs had acted, which did
not alter the general character of the reaction. In stronger concen-
trations barium chloride caused, sometimes after a slight prelim-
inary dilation, a marked and rapid constriction of the cerebral vessels,

1 Cf. Experiment 5 with Experiment 31, Table I.

? BorHM and Micxkwitz: Archiv fiir experimentelle Pathologie und Pharma-
kologie, 1875, iii, p. 216.

8 RINGER and SAINSBURY : British medical journal, 1883, ii, p. 265.

* RINGER: Journal of physiology, 1886, vii, p. 306.

5 Brop1E and Dixon: Journal of physiology, 1904, xxx, p. 486.

Carl F.

218

Wiggers.

as shown by a rapid decrease in outflow and a diminution in size of

the pressure curve oscillations.
reaction occurred was zero.

ie)

Oo

O—NWOKOIDNOSO

FIGURE 5.— Outflow from the cerebral ves-
sels per thirty seconds after application of
barium chloride. (4) The effect of strong
solutions BaClg, 1-10, (4) effect of weaker
solutions BaCly, 1-50. No recovery in
two hours.

The actual latent period before the
No sign of recovery was present, al-

though one experiment was con-
tinued for two hours. The vessels
no longer reacted either to heat
or cold, either to the nitrites or
an increased perfusion pressure
(Figs, )s

In weaker applications the drug
caused, without any actual latent
period or preliminary dilation, a
smaller but easily recognizable con-
striction, which was followed by
recovery in three minutes, as
shown by changes in oscillations
and outflow (Fig. 5,4). A 1:500
solution seemed to be the weak-
est to cause any distinct reaction.
After recovery from such dilutions

other drugs were still able to act. The following synopsis of a typical
experiment shows the effect of weak solutions more in detail:

Experiment 65. March 12, 1907.— Dog decapitated at 2 P. M.

fused at 3 P.M.

Brain per-

3-15. Barium chloride 1: 1000, 3 c.c. No result.

3.19. Barium chloride 1: 1000, 3 c.c. No result.

3-26. Barium chloride 1: 500, 3 c.c. Very slight, slow constriction.
3-30. Barium chloride 1: 200, 3.c.c. Constriction. Latent period, o ;

maximal constriction in one hundred seconds ; change in outflow per thirty
seconds from 1.8 to 1.3 c.c. ; recovery in six minutes to 1.8 c.c. Diastolic

pressure rose.
3-38. Barium chloride 1:

ZOO, 3 /C.€s

Constriction. Latent period, 0;

maximal constriction in one hundred and seventy seconds; change in
outflow per thirty seconds from 1.8 to 0.7 c.c.
3-52. Chloretone 1:125, 3 c.c. Dilation. Change in outflow per

thirty seconds from 0.7 to 1 c.c.

The kidney vessels showed essentially the same reactions.
The reactions obtained in these experiments differ from those
obtained by Brodie and Dixon in the absence of any latent period

The Innervation of the Cerebral Vessels. 219

and in the rapid course of onset. Their latent period results can
probably be accounted for by inclusion of the period of the apparatus.

3. The action of chloroform. JLzterature.—In the Transactions
of the Royal Society of Edinburgh! Schafer and Scharlieb review the
literature on the peripheral action of chloroform. They themselves
studied the action of chloroform on the coronary, limb, and renal ves-
sels by the perfusion method, and found that a constriction occurred
with dilutions as great as 1: 20,000, in the coronary and limb vessels,
while the kidney vessels were dilated by those beyond 1:1000. The
actual number of cubic centimetres of the chloroform solution allowed
to go through the organ was not stated, but one infers that it was
quite a large quantity.

The action of chloroform on the cerebral vessels has never been
made the subject of any very accurate work. Schiller? thought he
could observe a progressive constriction of the pial vessels with each
inhalation of chloroform. Gaertner and Wagner,? by measuring the
rapidity of flow from the jugular vein, and Hiirthle,t by measuring
the pressure in the peripheral end of the internal carotid, came to the
conclusion that an active dilation occurred. Bayliss and Hill ® showed
that these results could be as well explained by an increased general
venous pressure, and assumed that no active process in the brain
vessels occurred. Roy and Sherrington® had previously reported a
diminution in the size of the brain during narcosis, and therefore
assigned a constricting power to chloroform. Pick’ found an in-
creased outflow from the external jugular vein at the same time that
the flow from the femoral vein was slowed, —a combination of events
which he deemed sufficient proof of an active cerebral dilation.

Solution used. —In all experiments a saturated chloroform solu-
tion (1: 200) was used for a stock supply from which dilutions were
made when necessary. Inevery case 3 c.c. of the solution to be tested
was allowed to flow through the organ.

Course of reaction. — In studying reactions of the vessels to chloro-
form it was soon discovered that even a continuous outflow record did

1 SCHAFER and SCHARLIEB: 1904, xli, p. 3II.

2 SCHULLER: Berliner klinischer Wochenschrift, 1874, p. 295.

8 GAERTNER and WAGNER: Wiener medicinische Wochenschrift, 1887, p. 602,

4 HURTHLE: Archiv fir die gesammte Physiologie, 1889, xliv, p. 561.

5 Bay.iss and HILv: Journal of physiology, 1895, xviii, p. 334.

6 Roy and SHERRINGTON: Journal of physiology, 1890, xi, p. 85.

7 Pick: Archiv fiir experimentelle Pathologie und Pharmakologie, 1899, xlii,
p. 412.

220 Carl F Wiggers.

not present as trustworthy a picture of the behavior of the vessels as
a record of side pressure oscillations, and these were largely made
use of in interpreting results. The kidneys of dogs, cats, and rab-
bits were used, while the study of the brain reactions was of ne-
cessity limited to dogs. The results of twenty applications show
that in both brain and kidneys a 1: 200 solution of chloroform
constricted the vessels. The differences between the reactions in
these organs were similar to those found with adrenalin, and were
probably attributable to similar causes. The constriction occurred
without an appreciable latent period, and reached its maximum rap-
idly in both brain and kidney. Recovery was slower in the former.
With progressive dilution of a 1: 200 solution to a 1: 2000 solution,
the vessels not only reacted less and less by constriction, but this
diminished constriction was followed by a dilation more and more
marked. Thus, in Experiment 59, on acat’s kidney, 3 c.c. of 1: 200
solution caused constriction from which the vessels recovered in sixty
seconds. A 1: 400 solution caused a similar constriction of shorter
duration. A 1: 600 solution caused a slight constriction, as shown by
the oscillation curve, but not in the outflow record. This was fol-
lowed by a dilation distinctly evident in both curves. A I: 1000
solution showéd no evidence of constriction, but only a dilation.

The same type of reaction was true of the cerebral vessels. In
Experiment 64 a 1: 4000 chloroform solution caused a dilation only,
with recovery in eighty seconds. A 1: 2000 solution as well as a
I: 1000 solution caused this dilation only after a preliminary constric-
tion. A 1: 500 solution caused a constriction only. Similar results
were obtained in other experiments. These are again referred
to.

These results warrant the following deductions concerning chlo-
roform :

1. The kidney vessels are not the only vessels, as Schafer and
Scharlieb claim, on which the weaker solutions of chloroform exert a
dilating influence. The cerebral vessels at least react like those of:
the kidney. This makes it questionable whether such a difference
between various vessels of the body, as Schafer and Scharlieb
describe, exists.

2. No definite concentration of the chloroform in the perfusion
fluid can be placed as the limit which determines the reaction as
being one of constriction or dilation. The strength which in 3 c.c.
doses produces dilation first may be 1: 4000, as in one brain ex-

*

The Innervation of the Cerebral Vessels. 221

periment: it may be 1: 600, as in one kidney experiment. In fact,
the amount injected and the rapidity with which this flows through
the organ largely determines the result.

3. Chloroform in solution has a twofold action on blood vessels,
constricting andrelaxing. In many cases where a dose was adjusted
to the outflow so as to fall between the doses producing only con-
striction and those producing only dilation, it could be determined
that a dilation followed the constriction. Any stronger dose dimin-
ishes the constriction, while any weaker dose increases the following
dilation. The predominant changes regulate the amount of fluid
passing through the organ in definite time intervals.

4. The action of digitalis. Previous work.—It has been shown
repeatedly by perfusion experiments that digitalis in different prepara-
tions has a peripheral action on blood vessels. Gottlieb and Magnus!
have given a complete review of the work done. These authors
studied its action on the cerebral vessels by the use of three methods,
namely, (1) by inspecting the pial vessels, (2) by recording changes
in the volume of the brain, and (3) by noting changes in outflow from
a cerebral vein consequent to the injection of the drug. Asa result,
they report that strophanthin as well as digitoxin constricted the
cerebral vessels. As these results were obtained, however, by
methods which introduced the variable general blood pressure as a
factor, they should be accepted with reserve, and should be subjected
to reinvestigation.

Preparations used. — Two preparations of digitalis were chosen for
this work, the pure principle digitalein (Merck), and an infusion pre-
pared to U. S. P. strength by steeping 15 parts of bruised digitalis in
1000 parts of Locke’s solution. These preparations were found most
suitable, as they were easily dissolved in Locke’s solution and con-
tained no substances which modified the pure digitalis action. The
latter objection was found against such preparations as the tincture,
fluid extract and digitalone, which were also tested.

Nature of the reaction. — The active principles of digitalis, as shown
by the use of digitalein and the infusion, affect the peripheral blood
vessels of the brain as well as the kidney by inducing a constriction.
In no instance was a dilation produced. Only small doses of digi-
talein were required to cause a reaction (0.01 mg. in the kidney and
0.05 in the cerebral vessels). The constriction resulting from such

1 GoTTLiesB and MaAGnus: Archiv fiir experimentelle Pathologie und Pharma-
kologie, 1901, xlvii, p. 135; also 1902, xlviii, p. 262.

222 Carl ¥. Wiggers.

doses reached its maximum rapidly (7-11 seconds), and complete recov-
ery followed (Curve 1, Fig. 6). Doses larger than 0.1 mg. caused
a more marked constriction, from which the vessels either failed to re-
cover at all or did so only slightly. The induction of a tonic state of
contraction seemed to be the most characteristic action of the drug
(Curve 2, Fig. 6). Each succeeding dose of digitalein caused more
and more of a permanent constriction. After the vessels were thus
markedly contracted by
TH these doses, they still re-
Pc a AK wane tained the ae of re-
T acting to such drugs as
| : adrenalin, chloroform, and
KARA barium chtorsde, and re
sponded to an increase in
i the perfusion pressure with
wansan yl LT i RA a greater outflow. It may
Hil i iil be added that the reaction
to the infusion of digitalis
was of the same nature.
5. The action of chlor-

ik

FicurRE 6. — Effect of digitalein on blood vessels as
determined by changes in the diastolic pressure. ;
(1) 0.05 mg. on the cerebral vessels. (2)2mg. etone. Literature. — As

digitalein on the renal vessels. (3) 3 mg. digitae chloretone represents a

lein on renal vessels. Time in seconds. product closely related to
chloroform, and as no exact investigations of its peripheral action have
come to my notice, a few observations in regard to this drug were
made on the brain and kidney vessels with the hope that it might
prove an important drug for us.

The only reference to its peripheral action that could be found was
that of Impens,! who observed a dilation of the blood vessels of a
frog’s foot after its injection. This, however, he attributed to a
paralysis of the vaso-constrictor centre, inasmuch as a saturated chlo-
retone solution perfused through a decapitated frog did not increase
the outflow, but decreased it. Thus he supposed its purely peripheral
action to be one of constriction.

Solution used.— One part of chloretone is soluble in 125 parts of
Locke’s solution. In this strength it was perfused through the brain
and kidney vessels.

Results. —Chloretone caused only a dilation of the cerebral and

* IMPENS: Archives internationales de pharmacodynamie et therapie, viii,
Pp: 77-

The Innervation of the Cerebral Vessels. 223

renal vessels. The dilation of the brain vessels lasted from one to
two minutes, while in the kidney vessels their action was, for mechani-
cal reasons, about half as long. The synopsés of two typical
experiments show the general character of the reactions.

Experiment 52. December 16, 1906. — Dog decapitated at 2 p.m. Perfusion
of kidney at 10.25 P. M.

10.59. Chloretone 1: 125, 3 c.c. Maximal dilation in twenty-five sec-
onds ; change in outflow per thirty seconds from ro to 11.3 ¢.c. ; recovery
not waited for.

Experiment 86. March 21, 1906.— Dog decapitated at 5 p.m. Perfusion of
brain at 7 P. M.

7-08. Chloretone 1/125, 3c.c. Maximal dilation in twenty-three
seconds ; change in outflow per thirty seconds from 10.5 to 12.2 c.c.;
recovery in fifty seconds.

Some estimate of the power of chloretone to dilate vessels may be
gained by studying its neutralizing action of some known constricting
drug. Digitalis supplies a convenient drug for this purpose, and the
following experiment illustrates the neutralizing effect. Three cubic
centimetres of the infusion of digitalis caused a constriction reaching
its maximum in twenty seconds, with partial recovery in fifty seconds.
Then 3 c.c. of a mixture of equal parts of the above infusion and chlore-
tone (1:125 solution) were perfused. Theconstriction occurring after
digitalis alone was now replaced by a dilation lasting for about two
minutes. In one case the dilation was preceded by a temporary
constriction lasting five seconds. As small a quantity as three drops
of a 1: 125 solution of chloretone added to 3 c.c. of digitalis infusion
was sufficient to diminish the constriction and the permanency of its
duration. We may conclude, then, that chloretone has a marked,
direct dilating effect on blood vessels, and, added to digitalis prepa-
rations, abolishes their action. It does not, however, in any way
modify the subsequent action of digitalis; for if a dose of chlore-
tone was injected and allowed to cause a dilation, and then this was
followed, before recovery, by the injection of a dose of digitalis in-
fusion, the ordinary constriction course is run by the latter. As far
as other experiments were conducted, they failed to show that chlore-
tone affected, to any extent, the subsequent action of other drugs, as
chloroform, barium chloride, or adrenalin.

6. The action of apocodein. Previous work. — Brodie and Dixon,!
in their study of peripherally acting drugs, demonstrated some valu-

? BRopIE and Dixon: Journal of physiology, 1904, pp. 97 and 476.

224 Carl F Wiggers.

able physiological reactions by means of apocodein. Injected into
an animal in moderate doses, it caused a permanent fall of blood
pressure, sometimes preceded by a slight preliminary rise. This fall
was uninfluenced by section of the vagi, but was abolished by
nicotine and diminished by section of the splanchnics. They assumed
the action to be primarily on the ganglion cells, and the nature of the
reaction a paralysis. Ifadditional larger doses of apocodein were
then injected, they caused no further fall, but a rise in blood pressure,
which was attributed to a peripheral-action of the drug. The periph-
eral action was more clearly demonstrated in perfusion experiments.
For the characteristics of this peripheral action the above-mentioned
authors may be quoted (page 497): “‘ The first effect of an injection
of apocodein during the course of a perfusion through the limbs, kid-
neys, or intestines, is to produce a constriction. This quickly
passes off, and subsequent injections produce less and less effect, until
after the injection of a few cubic centimetres of a I per cent solution
no constriction is produced, and in most cases a dilation takes place.
It is now found that the nerve endings are paralyzed. Stimulation of
the vaso-constrictor fibres is entirely without effect.” Adrenalin was
also found to be without effect.

Solution used.— A 3 to I per cent solution of apocodein was
used. In every case the solutions were prepared just before their
use.

Results. — The results here reported were gathered from fifteen
applications of the drug to the brain vessels, and a like number of
controls on the kidney. All experiments were made on dogs. Ap-
plied to the cerebral and renal vessels in doses ranging from 0.5 to
20 mg., the drug invariably induced a constriction, attaining its
maximum rapidly. This constriction never passed off to any extent,
whether caused by weak or strong solutions, there being only a differ-
ence in the degree of reaction. In the kidney vessels the constric-
tion was sometimes preceded by a slight temporary dilation, as shown
by the diastolic pressure change.

The reaction of the cerebral vessels to successive injections varied
in different experiments. The experimental data collected show that
after 2 c.c. of a I per cent solution, further injection was without
effect; while in other cases this dose was repeatedly introduced into
the cerebral vessels, each time causing a constriction superimposed
on that already existing. This reaction was well shown by the pres-
sure oscillations as shown in Fig. 7, which represents segments of

The Innervation of the Cerebral Vessels. 225

the oscillations after the injections of this drug. After several appli-
cations of such doses, the reaction of the vessels was only temporary,
with a prompt return to the tonic state of constriction established by
the previous doses. Such reactions were usually slight constrictions
(Fig. 7), but in two cases slight dilations occurred.

The kidney vessels, like those of the brain, reacted to each succes-
sive dose by an additional tonic constriction. In the experiments in
which the vessels reacted to the first dose by a temporary dilation

| | | \
ai tN RARILANCUN NL LMRAANN ayaa A

lec.1% Sol. 4.18 P.M. 2cc. 4.25PM. 2cc. 4.35 P.M, 2cc. 4.43 P.M.

FiIGuRE 7.— Effect of successive injections of apocodein on the cerebral vessels as shown
by sections of pressure oscillations.

followed by constriction, each succeeding dose caused less constric-
tion, but the temporary dilation increased in degree and duration.
When this dilation was maximal, the calibre of the vessels never ap-
proximated the normal, as shown by a diminished outflow rate, and
the height of the diastolic pressure. Even after large doses had been
introduced, the vessels still reacted to an increase in the perfusion
pressure by a greater outflow. This was tested in two experiments
on the cerebral vessels and in three on those of the kidney.

The action of apocodein after digitalein. —The great similarity in the
tonic constriction produced by these two drugs suggested the question,
To what extent does the previous action of digitalein modify the
action of apocodein ? Only two experiments were made on the cere-
bral, and six on the renal vessels. The kidney was better suited for
this investigation than the brain, because of its greater vascular supply
and because a control could be made on the opposite organ. The
results in each case show, not only that digitalein possesses the power
of reducing the action of apocodein, but also that large doses could
convert its action into a temporary constriction, followed even by a
slight dilation. In Fig. 8 curve I was made as a control on the right
kidney of a dog. Curves 2, 3, and 4 were taken on the left kidney
of the same dog after previous application of digitalein had been
made, under as similar conditions as possible. These latter curves
show that the degree of constriction, as well as its duration, was
markedly diminished. This could not be assigned entirely to the fact
that the vessels were constricted to their greatest degree, for in the

226 Carl F Wiggers.

same experiment adrenalin and chloroform still caused their typical
reactions.

7. The action of drugs after apocodein. The action of adrenalin. —
Experiments performed on both the brain and kidney vessels showed
that the action of adrenalin could be entirely abolished by previous

application of apocodein. The dose of
ca r wn boy " ww apocodein necessary to do this was

iN MN nm found to vary. In some experiments 10
mg. were sufficient while in others
three times this quantity did not entirely

ga Minin abolish its action. The dose required
was smaller if injected in a few larger
doses than if introduced in numerous
small ones. The time interval between

the injection of the drugs, apocodein
and adrenalin, varied from one hour to
one minute, and seemed to be without
influence on the effect.

If the dose of apocodein was not suf-
ficient to cause complete abolition of the
adrenalin action, it reduced it. A com-
parison of the pressure curves resulting
from the introduction of adrenalin after
previous action of various-sized doses of
the apocodein showed that the latter
shortened both the degree and duration
of the reaction of the former (Fig. 9).

The action of barium chloride. — After

I (i ci iN

1 =
vk wis AAA

FIGURE §.— Effect of apocodein
after digitalein: (1) One-half c.c
apocodein (1 per cent) on nor-
mal right kidney. (2) One-half

c.c. apocodein (1 per cent) on
left kidney after 3 mg. digitalein.
(3) Same after 6 mg. digitalein.

the administration of apocodein in such
dose that adrenalin failed to act, barium
chloride, applied from seven to twenty-

(4) Same after 12 mg. digitalein. . ;. :
Pye eight minutes after, did so. The reac-

tions did not differ in their general character from those obtained on
the normal vessels of the brain and kidney, although there was no
longer so marked a tendency for the reactions from weaker solutions
to pass off. This was probably partly due to some mechanical cause,
as discussed under adrenalin. In two experiments the reactions to
strong solutions (2 to 5 per cent) differed from those normally ob-
tained by an augmentation of the dilation preceding the constriction

(Exp.°33.and35).

The Innervation of the Cerebral Vessels. 227

The action of chloroform.— As far as the reaction to chloroform
was concerned, the vessels were not altered by the previous introduc-
tion of apocodein. This may be said both of strong solutions, which
still constricted, and weak ones, which still dilated after the vessels
failed to react to adrenalin. Moderate doses still caused both dila-
tion and constriction, and, as far as could be determined by the limited
number of experiments, the
relative degree and duration eR
of each was not affected. Iara “cK

The action of digitalis. —

Z
It will be recalled that digi- Mii "HAA

talis, by inducing a tonic con-

. . , . BI
striction, had the power in veneer"
suitable doses of preventing
the subsequent action of apo- ri
codein. The results cannot so ™tintinntaaniannninintay

readily be reversed, as the ex-
periments, with one exception,
show, for digitalein still acted
after large doses of apocodein TURF 2 Met of ateatin er neon.
had been administered from on normal vessels of kidney. (2) Same five
two and one half to sixteen minutes after 1 c.c. apocodein (1 per cent).
minutes before. In each case (3) Same after 2 c.c. apocodein (1 per cent).
Bt fs ; (4) Same after 3 c.c. apocodein (1 per cent).
the constriction was superim-
posed on the one previously induced by apocodein, and for this reason
was necessarily less marked than on the normal vessels.

The general relation of apocodein to other drugs. — Two experiments
—one on the brain, the other on the kidney—could be regarded as
model ones, for they permitted, before cedema set in, the introduction
of a number of drugs, and thus condensed into one the results of other
experiments showing the relation of apocodein to other drugs. A
brief account of the experiment on the cerebral vessels will be given.
That on the kidney was practically identical. Seven milligrams of
digitalein were injected in three doses, each causing an additional con-
striction. After this neither a 1 nor a 2 c.c. dose cf apocodein (1 per
cent) caused any reaction. Three cubic centimetres of adrenalin
(1: 1000) caused a slight reaction, lasting for one hundred and thirty
seconds. Again 2 c.c. of apocodein (1 per cent) were injected, produc-
ing a very slight dilation. Subsequent to this, adrenalin failed to act,
2 mg. of digitalein still caused an additional reaction ; 3 c.c. of a1: 200

TU

228 Carl F. Wiggers.

solution of chloroform caused a slight reaction, and in spite of the
marked cedema present at this time, 2 c.c. of a 2 per cent barium
chloride solution caused a further permanent constriction.

III. SuMMARY OF RESULTS.

Adrenalin. — Weak solutions, 1: 100,000 to 1: 5000.
Constrict vessels of brain ; recovery complete in 2 to 3 minutes.
Constrict vessels of kidney ; recovery complete in 0.5 to 1.5 minutes,
Moderate strength solutions, 1: 4000 to 1: 1000.
Constrict vessels of brain; recovery incomplete in 2 to 5 minutes.
Constrict vessels of kidney ; recovery complete in 1 to 1.5 minutes.
Strong solutions, 1 : 750 to 1: 250.
Constrict vessels of brain ; no recovery.
Constrict vessels of kidney ; recovery incomplete in 2 to 6 minutes,
Barium chloride. — Weak solutions, I: 500 to rT: 5o.
Constrict cerebral vessels ; recovery complete in 3 to 6 minutes.
Constrict renal vessels ; recovery complete in 1.5 to 6 minutes.
Strong solutions, 1: 40 to 1: fo.
Constrict cerebral and renal vessels, preceded sometimes by a slight
temporary dilation ; no recovery.
Chloroform. — Weak solutions, 1 : 4000 to I : 1000.
Dilate the cerebral vessels ; recovery complete in 1 to 1.3 minutes.
Dilate the renal vessels ; recovery complete in 0.5 to 1.5 minutes.
Moderate strength solutions, 1: 1000 to 1: 400.

Constrict the cerebral vessels for about 0.5 minute, then dilate ; re-
covery in r to 1.5 minutes.

Constrict the renal vessels for about 0.6 minute, then dilate; re-
covery in 1.5 to 2 minutes.

Strong solutions, 1: 300 to 1: 200.
Constrict cerebral vessels ; recovery in 2 to 4 minutes.
Constrict renal vessels ; recovery in 1 to 1.5 minutes.
Digitalis. — Weak doses, 0.01 to o.1 mg. digitalein.

Constrict cerebral and renal vessels; recovery AGE complete in
0.5 to t minute.

Moderate doses, 0.1 to 1 mg. digitalein.

Constrict cerebral and renal vessels tonically ; no recovery ; subse-
quent action of apocodein diminished; subsequent action of
adrenalin and chloroform not affected.

Large doses, 1 to to mg. digitalein.

Constrict cerebral and renal vessels ; no recovery ; subsequent action
of apocodein entirely prevented or changed into a slight dilation ;
constricting power of adrenalin only slightly diminished.

The Innervation of the Cerebral Vessels.

to
i)
No)

Chloretone. — Saturated solutions, 1: 125.
Dilate cerebral vessels ; recovery in 1 to 2 minutes.
Dilate renal vessels; recovery in 0.5 to 1 minute.
Do not prevent subsequent action of any drug.
Apocodein. — Weak solutions, 0.1 to 0.4 c.c. of 0.5 per cent solution.
Constrict cerebral vessels ; no recovery.
Sometimes dilate temporarily, then constrict renal vessels; no
recovery.
Strong solutions, 1 to 2 cc., 1 per cent solution.
Constrict cerebral vessels ; no recovery.
Dilate renal vessels temporarily, then constrict ; no recovery.
Diminish or abolish subsequent action of adrenalin; somewhat re-
duce the action of digitalein; but do not affect the subsequent
power of chloroform, chloretone, or barium chloride.

From the above summary the following general conclusions may be
drawn:

I. It is evident that the cerebral as well as the renal vessels react
to the various drugs, and sometimes to different dilutions of the same
drug in very characteristic ways. Thus it is found that chloretone is
capable only of dilating vessels separated from the central nervous
system, while adrenalin and digitalein only constrict them. With
another group of drugs dilation or constriction, or one following the
other, may be produced, depending on the strength employed. Some
drugs, as chloroform, cause dilation in weak solutions, constriction
in stronger ones; others, as apocodein, cause constriction in weaker
solutions, and tend to dilate in stronger ones. If a moderate dose of
these drugs is injected, both constriction and dilation may follow. In
these cases the dilation may precede the constriction, as with apo-
codein, or follow it, as with chloroform.

2. The statement made by other workers that apocodein is capable
of abolishing or diminishing the action of adrenalin, the strongest
vaso-constricting drug known, while other less powerful drugs, as
chloroform, chloretone, digitalein, and barium chloride, still react, has
been corroborated in this research, and extended to the cerebral
vessels.

3. The results also show that digitalis, which does not-prevent the
action of adrenalin, is capable of diminishing or abolishing the action
of apocodein, or of converting it into a dilating drug.

230 Carl F. Wiggers.

IV. THe PHYSIOLOGICAL SIGNIFICANCE OF THESE REACTIONS.

These varied reactions given by the cerebral and renal vessels are
difficult to explain unless it be assumed that apocodein and adrenalin,
and probably other drugs, act elsewhere than on the muscle substance
of the vessel walls. Since there remains as a point of action for
such drugs only the nerve terminals, or some junctional structure
between nerve end and muscle, the results of this research seem to
supply proof of a cerebral innervation.

The evidence on which it is assumed that adrenalin stimulates the
nerve terminals in the blood vessels depends largely on the paralytic
action of apocodein on the same structure. For this reason it be-
comes necessary to analyze more ‘in detail those reactions of apo-
codein which might cast doubt on the existence of this paralyzing
power.

In the first place, it was found in many experiments, especially on
the renal vessels, that repeated injections of apocodein failed to abol-
ish, but merely diminished, the action of adrenalin. This may be
explained by supposing that in these cases apocodein had a paretic
rather than paralytic effect, lowering rather than destroying the irri-
tability possessed by these particular structures.

In other experiments the action of adrenalin was abolished only
when the vessels were already extremely constricted by the apocodein.
The question arises, Did apocodein abolish or lessen the action of
adrenalin by constricting the vessels so far that the latter could pro-
duce little or no additional constriction? ‘The fact that chloroform,
barium chloride, digitalein, etc., which normally have a much weaker
action than adrenalin, were still able to act after this extreme degree
of constriction had been reached, disproves this. By some this argu-
ment might not be accepted, as it has been asserted that some of
these drugs produce their changes by a rigor process. Experimental
evidence refutes this statement, as the vessels still react to other
drugs (presence of contractility), and passively respond to an increase
in the perfusion pressure (presence of elasticity).

Lastly, doubt is cast on the paralytic action of the drug by the fact
that, contrary to the results of Brodie and Dixon, the vessels remained
tonically constricted after an application of apocodein, and each suc-
ceeding dose caused an additional constriction. In an occasional
experiment there occurred, after many injections of apocodein, a very
slight and temporary dilation, but even then the vessels were far from

The Innervation of the Cerebral Vessels. 231

reaching their normal calibre, whereas, were the reaction due to
paralysis, they should pass beyond. The action of apocodein after
digitalein suggests an explanation for this tonic constriction. The
action of either of these drugs was characterized by a persistent tonic
constriction. That induced by digitalein could apparently replace that
of apocodein in the sense that when the latter drug was injected after
the former, its characteristic action was absent. Digitalein, however,
differed from apocodein in its action in that it could not abolish the
subsequent action of adrenalin. When digitalein was followed by
apocodein, however, the latter drug, although giving no evidence of
a reaction itself, was capable of doing this. These facts seem to indi-
cate that apocodein also has a muscular action which causes a tonic
constriction replaceable by that of digitalein.

In summing up the action of apocodein we find that it apparently
paralyzes some nervous mechanism in the vessel walls, as shown by
its abolishing the action of adrenalin while other drugs still acted.
On the other hand, the tonic state of constriction, shown in its reac-
tion, seems to be due to a stimulating muscular action of the drug.
These paradoxical reactions may be explained by assuming that the
constriction due to muscular stimulation prevented the paralysis of
the nerve terminals from being indicated by a dilation. It is a most
plausible assumption that the action of any peripherally acting drug
is not limited to any definite structure, histological or imaginary, but
that it has a graded affinity for all the structures comprising the pe-
ripheral mechanism. It may be assumed that, as the nerve-end sub-
* stance gradually shades into that of the muscle, the affinity which
the terminals have for a drug is passed over to some extent with
it, there being no point where the action of the drug abruptly
ceases.

By the judicious application of one set of drugs, the action of another
set may be profoundly altered. This may happen because the first
drug destroys the affinity of the second for a certain structure, or
substitutes a similar action for it. The experiments of this research,
though far from being conclusive, have been suggestive enough to
point to a tentative classification of peripherally acting vaso-motor
drugs, not according to any action on one definite structure, but ac-
cording to a system of graded neuro-muscular action. This was done
by the following process of eliminating the point where it seemed
improbable for each drug to act:

333 Carl F Wiggers.

o

1. Apocodein abolishes the action of adrenalin, while weaker drugs,
as digitalis, barium chloride, and chloroform, still act. Therefore
adrenalin is classified as having more affinity for the nervous side of
the peripheral mechanism, while these other drugs have more affinity
for the muscular side. Apocodein, accordingly, must be supposed
to paralyze the nerve-end por-
tion which adrenalin stimulates
D (Fig. 10, A to B).

2. Digitalis does not abolish,
nor react in the same manner as
adrenalin; therefore this drug

probably does not act to any

OIGITAUIS
‘

?---

ee i eee great extent, at least, on the

MUSCLE SUBSTANCE "TRANSITION SuasTance™—Susstance | Same Structure as adrenalin

(Fig. 10, € to D).

FiGuRE 10.— Diagram to express the over- 3. Full doses of digitalis com-
lapping action of a number of vaso-motor pletely prevent the subsequent
drugs. No definite division between muscle tonic constriction of apocodein,
substance (toward left) and nerve-end sub- sometimes even converting it

stance (toward right) is made. Greatest , SSvEee :
points of predilection for various drugs are into a dilating drug, while full

represented by the widest portion of the de- doses of the latter drug merely

crescendo, points least affected by the apex. weaken the subsequent constric-

The uncertainty as to how far action occurs .. Pheer Ss:

is expressed by question marks. Lines 4, B, tion of digitalis. The first of

C, and D are referred to in the text. these results indicates that apo-

codein has a muscular as well
as nerve-end action. The second shows that its affinity for the mus-
cular portion is exceeded by that of digitalis (Fig. 10, C to D).

4. The action of either weak or strong doses of chloroform and
barium chloride is not appreciably affected by large doses of apo-
codein or digitalis. Therefore their action is represented as leaning
still more toward the muscular side.

Present-day methods are still too crude to reveal much more than
the structure for which each drug has the greatest affinity, while the
extent to which it acts on neighboring structures cannot be so clearly
established. A careful analysis of drug reactions, however, indicates
that probably no drug possesses an action absolutely confined to either
nerve terminal or muscle substance. This hypothesis that a graded
affinity exists for all peripherally acting drugs does not interfere with
their use to prove the presence or absence of an innervation of blood
vessels as long as the greatest action of these drugs may be assumed

yy

The Innervation of the Cerebral Vessels. 233

to be on the nervous side. The reactions of the cerebral vessels to
adrenalin and apocodein are such as to necessitate the assumption
that their greatest action is on the nervous side. Thus physiological
evidence of a functioning of the nerve terminals, histologically dis-
covered in the cerebral vessels, is partly supplied by the reactions
obtained in this research.

My thanks are due to Professor Lombard for his useful suggestions
in this research.

THE NUTRITIVE VALUE OF GELATIN. —Ils9 SIGNIFI
CANCE OF GLYCOCOLL AND CARBOHYDRATES
SPARING THE BODY'S: PROTEID:

By JOHN R. MURLIN.

[From the Physiological Laboratory of the University and Bellevue Hospital Medical
College, New York.}

N a recent communication by the writer! it was shown that nitro-

gen equilibrium at or near the fasting level may be maintained in
man and dogs, for short periods of time at least, after approximately
two thirds of the supply of proteid (meat) nitrogen has been replaced
by gelatin nitrogen. This is the highest replacement of proteid by
gelatin thus far reported, for no one has before obtained a replace-
ment of as much as one half of the proteid at the fasting level. This
result, however, was obtained only when the amount of carbohydrate
in the food was very high (two thirds of the total supply of energy),
and when the total supply of energy was considerably in excess of
the requirement. The advantage of high carbohydrate calories was
not unexpected in view of the well-known protecting power over the
body’s proteid of this class of foodstuffs. But it suggested the idea
that the high sparing effect of gelatin itself might possibly be due in
part to carbohydrate capable of being synthesized from it rather
than to the amino-acids which it is capable of yielding.

The line of reasoning was as follows: Landergren* had performed
some experiments which he interpreted as showing the body’s adso-
lute requirement for carbohydrate. For example, by exclusive feeding
of carbohydrate to the amount of 38 to 45 Cal. per kgm., he was able to
reduce the amount of nitrogen output of a man weighing 70 kgm. to
between 3 and 4 gm.; by exclusive fat feeding, to the amount of 54
Cal. per kgm., the nitrogen output was 5.7gm. ‘This great superiority
of carbohydrate in sparing approximately twice as much proteid as

1 MouRLIN, J. R.: This journal, 1907, xix, p. 285.
2 Reviewed by HAMMERSTEN in MAty’s Jahresbericht, 1902, p. 685
, 234 "

The Nutritive Value of Gelatin. 235

fat is met with only in exclusive ingestion of carbohydrate or of fat,
for with a certain minimum of carbohydrate added to the diet, fat
spares proteid as well as carbohydrate itself. Hence Landergren
concludes that the increased proteid combustion which takes place
in exclusive fat feeding corresponds to that quantity of proteid which
is required for the formation of carbohydrate when no carbohydrate
is being fed.

Whether Landergren’s interpretation as to the quantitative require-
ment of the body will stand or not, there is no doubt that under cer-
tain circumstances (¢. ¢., Diabetes mellitus, phlorhizin glycosuria, etc.)
proteid can yield carbohydrate. Moreover, according to modern ideas
of protein metabolism,’ when the amino nitrogen is split off normally
in the form of ammonia and transformed into urea (because it is of
no use to the organism), there remains a carbonaceous residue capable
of being burned. We do not know that in the case of every amino-
acid this residue is convertible into carbohydrate, but it has been
demonstrated that alanin, asparagin, and others can yield dextrose?

It has been shown by Reilly, Nolan, and Lusk,’ that gelatin, like
protetd, can yield carbohydrate. They observed, after feeding gelatin
to a dog which was under the influence of phlorhizin, that the in-
crease in the amount of dextrose eliminated in the urine was as much
as 60 per cent of the weight of the gelatin fed. May not this dextrose
from gelatin (equal in potential energy, gram for gram, with the gelatin
itself) be formed normally also, and be available for the protection of
the body’s proteid? It has long been generally believed that the
nitrogen of gelatin cannot be retained in the body, and that its effect
of reducing the body’s waste of nitrogen is due to the fact that it
furnishes something which can take the place of proteid in combus-
tion. Is that something carbohydrate, z. ¢. the non-nitrogenous resi-
due, or does the gelatin protect the proteid in virtue of the amino
bodies which it furnishes?

It appears, from a study of the literature, that nobody has observed
the effect on protein metabolism, as measured by nitrogen elimina-
tion, of small quantities of carbohydrates fed alone. Landergren’s
experiments show the effects of carbohydrates in large quantities,
and from his results, together with those of Kirchmann* on gelatin

1 FoLIn: This journal, 1905, xiii, p. 117.

2 STILEs and Lusk: This journal, 1903, ix, p. 380.

8 REILLY, NOLAN, and Lusk: This journal, 1898, i, p. 395.
*# KIRCHMANN: Zeitschrift fiir Biologie, 1900, xl, p. 54.

236 Fohn R. Murlin.

fed alone, it may be calculated that the amount of dextrose derivable
from large quantities of gelatin would have a sparing effect equal to
that of the gelatin itself. But we know that gelatin in small quan-
tities can exert an effect almost as great as when fed in large
quantities,! and we do not know that this smaller proportional effect
of gelatin in large quantities may not be due to its incomplete deni-
trogenization.? If it were not completely denitrogenized, the non-
nitrogenous portion would not all be available for combustion. At
all events, there is nothing in these experiments, or in any others on
gelatin or carbohydrates with which I am acquainted, to show that
gelatin in small quantities may not spare the proteid in virtue of the
carbonaceous part of its molecular complex rather than because of the
amino-nitrogen which it affords.

METHOD,

The question may be put to test (1) by feeding an animal whose
fasting metabolism is known, a small quantity of gelatin, and com-
paring its effect with 60 per cent of that quantity of dextrose. It
is scarcely possible that gelatin can yield more than 60 per cent of
its weight as dextrose. Hence, if the reduction in nitrogenous waste
from the body is as great with this amount of dextrose as with the
amount of gelatin from which it could be derived, we may safely con-
clude that gelatin owes its sparing effect to the non-nitrogenous com-
pounds which it can furnish. (2) The amino-acids contained in
gelatin may be fed separately and their effects compared with those
of gelatin itself.

EXPERIMENTS WITH DEXTROSE.

Dog A.— Female fox terrier in good condition and weighing 13.8 kgm., was
made to fast for four days and was then fed with 34 gm. gelatin, contain-
ing approximately 20 per cent of the required energy. This was imme-
diately followed by a three-day period in which 12 per cent of the required
energy was supplied by carbohydrates, and the experiment was terminated
by another fasting day.

1 Cf. KIRCHMANN’S curve, /oc. czt., p. 81.

2 Or even incomplete digestion, since disturbances of digestion are always ap-
parent when gelatin is fed in excess. RUBNER, for example, has reported the
presence of gelatin itself in the urine. It is possible also that the destruction and

elimination of a large quantity of gelatin may itself entail a loss of nitrogen from
the body.

The Nutritive Value of Gelatin. 227

SABIE, Te
Doc A.

ate Cal. of | N N N Body
1906 Weight. Food. require-| in in in | Total | Fast- N
ment. | food. | urine. | feces.| N. | ing N.| spared.
|
July kgm. 5 per cent. per cent.
a) daa. |e as Bee echoed 182500 O09. |b Besa.
SiO, 34gm.
28 13.2 es 20 | 5.242.) 6.932 | 0.13 | 7.062 | 2.6 35.0
g | 13.1 Shee BE RE ayy 1). O40 | ah | OM toe ca NP eS 30.0
30 | 129 sees Bela choy © 1 59404l ... BONS Leta) 24 30.0
31 12.5 |Si02195gm.| 45 | ... | 2566] 0.10 | 2.666] 23 | —13.0
Aug. 904 » gee
i! 12.6 i pe Cer 12 29350 | OME Saeet hk Be —4.5
rose.
18.8 gm. corn- ze
2 12.5 eee 12 0,02841..2205 | O10) 12.305 | 21 8.0
123 fasting bate, tects. fe 1.0664. O10, I 206Gul t 2:0

1 A portion of the gelatin was vomited during the night and was found mixed with
the urine, making analysis impossible. Accordingly the sparing effect for the second
and third days of gelatin feeding are estimated from Kirchmann’s curve (oc. cit.).

It is clear that in this experiment the small quantity of carbohy-
drate which might be available by synthesis from the gelatin fed would
have no sparing effect. The fasting nitrogen, it will be seen, has
fallen, in the eight days of the experiment, from 2.7 gm. to 2.0 gm. If
we suppose this decline to have been a gradual one, and estimate the
fasting nitrogen for the carbohydrate days at 2.3, 2.2, and 2.1 gm.,
respectively, the results indicate an additional waste of nitrogen be-
cause of the carbohydrate ingestion. This I have expressed as a
minus sparing of 13, 4.5, and 8 per cent, respectively, for the three
days.! No particular significance should be attached to the different
figures obtained with the different carbohydrates fed on the three
days.. Except between the cane-sugar day and the dextrose day the
difference is not greater than might be expected between any two
days with the same carbohydrate, and the period is entirely too short
to make the one considerable difference of any value.

Dog B. — Female mongrel, weighing 12.7 kgm., was made to fast for 5 days
and was then given a small quantity of beef heart for three days. For
1 It was learned subsequently that the dog was in the early stages of pregnancy

at the time of the experiment. I am not prepared to say what influence, if any,
this condition might have had on the results as tabulated above.

238 Fohn R. Murtin.

three days immediately following this, 20 gm. of dextrose were added to
the meat, and the experiment was closed with another fasting period. The
dog’s appetite was very poor, so that even the small quantity of meat had
to be given by hand. The dextrose was taken readily dissolved in 100 c.c.

water (see Table II).
TABLE II.

N in | N in} Total N Fasting | Body N
urine. |faeces.| excreted. N. spared.

Weight. Food.

kgm. per cent.

12.04 | 5th fasting day Soea |} Zant) || CHS | ZAS00! cass 2.6
11.98 | 6th fasting day 6500 || ZaO |) OS | ASO. |) socc Ds
S10; 30.gm: beet | ae Ge |e sole) tr eee

heart
30 gm. beef heart 2.474 | 0.16 | 2.634 : Be
30gm. beef heart 2.407 | 0.16 | 2.567 Dae.
SUL SESE) Ne 2.674 | 0.19 | 2864 | 21

20 gm. dextrose

30 gm. beef heart
20 gm: dextrose 2.457 | 0.19 | 2.647 : 2.0

30 gm. beef heart 3
20 gm. dextrose 2.356 | 0.19 | 2.546 1.9

Fasting peel 2-390) | 00601 2-416

Fasting cele | l-d38"| 0106) e598

1 Dog vomited part of the meat, probably owing to the silicic acid given with it as a
feecal mark.

2 Owing perhaps to the poor appetite for the meat, it was imperfectly digested, and
consequently a small amount of meat fibre was found in all the faeces of the two feeding
periods. This was carefully separated, and its nitrogen deducted from the nitrogen
contained in the food. The figure given is the average per day after deduction.

Here the singular effect of a small quantity of carbohydrate to
increase the nitrogen loss is again seen. The sparing of the body
proteid effected by the meat alone is twice as great as the sparing
effected by the meat plus 20 gm. of dextrose.1

Dog C. — Female bull-terrier which had been very much reduced by suckling
a large litter of puppies. The dog was made to fast for eight days and
was extremely emaciated at the time feeding was begun. The nitrogen
in the urine rose from 2.581 gm. on the second fasting day to 3.104 on
the eighth. The dog was fed 21 gm. of dextrose (12 per cent of the energy

1 It is possible that this dog also may have been in the early stages of preg-

nancy. The dog was killed for another purpose before the importance of deter-
mining this point was appreciated. Dog C is known not to have been pregnant.

The Nutritive Value of Gelatin. 239

requirement) for four days, after which another fasting period of two days
was taken. On the second day of this period the nitrogen output rose
suddenly to a very high figure (probably the premortal rise), and the ex-
periment was terminated (Table IIT).

TABLE, Ti

Date | weich N in Nin | Total N| Fasting | Body N
1907. ee urine. | fzeces. |excreted.| N. spared.

Feb. kgm. gm.

18 11.90 7th day of fasting 3.067 0.028 3.095
19 11.62 San CO & sf 3.104 0.028 3.132

20 1152 21 gm. dextrose 2.829 0.039 2.878

21 11.38 3.033 0.039 3.072
22 11.22 3.135 0.039 3.174
23 AZ 3.237 0.039 3.276

24 10.82 Fasting 3.833 0.028 3.861
2) 10.65 3 4.4991 | 0.028 4 eva

1 Premortal rise.

In this experiment there is an evident sparing effected by the dex-
trose. Estimating the fasting metabolism by the amount of nitrogen
which would have been eliminated if nothing had been fed on the four
dextrose days, we learn that the sparing is exactly 0.33 gm. N, or
2.06 gm. proteid on each day. In other words, 12 per cent of the
body’s requirement fed alone in the form of dextrose has protected
the body proteid (held back the nitrogen) to the extent of ro per
cent of the fasting requirement. This is the only dog in which any
sparing has been obtained with 12 per cent of the requirement, — a
result probably due to the extreme state of impoverishment to which
the dog had been reduced previous to the experiment. Judging by
the steady rise in the nitrogen output previous to the feeding, there
could not have been a large reserve of fat available for the fasting
metabolism; hence the living substance was levied upon to a greater
and greater extent. In the other dogs the nitrogen output was on
the decline, as is usual in all but the very last stages of fasting.
Hence it would appear that under the circumstances which this dog
presents, a small quantity of carbohydrate is much more readily sub-

240 Fohn R. Murlin.

stituted for proteid (if that be the action of the carbohydrate) in the
katabolic processes of the body than when, as in the other dogs, suff-
cient fat (it may be presumed) was available. The sparing is slight,
however, even under these circumstances, and is by no means to be
compared with the protecting influence of that quantity of gelatin
from which, at the best, this quantity of dextrose could be derived.
In the next experiment the protecting influence of gelatin is directly
compared with that of 60 per cent of its weight of dextrose.

Dog D. — A small female mongrel in a fairly good condition as regards body
fat, weighing 6.28 kgm. at the beginning of the experiment, was made to
fast for thirteen days. The nitrogen output declined from 2.665 gm. on
the second day to 1.959 gm. on the eleventh day, then remained station-
ary at 2.010 gm. on the twelfth and thirteenth days. For four days there-
after 12 gm. of dextrose (equivalent to 12 per cent of the requirement for
potential energy) dissolved in roo c.c. water were given in five equal
quantities at ten, twelve, two, half-past three, and five o’clock. A single
fasting day followed, and for four days more 20 gm. of gelatin (equiva-
lent to 20 per cent of the energy requirement) were given at a single
feeding. The gelatin used in this and subsequent experiments reported
in this paper was a powdered commercial article ’ found to be entirely free
of proteid (see Table IV).?

There is no sparing in the dextrose period, but about the usual
sparing (31 percent) for the amount fed in the gelatin period. This
dog was absolutely normal so far as could be determined. It is clear,
therefore, that even in the condition of extreme nitrogen hunger in
the dog the protection of the body’s proteid by gelatin is not due to
dextrose, which may be formed from it, but to some nitrogenous
constituent.

In the following experiment on a man it will be seen that the same
must be true also of the human organism.

EXPERIMENT ON MAN (A. I. R.).

Subject was a medical student twenty years of age, slender of build and weigh-
ing at the beginning of the experiment 45.4 kgm. He was first placed in
nitrogenous equilibrium, or nearly so, on a diet containing two thirds

1 PETER COOPER’S, made in New York.
? A 5 per cent solution gave no turbidity with acetic acid and potassium ferro-
cyanide, and no colored precipitate with MILLon’s fluid.

The Nutritive Value of Gelatin. 241

of the estimated requirement for nitrogen in beefsteak, the remaining
third in oatmeal, soda biscuit, etc. Then the beefsteak was replaced by
gelatin containing the same amount of N for six days, at the end of which
time the gelatin was dropped and in its place 60 per cent of the weight of
gelatin was given in dextrose. The gelatin had the usual effect on the appe-

TABLE IV.

Cal.
Weight. require-
ment.

N in Total N
urine. .| excreted.

kgm. per cent
5.00 | 12th fasting day sf 55502 || ZAOKY 2.046

4.86 | 13th fasting day 5¢ saver | 2.010 2.046

4.74 | 12 gm. dextrose

4.70 | 12 gm. dextrose
4.68 | 12 gm. dextrose
4.63 | 12 gm. dextrose

4.60 | Fasting

4.58 | 20 gm. gelatin
4.56 | 20 gm. gelatin 4.156
20 gm. gelatin 4.156

20 gm. gelatin 4.021
Fasting sc 5600 |) decal

Fasting xc sean i 1Ly2ill

1 Not analyzed.

tite, and on the last days of the period was eaten only by the greatest
effort. The subject also experienced considerable languor and disincli-
nation to do mental work. A light wine was allowed by way of a tonic
for the stomach, and on the last two gelatin days a light dose of Zimct.
Zingiberis was given just before mealtime. Two meals each day at eight
A.M. and five P.M. were eaten. ‘The diets follow:

ALL PROTEID NITROGEN.
Breakfast.

40 gm. oatmeal. . . (2.14 per cent N) = 0.86 gm. N and 158 Cal.
75 gm. beefsteak . . (3.4 percent N) = 2.55 gm. Saal i cca

1 1.5 per cent fat estimated.

242 Fohn R. Murlin.

Breakfast (continued).
50gm.cream .. - (0.3 percentN)=3.15 gm. and 220 Cal.

50 gm. cane sugar. . aoe Uy coke 195
35 gm. soda biscuit . (1.68 percent N)= o58gm.N “ 125 “
109m. butter . . « (02 percent N) = 0.01 pm N “9 yom
200 pm. coliee , . - eee 0.05 gm. N “oom
“4.29 pm. N “ “O40ee

Supper. .

12s gm. beefsteak . . (3.4 per cent N) = 4.25 gm. N and 126! Cal.
Bo eMATICe. 4 ss ms == i052 gic =f) eae
75 gm. cane sugar . ies ° 287 Be
50 gm. soda biscuit . (1.68 per cent N) = shale pm. N “. 180.
50 gm.cream . .. -( og) percent N) = 0.15 em. N “\ 220mm
zo em. butter. . « (‘ecrper cent N) = ©.02 em. No“ \ir5cueee
200em\COlce™ .P x. ys Tess o.g5:gm. IN, scone
50 c.c. light wine . . isis 0.02.2m. N’.“) ome
5.85 gm.N “ 1122 °
Total for day .. « - .) “to.os gm. N “rege oe

43 cal. per kgm. ©

Two THIRDS (67 PER CENT) GELATIN NITROGEN + ONE THIRD PROTEID
NITROGEN.

Breakfast.

40 gm. oatmeal . . (2.14 per cent N) = 0.86 gm. N and 158 Cal.
5o.gm.cream . . . “(0-3 percent N) = ¢.15 gm. N “) @eommmee
50 gm. cane sugar. fees Eel 95. =.
35 gm. soda biscuit . (1.68 percent N) =0.58 gm. N “ 125 “
ro'gm. butter « «ss “(oper cent -N) ="o.er gm. N “ee
200 c.c. coffee . . . S555 0.05 gm. N
otal, 2: Gar Geos hey OS oe NT

6c 774 6é

Half at Each Meal.
( 43-5 gm. gelatin . . (15.4 percent N)=6.7 gm.N and 165 Cal.

ls50 c.c. light Rhine wine we o,02:gm. iN. “) somes
Supper.
4ogm.rice . . . (4.3 percent N)= o.52 9m. Nand aq7ear
50 gm. cream ... . (0.3 pericent (N)—= o1¢5 pm. NW eesceee
75 gm. cane sugar. . ere eae 209) ie

1 1.5 per cent fat estimated.
2 50 per cent of total calories were from carbohydrate.

The Nutritive Value of Gelatin. 243

Supper (continued ).

50 gm. soda biscuit . (1.68 per cent N) = 0.84 gm. N and 180 Cal.
Sone: Nitter:. sms.) 46-8 percent N) = o-o2;.em..N) *), 152... “
Saore.c. coliee:«)-.5, |. foes ORC INE EE baccya set any, bs

Pao pm. NN“ -ogb..
Grand fotalforday «= 9.95 gm.N “ 1935. “
== 42 cal. per kgm.

OnE THIRD PROTEID NITROGEN; DEXTROSE INSTEAD OF GELATIN.

Breakfast.
Same asin previous diet . . . . . . 41.65 gm. N and 774 Cal.
Supper.
Same as in previous diet . . . . . - 41.58 gm. N andgg6 Cal.
Weasels dextrose! = 20 a ae ae
Total forday . . . . 3.23gm.N “1858 “
= 4o cal. per kgm.
TABLE V-.

SUMMARY OF EXPERIMENT. FOR DETAILS SEE TABLE X AT END OF PAPER.

Number | Weight | Cal.per| Nin | Total N| N. dif.

pemeeece em: of days.| kgm. kgm. food. | excreted.| per day.

All-proteidN .... . 46.4 43 10.05 104477) —0:39
Two thirds (67% ) gelatin N

+ one third proteid N . 46.4 42 9.95 11.284 | —l.44

One third proteid N only 46.4 40 6.269 | —3.02

1 Average of last two days only.

The amount of nitrogen supplied in the gelatin period was 0.1 gm.
less than in the all-proteid period, but the average loss was fully 1.0
gm. more. In my former paper on the subject of gelatin, I reported
an experiment on myself, in which I was able, with the help of extra
carbohydrate calories, to maintain almost exactly the same state of
nitrogen balance on two thirds (63 per cent) gelatin nitrogen plus
one third proteid nitrogen as on all-proteid nitrogen. The present
subject (not being fond of sweets) was unable to take the extra car-
bohydrate in the form of sugar, and an attempt to readjust the diet so

1 so per cent of total calories were from carbohydrate.

244 Fohn R. Murkin.

as to give two thirds of the total calories in carbohydrate (see Table X
at the end of the-paper for details) by substituting a cornstarch pud-
ding for rice was not successful. The subject experienced nausea
and more or less pain in the stomach. In spite of this, however, on
the last gelatin-proteid day the sparing was only a little less favorable
than on the best all-proteid day, and on the fourth gelatin-proteid
day, when the amount of carbohydrate calories was only 50 per cent
of the total, the sparing was a little better than on the best all-
proteid day. The experiment is not to be construed as contradicting
the former one, for the reason that the conditions were not identical
in the two.

The object of the experiment was to observe the effect of substi-
tuting for the gelatin 60 per cent of its weight in dextrose, and the
result is perfectly positive. The advantage, from the standpoint of
nitrogen loss, of the gelatin over dextrose is very evident, for the
average daily loss in the latter period is more than twice as great as in
the former. The conclusion from this experiment on a man and from
the preceding ones on dogs is, obviously, that gelatin does not pro-
tect the body’s proteid in virtue of dextrose synthesized from it, even
if as much as 60 per cent of its weight should pass through a dextrose
stage. Its value to the organism must consist in the fact that it
contains nitrogen, even though that nitrogen itself may be, as is gen-
erally supposed, immediately eliminated. Possibly a detailed study
of the nitrogen distribution in the urine in fasting and after feeding
gelatin alone will afford some light as to the exact rdle played by the
gelatin nitrogen.

EXPERIMENTS WITH GLYCOCOLL.

It has been customary,! in explaining the behavior of gelatin in me-
tabolism, to refer its inadequacy as a substitute for proteid to the
absence of tyrosin and indol groups inits molecular complex. Kauff-
mann,? in fact, proceeding on this idea, attempted to restore gelatin to
the full nutritive value of proteid by adding to it appropriate quan-
tities of tyrosin and tryptophan and of these two together with cystin.
His conclusion was that the latter mixture ‘“ probably had the full
physiological value of proteid.” Kauffmann’s results, however, have
recently been called in question by Rona and Miiller,? who failed to

' Cf, for example, Tigerstedt in NAGEL’s Handbuch, i, p. 423.

? KAUFFMANN: Archiv fiir gesammte Physiologie, 1905, cix, p. 440.

* Rona and MULLER: Zeitschrift fiir physiologische Chemie, 1907, 1, p. 363.

The Nutritive Value of Gelatin. 245

confirm his findings after adding tyrosin and tryptophan. Whether
it is possible thus to raise gelatin to the physiological status of proteid
by addition of artificially prepared amino-acids must, therefore, be left
in abeyance.

It seems to me, however, that this conception that gelatin falls
short of proteid merely in that it lacks tyrosin and indol or these and
a sulphur-containing group, overlooks the most important (from the
quantitative point of view) constituent of gelatin, namely, glycocoll.
Let us see what is the nutritive significance of this constituent. It is
well known that glycocoll fed alone to an otherwise fasting organism
is completely oxidized at once to urea. Thus v. Briigsch and Hirsch !
fed 20 gm. of glycocoll to a fasting woman on the twelfth day of her
fast, and observed that all of its nitrogen appeared in the urine as urea.”
Samuely 3 had likewise observed this fate (chiefly) of glycocoll nitro-
gen when glycocoll was added to meat and other foodstuffs in the
diet of dogs suffering from artificial anemia. But Liithje* reports a
nitrogen retention with asparagin and glycocoll as the only sources of
nitrogen, provided abundant carbohydrate was fed at the same time.
He thinks there may have been here a synthesis of amino-acids with
carbohydrate, 7. ¢., a formation of ‘ amino-sugar” which can escape
the destructive processes of the body.

Whether the nitrogen retained in Liithje’s experiment was the
nitrogen of glycocoll or of asparagin or both, there was no way of
deciding, just as in my own experiments (where nitrogen was retained
after feeding but little more than the fasting requirement, two thirds
of it in the form of gelatin and one third in the form of meat, but
under the protection of abundant carbohydrate) there was no way
of telling whether the nitrogen retained was nitrogen of gelatin or
meat or of both. But if it could be shown that glycocoll can be
retained, even temporarily, under conditions similar to those under
which I obtained nitrogen retention with two thirds gelatin, we should
have an additional explanation of my results; while if it should prove
that the retention with glycocoll is oz/y temporary, even under the
most favorable circumstances for retention, we should have strong
ground for believing that the inadequacy of gelatin as a source of

1 Briiescu and Hirscu: Zeitschrift fiir experimentelle Pathologie und Therapie,
1906, p. 638.

2 This was not the case with alanin and leucin.

8 SAMUELY: Deutsches Archiv fiir klinische Medizin, 1906, p. 220.

4 Lutuje: Kongress fiir innere Medizin, 1906, p. 440.

246 Fohn R. Murln.

nitrogen might, in large measure, be due to its high content of
glycocoll, and this condition would need to be taken into account
in any attempt to raise gelatin to the physiological status of proteid.

METHOD.

It has appeared in my experiments with gelatin! that the tendency
to nitrogen retention is stronger the lower the proteid condition of
the animal at the time of feeding. For the purpose of testing the
power of retaining glycocoll nitrogen, therefore, I prepared the dogs
by subjecting them to long periods of fasting or under-nutrition.

Dog D. — Female mongrel, weighing in good condition 6.3 kgm., was fasted
for thirteen days, during which the weight fell to 4.8 kgm. Then for two
weeks more the dog was kept in undernutrition for the purpose of the
dextrose and gelatin experiment reported on page 240. At the end of
the month the weight had been reduced to 4.2 kgm. ‘The total output
of nitrogen on the second day of a fasting period taken at this time was
1.247 gm.” Ten days later, on a diet containing 0.491 gm. N in the form
of beef heart and carbohydrate enough to make 150 Cal. per kgm. (weight
still, 4.2 kgm.), the dog was almost in nitrogen equilibrium (— 0.021 gm.
N). To this diet were then added 5 gm. of glycocoll, containing 0.951
gm. nitrogen, bringing the total amount of nitrogen ingested up to 1.442 gm.
This was continued for five days, after which the glycocoll was dropped

’ for two days and the experiment was terminated with a fasting period.
The results are given in Table VI.

On four of the five glycocoll days there was retention of nitrogen,
—a total for the period of 0.259 gm. Whether the nitrogen retained
was glycocoll nitrogen or nitrogen of the meat, we do not know posi-
tively. All we can say with certainty is, that so long as the diet con-

1 Muruin: Loc. cit., p. 303.

2 By feeding for a period of five days 6.4 gm. of proteid-free gelatin containing
0.990 gm. N and supplying over 100 Cal. per kgm. of energy (chiefly in the form
of carbohydrates), the output of nitrogen was reduced to 0.406 gm. —less than one
third of what it was on the fasting day just preceding. Theoretically it should have
been possible now, by adding to the diet this one-third of the fasting requirement
for nitrogen in the form of proteid, to establish perfect nitrogen equilibrium.
Some difficulty, however, was experienced ; for when 4.07 gm. of plasmon, contain-
ing 0.495 gm. N, were added, making the total supply 1.485 gm., there was a loss
of nitrogen on the second, third, and fourth days amounting to 0.114, 0.249, and
0.148 gm. respectively. Since the total supply of energy was not changed, it
would appear that the utilization of the plasmon itself involves an expenditure of
nitrogen,

_—————

The Nutritive Value of Gelatin. 247

tained but little more than one third the fasting requirement for
nitrogen, all of it in the form of proteid, there was considerable loss
of nitrogen from the body; but as soon as the glycocoll, containing
approximately two thirds the requirement for nitrogen, was added,
there was nitrogen retention.

TABLE VI.

Food; Total | Total | N in

Weight. source of N. cal. | N fed.| urine.

Kgm. per kgm.
4.20 16 gm. beef heart 109 | 0.491 | 0.780

4.20 1501! | 0.491 | 0.713
4.26 5 0.491 | 0.474 |

4.34 Fasting 2 SO6¢ slaise) ||) O194S

. beef heart
CEE : . glycocoll

. beef heart
1.26 . glycocoll
. beef heart
. glycocoll

1.324
1.324

4.33 | 1.429

Ps 6 . beef heart
4.36 - glycocoll 1.293

. beef heart z
4.48 . glycocoll & | 1.293

4.54 . beef heart

4.48 eo 4

4.34 Fasting

4.24 -

1 Raised by addition of 50 gm. cane sugar.
2 For the purpose of removing glycogen previously stored.
8 Dog vomited during night, contaminating urine.

The retention, however, is not permanent; for on the day follow-
ing when the glycocoll was dropped, and ‘there should have been ni-
trogen equilibrium, as there was previous to the glycocoll feeding,
there was a loss of 0.435 gm. N. It is more than probable that
this includes the very nitrogen retained during the previous period,
and it is perhaps safe to say that the nitrogen thus retained tem-
porarily represents glycocoll itself.

Dog EB. — Very much like dog D, weighing in good condition 5.4 kgm. Was
fasted for ten days, during which the weight fell to 4.6 kgm. The nitrogen

248

Fohn R. Murlan.

output in the urine on the last fasting day was 1.697 gm. For five days
immediately following this 1.506 gm. N were given, two thirds of it
(1.006 gm.) in the form of proteid-free gelatin and one third (0.500 gm.)
in the form of beef heart. Then for two days glycocoll was substituted
for gelatin ; and finally, the glycocoll was dropped and the beef heart con-
tinued alone for two days. A total energy supply of 130 cal. per kgm.
(chiefly carbohydrates) was maintained throughout. Only the urine was
analyzed. The conditions throughout the feeding periods being so nearly
the same as regards absorbability, it was not considered necessary to analyze
the feeces (Table VII.)

TABLE VIL.

1907, (REAR) PEE S SES a ane a ae ait ce ae
April | kgm. | el ey | per day

29 4.72 | 10th day fasting ar Axon || HELO Elk cere eer dhe
30 | 462 | eee eae 130 | 1.506:)".- 25) es]. Seen ee

tl AI6 era eens 130, +1 1.506), sescl| sec coe

2 Nbdeae Ge Peer wee 130 | 1.506 | 1:593 | —0.087 | 1.55

Bo 482 es ene 130 | 1.506 | 1.525 | 0.019] +0.053 | 15

a 980) te oe ee 130. | 1.506 | 1.559 | -0.053 1.45

Bt Be) Wes on ences 130 | 1.508) 1.559 | —0.051 lo ee |i

Be 290 ee eee 130 | 1.508 | 1.898 | —0.390)) © 1.35

7 4.86 |16.3 gm.beefheart only; 130 | 0.500 | 1.050 | —0.550 lage 1.3

8 4.80 | 16.3 gm.beef heart only) 130 | 0.500] 0.779 | —0.279 J 1.25

9 | 482 | Fasting d. | Sons 10017 | Gleg | Senge
10 4.68 | Fasting 5c sraveror | el 2 || Meters 5005 125
132| 4.32 | Fasting oo Newest LB8S1 "2k. | cee

1 Not analyzed. 2 Fifth fasting day.

Estimating the decline in the fasting proteid metabolism from the
time the feeding began, at the rate of 0.5 gm. N per day, we find

that
the

on the best gelatin day, when the animal was receiving exactly
requirement of nitrogen (1.5 gm), two thirds of it in the form of

gelatin, and the other one-third in the form of proteid, there was a
sparing of- practically 100 per cent, and the average sparing for the

last

two gelatin days was 90 per cent.

The Nutritive Value of Gelatin. 249

The amount of nitrogen in the urine on the first glycocoll day is
exactly the same as on the last gelatin day. This might be taken on
first sight to mean that glycocoll has an equal “sparing power” with
gelatin; but it is probably mere coincidence, meaning rather that on
the first day some of the glycocoll escaped oxidation. On the sec-
ond day there is a considerable increase in the amount eliminated, or,
calculated as a “‘ sparing,’ 84 per cent for the two days, as against
go per cent for the best two gelatin days. When we remember that
the amount of nitrogen fed on these two days was considerably
greater in proportion to the absolute requirement, as determined by
the fasting metabolism, than on the gelatin days, we see that from
the standpoint of any sparing effect the glycocoll suffers still fur-
ther by comparison with the gelatin.

With the meat alone the loss from the body is twice as great as
with the meat and glycocoll. Here, again, we must give a guarded
interpretation of the results. Do the figures mean that glycocoll
serves in some way to reduce the destruction of the body proteid,
which immediately rises-as soon as the glycocoll is withdrawn, or
do they mean that some of the glycocoll has escaped oxidation until
the meat period? The difficulty, of course, lies in the fact that we do
not know whether the nitrogen held back is the nitrogen of glycocoll
or the nitrogen of the meat or the nitrogen of the body itself. The
presumption is, I think, that the nitrogen held back is the nitrogen
of glycocoll itself. The fact that the loss is twice as great on the day
immediately following its removal from the diet as on the second day
would favor that interpretation. Besides, judging from the experi-
ment with dog D (April 6, Table VI, page 247), it is likely that ona
third day with meat alone the dog would have been in nitrogen equi-
librium, and it is quite possible that this would have been the case
on May 5th and 6th if glycocoll had not been fed. That is to say,
we know that the conditions for nitrogen equilibrium were already
present without the glycocoll.

It would simplify matters if glycocoll alone were given as the
source of nitrogen. In the following experiment this is the case. A
large quantity of carbohydrates and a small quantity of lard (10 gm.),
making, all told, a supply of energy equal to 142 Cal. per kgm., were fed
with glycocoll for three days and then for two days without the gly-
cocoll. Assuming that all the glycocoll was absorbed, I paid no
attention to the fzecal nitrogen. The results are given in Table VIII.

The non-nitrogenous diet should have been continued for one or

250 | Fohn R. Murln.

two days more, since, as Landergren’s 1 experiments have shown, it is
possible with carbohydrates alone to reduce the output of nitrogen to
about one third of what it would be in fasting. At the same rate of
decrease a third day in this period would have shown a nitrogen waste
of 0.437 gm., which is just one third of the assumed requirement (1.3
gm.) for the beginning of the glycocoll period, though probably a
little more than one third for the day in question. The average loss

TABLE VIII.

Weight N in
in kgm. urine.

4.32 5th day fasting sex nd 1.288

4.22 7 gm. glycocoll 1.457

4.38 1.612
4.50 2.006
4.54 ne 0.987
4.54 see 0.713
0.437(?)

of nitrogen, however, for the three days during which glycocoll was
fed was only 0.359 gm. This shows either that glycocoll has caused
some of the body nitrogen to be held back, z.2., it has “ spared” some
body proteid from destruction, or that some of the glycocoll itself
escaped oxidation and elimination. In either case it is clear that the
effect is not permanent; for as soon as the glycocoll is dropped from
the diet an increase in the amount of nitrogen eliminated takes place.

Again, the weight of evidence is that we have to do here with a
temporary retention of glycocoll itself. Two considerations at least
favor this idea: First, if glycocoll had not been fed on these three
days it is practically certain that the loss of nitrogen on the two car-
bohydrate days (May 17th and 18th) would have been considerably
less. Secondly, if glycocoll were to reduce the metabolism of pro-
teid in the body itself, the reduction, we should expect, would be
greater on the second day than on the first and possibly also greater
on the third than onthe second. Instead of that we find the “ pro-
tection”’ to be less and less. A much more plausible explanation is

1 Loc. cit.

The Nutritive Value of Gelatin. 251

that on the first day more of the glycocoll escaped elimination than
on the second, and more on the second than on the third.

The observation of Parker and Lusk! that the amount of glycocoll ?
formed in metabolism and capable of being removed from rabbits in
combination as hippuric acid is greater on the first day of the admin-
istration of benzoic acid than on subsequent days of a fasting period,
may be cited as favoring the possibility of a temporary storage of
glycocoll.

In all three of the experiments here reported, therefore, it appears
that we have a temporary retention of glycocoll. But glycocoll fed
alone to an otherwise fasting organism is at once completely oxi-
dized to urea and removed. This is shown by the experiments of
v. Briigsch and Hirsch, already cited (p. 245). This is ordinarily the
case also in feeding glycocoll with other foods, as may be seen in the
experiments of Samuely. How, then, shall we account for a tempo-
rary retention of glycocoll such as we have witnessed in the above
experiment?

SIGNIFICANCE OF CARBOHYDRATES,

In my previous paper,’ and in fact as early as 1904,‘ I drew atten-
tion to the large part played by carbohydrates in obtaining the com-
paratively high replacement of two thirds the fasting requirement of
proteid by gelatin with maintenance of nitrogen equilibrium. Liithje
has recently emphasized the same factor in obtaining nitrogen re-
tention with asparagin and glycocoll, and, as already quoted (page
245), believes that the important condition is a chemical union of the
carbohydrate with the amino-acid (amino-sugar). Still further sup-
port for this idea of aspecific relationship between nitrogen retention
and carbohydrates is to be found in the old experiments of E. Voit and
Korkunoff,> by which it was learned that with carbohydrates much

1 PARKER and Lusk: This journal, 1900, iii, p. 472.

2 The statement of these authors that ‘‘in a fasting rabbit which had been fre-
quently fed with lithium benzoate the amount of glycocoll eliminated in the urine
as hippuric acid compared with the total nitrogen eliminated indicates that 4.0
grams of glycocoll may be derived from the metabolism of every too grams of the
body’s proteid,” becomes very striking in the light of the work of ABDERHALDEN,
GIGON, and Strauss, which shows (Zeitschrift fiir physiologische Chemie, 1907, li,
p- 321) that 100 gm. of the proteid matter composing the rabbit’s body may yield
3-27 gm. of glycocoll by hydrolysis.

8 MURLIN: This journal, 1907, xix, p. 307.

* Proceedings American Physiological Society, this journal, 1905, xiii, p. xxix.

5 Voit, E., and KorkunorFF : Zeitschrift fur Biologie, 1895, xxxii, p. 58.

252 Fohn R. Murtn.

less proteid is necessary for maintenance of nitrogen equilibrium than
with fats; in those of Parker and Lusk,! where it may seem that the
output of glycocoll in combination with benzoic acid is smaller when
carbohydrates are fed than in starvation; in those of Landergren,?
where it is observed that carbohydrates reduce the metabolism of
proteid considerably more than do fats; and finally, in those of Heil-
ner,’ where a reduction of from 12 to 25 per cent in the proteid
metabolism (nitrogen output) was obtained by feeding carbohydrate
enough to replace the fat of the fasting metabolism. Whether we
call this specific relationship, with Landergren, a “specific need” on
the part of the body for carbohydrate, in order that it may protect
its living substance, or refer it to a specific affinity between the car-
bohydrate and the amino-acids by which a larger molecule is formed
and the nitrogenous bodies escape oxidation for a longer time, in
accordance with Liithje’s idea, there is sufficient warrant for the
belief that in the case of glycocoll in the experiments above reported
the essential condition for the temporary retention is the presence in
the circulation of abundant carbohydrate.

INFLUENCE OF EXTRA-METABOLIC CARBOHYDRATE ON NITROGEN
EXCRETION.

In the first section of this paper it was seen that small quantities of
carbohydrate (12 per cent of the energy requirement) exercise little
if any effect on the nitrogen output of the body. And yet, from the
considerations reviewed in the preceding paragraph, we have every
reason to believe that carbohydrates in large quantity play a deter-
mining part in the disposition which the body can make of nitroge-
nous substances. In attempting to reconcile these facts it occurred to
me that it might be only the excess carbohydrate — that which is
not needed for purposes of metabolism, or what we might call extra-
metabolic carbohydrate — that is especially significant in aiding the
retention of nitrogen.

It has been the general belief that carbohydrates protect the body
proteid, z. ¢., tend to reduce its destruction, by taking its place asa
source of energy for the vital processes. If this be the only way in
which carbohydrates can protect the body proteid, then its action, as

1 PARKER and Lusk: Loc. cit.
2 LANDERGREN: Loe. Cit.
8 HEILNER: Zeitschrift fir Biologie, 1906, xlviii, p. 144.

The Nutritive Value of Gelatin. 253

judged by nitrogen excretion, ought to be proportional to the amount
fed until the total energy supply of the body is covered, beyond
which its action would be 2/7. Butif, as might be presumed from Lan-
dergren’s experiments with carbohydrates fed exclusively, where from
7 to 14 Cal. per kilogram were in excess of the requirement, and from
my own with carbohydrates fed with gelatin, where! from 10 to 20
Cal. per kgm. were in excess, carbohydrates act in some other way,
then, as the amount fed exceeds the requirement, its influence on the
amount of nitrogen eliminated ought to be more pronounced. With
these considerations in mind the following experiment was planned:

Dog F.— Same as dog C, page 238. For nearly two months previous to this
experiment the dog had been on a rich proteid diet and was in fairly good
condition as regards body fat, weighing 13.54 kgm. After fasting for four
days the animal was given successively increasing quantities of carbo-
hydrates for periods of two days each, alternating with fasting periods of
one day. In all of the urines but two creatinin and ammonia as well as
total nitrogen were determined. No attention was paid to the fecal nitro-
gen. ‘The urine at no time gave any reaction for albumin nor any reduc-
tion of Fehling’s solution, even after boiling with HCL (Table IX).

Estimating the energy requirement of the dog by Rubner’s? method
employing v. Meeh’s formula, we get for a body weight of 13 kgm.,
646 Cal. Supplying 12.5 per cent of this requirement in the form of
dextrose (22 gm.), we get a mean sparing for the two days of 0.279
gm. N, or 8.3 per cent of the fasting requirement (3.345 gm.). In-
creasing the supply to 25 per cent of the requirement (42 gm. cane
sugar) gives us a sparing of 0.460 gm. N, or 13.3 per cent of the
fasting requirement. In other words, by doubling the supply of car-
bohydrate calories we have increased the sparing by only 5 per cent.

Through an accident the urine for June 6th became contaminated,
so that the figure obtained for the total N, was plainly too high. The
number given for that day is, therefore, estimated so as to make this
period yield the same percentage increase as the preceding period.
The surprising thing is that by so doing we make the increased spar-
ing resulting from the next 100 per cent increase in the supply exactly
the same, namely, 5 per cent. That is, from each doubling of the
energy supply, up to 100 per cent of the requirement, we get the
same percentage increase in the effect on nitrogen elimination.

LLC, Clits DP, 205.
2 RUBNER: Zeitschrift fir Biologie, 1885, xxi, p. 370.

Fohn R. Murlin.

gt a ree
suse yg Nt)

Sutieds uvayy

254

‘Aep Suljsej puosas 107 pajeurnjsy z

cOb >
BTSs
L8LT

SIE'b Z8S'%
z (SZ2't)

62+’

€SLZ

SLS°€

bEO'r
1 SS8L'G
IIe
Ors €
C567
LIO'E
Shee
+866
Srl

N
Sulysvy
uv

“ULULJEIID

‘SOLIO[VD

“XI WIAVL

*poyeuinjss st oansy sy 7

”

* Suysey

IVBNS 9uvd ‘US OLT
dSO1}XOpP “WS OZ
YOIBSULOD "WIS Of}
Ome * SUSE
Ie3ns aura ‘ws OT
9SO1}XIP "UIS OZ
YIIVJSUIOD “US Cp
eS es onsem
IvBns auvd ‘ws $9
JeBns guvd “ws +9

= 8 + » Sunsey
Iesns auvo “wis Zp
esns auvo ‘ws Zp
a ice a Sunse gq
* @sO1}Xap ‘Ws ZZ
* QS8O1]XOP “UID ZZ,
* Sunsey Lep yp

“ws y
IYS19 M

The Nutritive Value of Gelatin. 255

It will be observed that there is a steady increase in the fasting
nitrogen metabolism throughout the experiment. This rather un-
usual phenomenon was observed in this same dog in the experiment
reported on page 239, and is interpreted here, as it was there, to
mean that there was not sufficient fat in the body to meet the needs
of the fasting metabolism. This is rendered probable by the fact that
the dog became extremely emaciated in a very few days. It is evi-
dent, in view of this steady increase in the fasting nitrogen, that the
figure obtained on June roth does not represent the real fasting me-
tabolism, or, in other words, that the carbohydrate fed on the two
preceding days contained something more than the requirement of
potential energy; hence some glycogen was stored, which served to
keep back the body nitrogen. If a second day of fasting had been
taken, the figure would probably have been close to the mean between
that for June 7th and that for June 14th, or about 4.228 gm. This
figure is used in estimating the fasting requirement for the two feed-
ing periods adjacent thereto. The fact of the increase in fasting
metabolism, instead of the usual decrease, as well as the two defects
in the record just pointed out, may well be held to invalidate the ex-
periment to some extent as a basis for arriving at any definite law of
the influence of what we may call z#tva-metabolic carbohydrate on
nitrogen elimination. At all events, it will be necessary to confirm
the facts before enunciating the law as above indicated.

When, however, the amount of carbohydrate is increased beyond
the requirement of potential energy (extra-metabolic), we see a
marked difference in the effect on nitrogen output. With 100 per
cent (or slightly more, as it proves) of the requirement supplied by
carbohydrate, the reduction in the amount of nitrogen eliminated was
23 per cent. Increasing the supply to 125 per cent gives us a reduc-
tion of nearly 50 per cent. In passing from 25 to 50 per cent of the
requirement (June 2d to 5th) an increase of exactly the same abso-
lute amount (42 gm. cane sugar) gave an increased sparing of not
over 5 per cent, whereas here the increased sparing is more than 100
per cent. The contrast may be more sharply expressed in this way:
so long as the total amount of carbohydrate is below the full require-
ment of potential energy, doubling the supply gives a five per cent
increase in the effect; beyond the full requirement a 25 per cent
increase in the supply doubles the effect.

These facts certainly speak for some influence of the carbohydrate
on protein metabolism (nitrogen elimination) other than that of sub-

256 Fohn R. Murtlin.

stitution in combustion. Whether that influence is to be explained
by a chemical combination of the unburned sugar with amino-acids,
as Luthje supposes, or by some combination necessary to the stor-
age of glycogen, or by a reduction of the metabolic processes through
osmotic effects, or in some other way, must be left for future
researches.

CREATININ AND AMMONIA.

With the hope of discovering what fraction of the total nitrogen is
especially affected by the carbohydrate, a record was kept of the
creatinin (Folin’s method) and of the ammonia (Schaffer’s method)
for every day but two. This record, based on the single experiment,
is too brief for more than passing mention.

Up to the point where the full energy requirement is supplied by
the carbohydrate there is a noticeable tendency for the creatinin out-
put to be higher on the carbohydrate days than on the intervening
fasting days. Within the same limit the tendency with respect to the
ammonia is just the reverse — higher on the fasting days than on the
food days. From the 100 per cent limit on, the creatinin exhibits
no fixed relationship, but the ammonia is markedly higher on the
food days than on the fast days.

SUMMARY AND CONCLUSIONS.

1. The well-known action of gelatin to protect the body’s proteid
is not due to the influence of any dextrose which may be synthesized
_from it in the course of metabolism. Its value as a proteid-sparing
agent must consist in the fact that it contains nitrogenous bodies.

2. Glycocoll, which is the chief amino-acid contained in gelatin,
when fed with abundant carbohydrate, either as the only source of
nitrogen or together with proteid (beef heart), can be retained tem-
porarily in the body. This fact may serve in part to explain the
unusually high replacement of proteid by gelatin, with maintenance of
nitrogen equilibrium, which I reported in a previous paper. The
fact that glycocoll cannot be retained permanently, even with large
quantities of carbohydrate available, may serve in part to explain
the inadequacy of gelatin as a source of nitrogen, and must be taken
into account in any attempt to “restore” gelatin to full proteid
value.

3. A specific relationship has been shown to exist between carbo-
hydrates ingested and the elimination of nitrogen (or protein me-

The Nutritive Value of Gelatin. 257

tabolism, as measured by nitrogen output). Carbohydrate, not
needed for combustion (extra-metabolic) is far more efficacious in
reducing the nitrogen output (therefore favoring the retention of
proteid) than carbohydrate coming within the requirement for poten-
tial energy. This fact indicates the importance of abundant carbo-
hydrates for convalescence and growth, and may explain the almost
universal craving for sweets, especially in the young.

Fohn R. Murlin.

£60°S—
6c
ale
bOS'0—
£9L'o—
1vc'0—-
bby I—
try I—
ILeo—
00e'0—
SO

cst+'0—

‘UNeIAyD 1

(MLV) LNaqaLg Ivolaay NO LNAWINAdXY

x

aTdvL

ror |. 8
or | L
ror | 9
ror | ¢
Co | +
yor | €
or | Z
vor | 1
“990
99 | O€
SO | 62
rsh | 82
Sb | LZ
rsh | 92

TaseBh

"wi3¥ UL) “906T

IYSII A | 23"

‘IIIT g “YOIBySusor) z f

| EE Et eos ieee ce eee
szos_| gogo | stos leze | op | ssst | +2 | OS | Cor | Of | OOT | S2t | sOr | SB | OF | ***
6sz9 | 9090 Isto |eze | ob | gsst | +2 | OS | Ooh | O€ | OOT | Szt | sOb | $8 | Ob | *°*
czr9 | gov | ztss jeze | ob | gsst | 42 | OS | OOF | OF | OOT | Szt | sOh | 98 | OF | °°"
ezror | oss'o | 46z'6 1296 | +p | 902 | oz | os | oor | oz | OS | Szt | sO | Tet | Ob [rset
sse-zt | ocso lesstt|zoe | +b | +02 | os | os | oor | oz | OS | S2t | OL | OZ | OF |rS'€+
16rot | osso | 1986 |so6 | ze | -scor | °° | OS | OOF | OF | OOT | SzT | Ob | SS | OF |rS"Eb
POEL | sso | +9S0T|so6 | ze | seor | ** | OS | OOF | O€ | OOT | S2t | OF | SB | Ob jrS'Eb
rect | ogo | 49sot|see | zp | seer | ** | OS | ooh | OF | OOT | SzI | OF | SS | Ob |rS"e+
1zzzt | o¢s'o | test] 66 | Te | seer | ** | OS | Oop | O€ 4 OOT | S2t | OF | SB | OF [LS'er
oscot | o6t't |ost6 |soot| sb | tet | ** | os | oob | O€ | OOT | S2t | Ob | S8 | Or | O02
sesot | ott | sseo ese | ce | ocer | ** | OS | OOF | Of | OOT s| S2t | OF | SL | OF | 002
“| g6R.|-5,, (S90L1 oe | S002: 4s || 0S"| Gor! 08 | Sk | .G0T | 09.7) .O0L || Oyj gene
Zot | ost't | zzs6 |zsot| sb | z90z | ** | OS | OOF | OF | OOT | ZIT | 09 | OOT | OF | 002
veo fowa} sou) | RN) og, |S 52/23 | Sat amnol Z| ont [Une eo | Soe

ur N
4yndjno N *SOLIO[VD “sues Ul yarp Ape

PERISTALTIC: RUSH,

By S. J. MELTZER anp J. AUER.

[From the Department of Physiology and Pharmacology of the Rockefeller Institute for
Medical Research.|

INTRODUCTION.

ao writers on the motor phenomena of the intestinal tract
distinguish at present three types of movement in the small gut.
(1) Pendular movements: rhythmical, swaying motions, which
apparently contribute but little to the forward expedition of the
intestinal contents. According to Bayliss and Starling, and to
Cannon, the pendular movements are chiefly concerned in the
thorough mixing of the food with digestive fluids; they are the
essential factors in the “rhythmic segmentation’”’ of the intestinal
contents observed by Cannon. (2) Peristaltic movements, which
consist in a contraction of the gut above a food mass and a relaxa-
tion below it; they are chiefly concerned in carrying the food
through the intestines in the aboral direction. The progress is very
slow, I cm. in two to ten minutes in a fasting state and in thirty
to forty seconds after a meal (Cash). The peristaltic waves never
travel far, the same wave never running through the entire length
of the intestines. (3) “ Rollbewegungen,”’ a fast, running move-
ment extending over the whole or a large section of the small in-
testines. It is this motor phenomenon we intend to deal with in
the present paper.

b

HISTORICAL.

This phenomenon was first observed and described by van Braam
Houkgeest! with Sanders-Ezn. The observations were made on
rabbits whose abdomen was opened and the intestines observed while
these parts of the animal were submersed in a warm saline bath.

1 HOUKGEEST, VAN BRAAM: PFLUGER’S Archiv fiir die gesammte Physiologie,
1872, vi, p. 266.
259

260 S. F. Meltzer and F. Auer.

When the animal was killed by suffocation or by inhalation of CO,,
a few seconds after the convulsions attending asphyxia ceased, a
strong peristaltic contraction set in, either at the pylorus or a little
distance away, which energetically and rapidly drove the contents
before it. This rapid peristaltic wave ran without interruption to
the czecum, especially when the duodenum was strongly filled with
contents. Sometimes another wave followed which originated in
the ileum, and which also ran down in the same manner to the
cecum. Mostly, however, the duodenal wave was the only one
which was observed. When the duodenum was less filled, or when
the animal was already for some time in the saline bath, the peristal-
tic wave did not reach the caecum, but stopped at the lower end of
the jejunum or at the upper end of the ileum; in the latter case
another similar wave arose later at some part of the ileum which
carried the contents rapidly to the cecum. None of the waves ever
continued their course in the caecum.

The rapid run of the peristaltic wave through the small intestine
gave the coils the appearance of turning wheels; the movements
were therefore designated by the investigator as “ Rollbewegungen.”’
At first in the protocols, but later for lack of a better expression,
van Braam Houkgeest accepted this term in the final presentation
of his results. Most of the subsequent writers have adopted this
term for the description of this special form of intestinal peristalsis.
Van Braam Houkgeest termed them in his protocols sometimes also
“post mortem movements ”’ on account of their occurrence only after
death of the animal. Only in one case did he succeed in reviving
the animal by prolonged artificial respiration after the Rollbewe-
gungen had made their appearance. Later on van Braam Houkgeest
was led to believe that under certain conditions he was able to pro-
duce the Rollbewegung also while the animal was alive. The dis-
cussion of this, however, will be deferred until later.

Observations of a similar character were also made by Engel-
mann? (with van Brakel) about one year previous to the publication
of van Braam Houkgeest. Ina cat, which was killed by chloroform,
and whose intestines were examined about half an hour after death,
a mechanical stimulus applied to some part of the intestines caused
a strong local constriction which ran down as a _ peristaltic
wave to the cecum with a rapidity of about 4 cm. in a second;
upward the contraction ran as an antiperistaltic wave, stopping at

1 ENGELMANN: Archiv fiir die gesammte Physiologie, 1871, iv, p. 33.

Peristaltic Rush. 261

the stomach. In animals which were killed by exsanguination
(rabbits, dogs, and cats) spontaneous movements which originated
in the duodenum were seen to run down at times to the ileo-czcal
valve; mostly, however, the rapid wave stopped at the ileum.

In the experiments of van Braam Houkgeest, as well as in those
of Engelmann, these movements were seen to occur after death;
they were characterized by the rapid propagation of the peristaltic
wave; sometimes the wave ran through the entire small intestine
without stopping.

Nothnagel! later described Rollbewegung, which he studied in
the living rabbit. He describes it as a violent peristalsis running
rapidly over some 20 cm. of the small intestine and coming sud-
denly to a standstill, as he suggests, by some inhibitory influence.
He has never seen the wave extending over the entire length of the
small intestine. Nothnagel believes that the distention of the in-
testines by liquid and gas is the cause of these Rollbewegungen, but
asserts at the same time that observation of these movements had
to be confined to their accidental ocurrence, and that it was impos-
sible to produce the Rollbewegungen at will. -

Bokai,2 on the other hand, stated that the presence of CO,, CE
or H,S in the intestines will cause “ rollende Bewegungen.” These
movements, when once started, are not propagated to a distant sec-
tion of the intestines where these gases are absent. The gases
produce violent but only local contractions. The presence of oxygen
in the Jumen of the intestines stops these movements; nitrogen and
hydrogen exert no effect.

From the writings of Mall,3 Cannon,‘ and others it appears that
the “ rollende Bewegungen” of Bokai are considered as identical
with the Rollbewegungen of van Braam Houkgeest.

Cannon describes the Rollbewegungen “as a rapid movement
sweeping the food without pause through several turns of the gut,”
and ‘‘is frequently seen when the food is carried on from the duo-
denum ”’; and states “ that it may readily be produced in other parts
of the small intestine by giving an enema of soapsuds.”’

1 NOTHNAGEL: Beitrage zur Physiologie und Pathologie des Darms, Berlin,
1884. See also NOTHNAGEL’S Handbuch der speziellen Pathologie und Therapie,
Die Darmbewegungen, 1898, xvii, p. 2. D

2 BoKAI: Archiv fiir experimentelle Pathologie und Therapie, 1887, xxiii,
p- 209.

8 MALL: Johns Hopkins Hospital reports, 1896, i, p. 51.

4 CANNON: This journal, 1902, vi, p. 251.

262 S. Ff. Meltzer and f. Auer.

We shall return now to a few statements of van Braam Houk-
geest. In the first place, he states that the Rollbewegungen did not
occur post mortem if both vagi were previously cut; only irregular
Rollbewegungen and constrictions appeared in various sections of
the small intestines. If however at the onset of these irregular
contractions the peripheral ends of the cut vagi were stimulated,
regular Rollbewegungen then ran through the entire small intestine.
Furthermore, if in a living rabbit both splanchnics were cut, then
stimulation of the peripheral end of either vagus would cause first
a strong constriction of the median part of the stomach which was
followed soon by a “true Rollbewegung,’ which in most cases
ran through the entire small intestine. In strong animals such
results could be obtained repeatedly by repeated stimulation of the
peripheral end of either vagus. Van Braam Houkgeest believes that
the Rollbewegungen produced by the stimulation of the vagi, after
cutting the splanchnics, are identical with the post mortem Roll-
bewegungen.

In summing up this review we find that Engelmann as well as
van Braam Houkgeest have observed in dead animals Rollbewe-
gungen of spontaneous origin running through the entire intestine;
that Engelmann produced in dead animals “ Rollbewegungen”’ by
mechanical stimulation. Van Braam Houkgeest produced true Roll-
bewegungen in living animals by stimulating the peripheral end of
one vagus after both splanchnics were cut. Nothnagel has seen
in the living animal Rollbewegungen, but they never traversed the
entire small intestine, and they could not be produced at will. Bokai
produced by the introduction of CO,, CHy, and H5s into the lumen ~
of the intestine of living animals violent peristaltic movements which
he termed “ rollende Bewegungen.”

We should add here that the statement of van Braam Houkgeest
concerning the production of true Rollbewegung in the living ani-
mal by stimulation of the peripheral end of the vagus was, as far
as we know, never tested by any one; in fact, we did not come
across any reference to that statement, nor to the statement that
the post mortem Rollbewegungen depend upon the vagi being
intact.

The true Rollbewegungen were not seen by many students of
intestinal movements. Those who had seen them would not have
failed to dwell upon them, as they present a striking phenomenon.

Peristaltic Rush. 263

In the extensive studies of Bayliss and Starling? this form of in-
testinal movement was apparently not observed by them. Had they
observed this mode of peristalsis, they would not have failed to
comment upon it, since it illustrated, as we shall show later, their
law of intestinal contraction in a striking way. Starling? says
that ‘‘ the post mortem vermicular contraction described by Engel-
mann in the rabbit is probably merely an exaggerated wave just
described,” meaning pendular movements. He states there further
that Mall places this form of contraction in a class by itself which
he terms “ vermicular.”” Mall, however, applies the term vermicular
to the normal peristalsis as well as to the “irregular rapid wave.”
The latter is not designated by Mall by any special name. The term
Rollbewegung does not occur in Mall’s paper in which the descrip-
tion of Engelmann of the effect of mechanical stimulation of the
intestines serves as a basis for the analysis of this mode of intestinal
movement. Mall considers the rapid irregular wave as a patho-
logical phenomenon “to rid the intestine rapidly of irritating prod-
ucts of decomposition,’ having in mind the above recorded
statements of Bokai.

We came across the phenomenon of Rollbewegungen in our
studies of the effects of saline purgatives,3 ergot,* and of magnesium
salts. We have seen it occasionally occurring in the living animal
in the very same striking manner as was described by van Braam
Houkgeest in animals dying from asphyxia. In a series of experi-
ments especially devoted to that subject we have traced at first the
conditions under which Rollbewegungen accidentally occurred, and
then attempted to bring out the phenomenon at will. It is the object
of this paper to give a brief account of the results we have thus
obtained. However, before entering upon a detailed description
of our experiments we shall dwell upon the general appearance of
the phenomenon of Rollbewegung and its essential features, by
means of which it may be distinguished from other forms of in-
testinal movements.

Rollbewegung, or peristaltic rush. — In moderately distended and

1 Bayliss and STARLING: Journal of physiology, 1899, xxiv, p. 99, and Igoo-
IQOT, Xxvi, p. 125.

2 STARLING: SCHAEFER’S Textbook of physiology, ii, p. 329.

3 AUER: This journal, 1906-1907, xvii, p. 15.

* MELTZER and AvER: /ézd., p. 143.

5 MELTZER and AUER: Jid., p. 313.

264 S. F. Meltzer and F. Auer.

moderatively active small intestines, suddenly, frequently without
warning, a rushing wave appears, which sweeps with great rapidity
over the entire small intestine to stop only at the cecum. Each
coil, as the rushing wave passes through it, gives the appearance of
a rapidly turning wheel. On account of twisted and intricate
relations of the intestinal convolutions which do not permit the
simultaneous observation of the entire gut, the peristaltic rush pre-
sents a confusing spectacle of whirling coils appearing, disappearing,
and reappearing, until the entire rush comes to a standstill. The
entire rush is at times accomplished in less than fifteen seconds.
In the wave, as it hurries through each coil, two parts may be dis-
tinguished which present completely different aspects. In the aboral
part of the wave the moving intestine appears in the shape of a
turning wheel, is greatly distended, perfectly smooth, and offers
apparently not the slightest resistance to the onward movement
of its contents which is rapidly driven through it, and which con-
sists of a dark brown or yellowish fluid intermingled with some
gas bubbles. This part is apparently completely relaxed, all tonic
or rhythmic contractions are inhibited. Closely at the foot of this
section, at the oral end of it, the other part of the wave follows in
which the intestine is contracted to a cord. The lumen of that part
of the intestine is completely obliterated, but the constriction fails
to produce such complete anemia as often attends strong contrac-
tions of the intestines caused by artificial stimulations, by the
administration of barium, etc. Even at the height of the contrac-
tion the color is still pinkish. The contraction lasts only a few
seconds, after which that part of the intestine looks patulous and
empty: it retains, however, the rounded shape for some time. On
account of the very rapid propagation of the wave of constriction
the contracted piece of intestine appears sometimes to havé a length
of 5to1ocm. The preceding wave of inhibition extends apparently
over a considerable section of the intestine; but its aboral end is
usually lost in a hidden loop and cannot be ascertained.

A complete peristaltic rush begins somewhere in the duode-
num and terminates at the cecum. The exact starting-point of the
wave is difficult to establish on account of the deep location of the
duodenum. |

Such complete Rollbewegungen are sometimes followed immedi-
ately by an incomplete wave which begins at some place in the

Peristaltic Rush. 265

jejunum or ileum, and runs either a full course terminating at
the czecum or dies out, terminating at some distance from it.

As a rule, after such rushing waves, the small intestine remains
completely quiescent, and for some time neither normal peristalsis
nor pendular movements make their appearance.

Accordingly, the true and complete Rollbewegungen, as we have
seen them, are characterized (1) by the great rapidity of progress
of the circular constriction; (2) by the extensive and complete
inhibition preceding the contraction, and (3) by the complete course
of the wave, which traverses without interruption the entire small
intestine.

In many instances, however, Rollbewegungen appear which are
incomplete in one or the other of the characteristics mentioned. In
the first place many rushing waves appear which run only short
distances, beginning at the duodenum and terminating in some part
of: the jejunum or ileum, or beginning in a still lower section and
dying out before reaching the cecum. It is this kind of Rollbewe-
gungen which were seen by Nothnagel in the living animal. In the
second place, the progress of the wave may sometimes be of only
moderate rapidity. Such slower waves may exceptionally traverse
the entire intestine; asa rule, such slow waves die out in the middle
of the circuit. However, even these slower waves move incom-
parably more rapidly than the waves of normal peristalsis. Finally,
in some cases of Rollbewegungen the constriction as well as the
relaxation may not be so extreme as described above; but here
again both features of the rushing wave are even in such incomplete
cases much more pronounced than in normal peristalsis.

Incomplete forms of Rollbewegungen are easily distinguishable
from normal peristalsis by the size of the section which is involved,
by the rapidity of the progress of the wave, and by the intensity
of the processes, especially by the striking relaxation of the part of
the intestine in front of the progressing constriction.

For those who have seen the phenomenon of Rollbewegungen
it is hardly necessary to point out the particulars which distinguish
them from pendular movements. The two types of movement can
hardly be confounded, as they have practically very little in common.

It is, however, necessary to state expressly that the Rollbewe-
gungen should not be confused with simple, violent constrictions
of the smail intestines, such as are frequently seen after intravenous

266 S. F. Meltzer and f. Auer.

administration of barium chloride, of eserin, or even of ergot, and
in many other conditions. These constrictions may even be
stronger than those seen in Rollbewegungen, and may extend over
6 or 8 cm.; they might even show some travelling. But it is
not difficult to distinguish them from the Rollbewegung. The
constrictions in violent intestinal movements last a good deal
longer than those of the Rollbewegungen, occur simultaneously at
many sections of the small intestine; they do not travel progressively
in an aboral direction, but move irregularly to and fro in a slow
fashion. The striking differential point, however, is the absence in
these violent constrictions of any inhibition in front of a constricted
section. The contents of the intestine which is driven out by such
violent constrictions have to pass through a more or less tonically
contracted piece of intestine at the other end of which the way
is frequently blocked by another strongly’ contracted part of the
intestine.

We may say here that the above-mentioned “ rollende Bewe-
gungen,”’ observed by Bokai after the introduction of CO, and other
gases into the lumen of the intestines, were simply violent move-
ments of the type just mentioned, and do not belong to the true
Rollbewegungen. Bokai himself states positively that the inhibi-
tory factor of the intestines was not involved in the action of these
gases upon the intestines. The gases introduced into the lumen
of the intestines stimulate the gut by contact to violent constriction,
extending over a few centimetres, which constriction may extend
farther down if the gases travel downward.

A few words more with reference to the nomenclature of this
phenomenon. As stated above, the term Rollbewegungen was used
by van Braam Houkgeest in his protocols. For lack of a better
term he retained it also in the final publication of his studies. How-
ever, Rollbewegung indicates only the incidental feature of the
phenomenon, the rapidity of the wave through the coils calling forth
the illusion of a turning wheel. In the English literature there
is no special name for this phenomenon. Cannon and other writers,
use the German term Rollbewegungen. As a fitting English desig-
nation for this phenomenon we adopted the term Peristaltic Rush,
indicating in the first place its main feature, namely, the rushing
character of the forward movement. Furthermore, “ peristaltic ”
conveys the important fact that these movements possess the essen-

Peristaltic Rush. 267

tial characteristic of peristalsis, namely, a contraction above and
inhibition below (Law of Intestine of Bayliss and Starling). When
the rushing wave runs through the entire small gut, we designate
it as complete; when the wave runs through only a part of the
course, or is deficient in other ways, we designate it as incomplete.

EXPERIMENTAL OBSERVATIONS.

Method. — The observations were made exclusively on rabbits,
some of which received subcutaneous injections of morphin. The
intestines were invariably observed in a warm saline bath. The
saline consisted of a solution of 0.92 per cent sodium chloride, which
is considered isotonic with the serum of the rabbit. (The solutions
employed by van Braam Houkgeest consisted of 0.6 per cent sodium
chloride.) In the present series of experiments the “ receptacle
for the bath” was prepared by flaps from the abdominal skin in
the manner described in our paper on the action of ergot.’ After
an incision in the middle line of the abdomen the skin was extensively
dissected on both sides from the underlying musculature, and by
an appropriate suspension a deep receptacle was formed. This was
filled up with a warm saline solution, and under cover of this the
abdomen was freely opened in the linea alba. By slightly retracting
the muscular walls by light weights the escaping intestines were
bathed on all sides with the warm saline. The solutions were kept
warm and at the proper level by frequent addition of fresh warm
saline.

Movements in normal rabbits. At the outset we may state that
in normal living animals we have never seen movements of the small
intestines which we could designate as rushing peristalsis according
to our definition. In the present series the behavior of the intestines
was watched for some time before the effect of any substance was
tested. It must be admitted that the time given to such preliminary
observations was necessarily not very long. But we have had suff-
cient opportunities in various series of experiments, carried out for
other purposes, to watch the behavior of the uninfluenced intestines.
Never have we noticed intestinal movements belonging to the rush-
ing type.

After intestinal stimulants. — Neither have we seen rushing peri-

! MELTZER and AUER: This journal, 1906-1907, xvii, p. 143.

268 S. F. Meltzer and F. Auer.

stalsis in experiments in which the intestines were stimulated to
greater activity by subcutaneous or intravenous injections of some
purgatives. In the experiments of one of us (A.) with subcutaneous
and intravenous injections of sodium sulphate, sodium phosphate,
and sodium citrate in which it was found, in agreement with the
statement of J. B. MacCallum,’ that the movements of the small
intestine are increased, no peristaltic rush ever occurred, although
the intestines were watched for hours. In experiments with intra-
venous injections of barium chloride and of eserin, in which the
intestines were stimulated to violent and extensive contractions,
no peristaltic movement was ever observed which bore the criteria
of peristaltic rush as set forth above. The same we may state with
regard to the effects of ergot, when employed alone. The intestines
were stimulated to rhythmic, travelling, and tonic contractions, but
none of them possessed the characteristics of rushing peristalsis ;
especially was the marked inhibitory wave absent from all the motor
phenomena produced or aggravated by the injections of ergot or
any of the other intestinal stimulants.

Stimulating and inhibitory factors. — The phenomenon of peri-
staltic rush we have seen to occur in such experiments in which
apparently two opposing elements were in operation, — factors which
increase intestinal activity and factors which as a rule inhibit this
activity. Of the elements of the first class there were employed
sodium phosphate, sodium sulphate, sodium citrate (“ saline purga-
tives’’), ergot, barium chloride, eserin, and destruction of some
part of the dorsal cord. As inhibitory agents we have employed:
calcium chloride, magnesium chloride, and magnesium sulphate.
The greatest number of experiments were made with ergot and
calcium, which gave, as we shall see later, very reliable results. We
have obtained, however, satisfactory results also in other combina-
tions. We shall illustrate our results only by a few greatly abbrevi-
ated protocols.

“‘ Saline purgatives ’’ and calcium. —
Experiment 1. — Gray female rabbit, 1750 gm... .
3:27 P.M. Abdomen opened in saline bath. ... No movements of
small intestines.
3.37 P.M. Injected subcutaneously 15 c.c. sodium phosphate (4 per
Cent). 9s ae

1 J. B. MACCALLUM: This journal, 1904, x, p. 107.

Peristaltic Rush. 269

3-47 P.M. Slight contraction of duodenum, balance of small gut empty
and quiets-.-.

3-52 P.M. Duodenum full of light yellow fluid, quite active ; jejunum
and ileum, still empty, show swaying movements. . . .

During next eighty minutes only slight changes.

5-10 P.M. Upper part of jejunum full, but not distended, shows con-
stantly swaying movements. Duodenum shows only moderate swaying. .. .

5.17 P.M. Injected through the external jugular vein 2 c.c. sodium
sulphate #/8 solution, followed by an injection of 1 c.c. saline. Small
intestines show soon after a moderate increase of swaying motions.

5:23 P.M. Movements definitely less again.

5.30 P.M. Injected 2 cc. sodium citrate /8 solution, followed by
1 c.c. saline. Soon after the injection the movement of small intestines
definitely increased, swaying and shortening movements, no constriction
waves seen.

5:35 P.M. Intravenous injection of 2 c.c. CaCl, m/8 solution, fol-
lowed by 1 c.c. saline.

Before injection was finished jejunum and ileum became quiet, but a
part of the duodenum showed very active pendular movements and strong
circular constrictions such as were not seen before.

5:40 P.M. Small intestines show again good swaying movements. .

5:50 P.M. Intravenous injection of 3 c.c. CaCl, m/8, followed by 1 c.c.
saline. . . . No marked change in either direction.

6.00 P.M. Intravenous injection of 1 c.c. CaCl, m/1 solution, followed
by 1 c.c. saline. Shortly after, a powerful contraction of the duodenum
set in, shooting the fluid contents swiftly through the coils of the small
intestines and stopping at the cecum —a complete peristaltic rush.

Repeated twice with 1 c.c. CaCl, m/1, each time with the same result.

The above experiment, the protocol of which was here greatly
abbreviated, is instructive in many directions. For two hours the
intestines were watched while they were under the influence of sub-
cutaneous and intravenous injection of the “ purgative salts.” As
a result of these injections the small intestines showed a moderate
increase of their motility which was confined to the swaying motions.
At no time was there an indication of a rushing peristalsis. After
the injection of 2 c.c. CaCl, m/8, which was equal to the preceding dose
of sodium citrate, the activity of the jejunum and ileum became in-
hibited, which is in harmony with the statement of J. B. MacCallum !
that calcium counteracts the stimulating effect of the purgative salts.

1 MAcCAL.uM, J. B.: This journal, 1904. x, p. 107.

270 S. Ff. Meltzer and Ff. Auer.

The duodenum, however, became more active. After a few minutes
the activity of the small intestines returned, and a second injection
of a similar dose of CaCl, had no decided effect. Finally, when a
dose of I c.c. of a molecular solution of calcium chloride was ad-
ministered, which had to be injected very slowly, a powerful wave
of rushing peristalsis swept over the entire small intestine. Repeat-
ing the injections with similar doses at proper intervals brought out
similar results.

While we can confirm in general the discovery of MacCallum re-
garding the inhibitory effect of the calcium salts upon intestinal
movements, we found at the same time that it is just the addition of
the calcium salts which brings out the rushing peristalsis.

Similar results were obtained in some other experiments in which
injections of calcium salts followed those of purgative salts.

However, we have not made many experiments with the purga-
tive salts. Our main experiments were made, as stated before, with
ergot and calcium, and we shall quote two abbreviated protocols to
illustrate the various results obtained with this combination.

Calcium and ergot. —
Experiment 2.—Black female rabbit, 2030 gm... . / Abdomen opened.

. . Recovering from ether.

11.15 A.M. No movements of gut. Intravenous injection of 1 c.c.
CaCle m/1, followed by 2 c.c. saline.

11.23 A.M. Intestines relaxed, flat, no movements anywhere. .. .

11.35 A.M. Intravenous injection of 1 c.c. CaCl: m/1, followed by
27C.c, Saline.

11.42 A.M. Lower small gut shows slight motions, coils flat... .

11.54 A.M. Occasional good swaying of some loops of the small
intestine.

11.55 A.M. Intravenous injection of 1 c.c. of fluid extract of ergot
(Squibb), followed by 2 c.c. saline. Shortly after, movements of small gut
definitely increased. .. .

11.58 A.M. Strong contraction, a few centimetres long, drives contents
swiftly into cecum — complete peristaltic rush.

12.11 P.M. Intravenous injection of 1 c.c. ergot, followed by 2 c.c.
saline.

12.13 P.M. Slight increase of swaying of small gut.

12.24 P.M. Swaying becomes once in a while more marked, but no
sign of Rollbewegung.

12.26 P.M. Intravenous injection of 1 cc. CaCl m/1, followed by
2 c.c. saline.

Peristaltic Rush. 271

12.27 P.M. Strong “Rollbewegung,” driving contents into caecum
(stronger than before).

12.30 P.M. Small gut relaxed, but not empty.

12.40 P.M. Intravenous injection of 1 c.c. ergot, followed by 2 c.c.
saline.

12.50 P.M. Pendular movements increasing, gut gradually filling up,
no Rollbewegung.

12.52 P.M. 1 cc. CaCl m/1, followed by 2 c.c. saline. Swaying
movements subsided for a while and then started again.

12.55 P.M. A good wave of rushing peristalsis... .

Animal killed by asphyxia. After convulsions subsided “ good travelling
peristalsis of small gut.” .

In this experiment the first injections of CaCl, in molecular solu-
tion produced no effect, there were previously no movements to be
inhibited; the additional injection of I c.c. of ergot brought on
within three minutes a wave of rushing peristalsis. A further in-
jection of ergot brought out only swaying movements, but now an
injection of CaCl, brought out within two minutes strong Roll-
bewegungen. Further injections of I c.c. ergot and 1 c.c. calcium
brought on another peristaltic rush.

Calcium chloride alone had no effect at all (when there were no
previous movements), and ergot alone brought only an aggravation
of the usual intestinal movements; but when ergot followed calcium
and when calcium followed ergot, the result was usually a complete
peristaltic rush.

Experiment 3. — Female rabbit, 1530 gm... .

11.35 A.M. Abdomen open, all operations finished. No movements
of intestines visible.

11.50 A.M. Occasional slight movements of small intestine.

11.54 A.M. Intravenous injection of 1 c.c. CaCle m/1, followed by
2 c.c. saline.

11.57 A.M. Moderate, but distinct, pendular movements of all visible
coils of small gut.

12.01 P. M.. Movements improved; some coils filled up, are round ;
duodenum invisible. .

12.11 P. M. Intravenous injection of 1 c.c. CaCl, m/1, followed by
2 c.c. saline. Before injection of the calcium was finished all movements
disappeared.

1215 P.M. Slight movements visible again.

12.20 P.M. Movements improved, present in nearly all visible coils.

a72 S. Ff. Meltzer and Ff. Auer.

12.28 p.M. Intravenous injection of 1 c.c. of ergot, followed by 2 c.c.
saline. Before the injection of saline is finished a moderate incomplete
peristaltic rush appeared.

. 12.32 P.M. Pendular movements in all coils.

12.34 P.M. Nearly all coils are perfectly quiet.

12.44 P.M. Intravenous injection of 1 c.c. of ergot, followed by 2 c.c.
saline.

12.46 P.M. Pendular movements in all coils, gradually increasing.

12.58 p.M. No Rollbewegung occurred. Intravenous injection of
CaClz m/1 was given again, followed by 2 c.c. saline. Before injection
of calcium was finished, az incomplete but strong peristaltic rush set in.

1.02 P.M. All coils very active.

1.12 P.M. Complete peristaltic rush.

1.27 P.M. Slight pendular movements in some coils.

1.28 p.M. Intravenous injection of 1 c.c. of ergot, followed by 2 c.c.
saline.

1.30 P.M. Pendular movements increased everywhere.

1.38 and 1.44 P.M. Jucomplete but good waves of rushing peristalsis
reach the cecum. Loops remain full, round, and show good swaying
movements.

1.45 P.M. Intravenous injection of 1 c.c. of CaCl, m/1, followed by
2 c.c. saline.

1.47PM. Entire gut quiet.

A few more alternating injections of calcium and ergot brought out only
increased normal activities and their inhibition, but no waves of rushing
peristalsis.

In this experiment the various alternating injections of calcium
and ergot brought out only one complete peristaltic rush and a few
incomplete rushes.

In many of the experiments upon normal animals, in which ergot
and calcium were used, more than one complete peristaltic rush
occurred, besides a few incomplete waves. Frequently, however,
after one or two rushes occurred, further injections of these sub-
stances became less effective and an incomplete Rollbewegung was
the most that could be obtained. We had no experiment in which
the injection of calcium and ergot failed completely to cause rush-
ing peristalsis, and we had only one experiment in which only in-
complete Rollbewegung occurred. In this experiment morphine
alone was used as an anesthetic, the respiration became very slow,
and ergot failed otherwise to cause the customary stimulating effect
upon the intestinal movements.

Peristaltic Rush. 273

Destruction of cord and administration of calcium. — In the follow-
ing instructive experiment the administration of calcium alone
brought out a complete peristaltic rush.

Experiment 4.— White female rabbit, 1560 gm. Morphine subcutaneous
0.015. . . . Spinal cord destroyed below fifth dorsal vertebra. .
Abdomen opened.

11.45 A.M. Good swaying motions and constrictions all over the small
gut.

12.00 M._ Intravenous injection of 1 c.c. of CaCle m/1. Before injec-
tion was finished, a strong wave swept over entire small intestine (duodenum
not visible) and drove contents into cecum. Complete peristaltic rush.
After this wave passed, gut became quiet.

12.06 P.M. Another wave swept down, stronger than before. Small
gut, after wave, empty and tape-like, moderately relaxed and shows some
swaying.

Three more injections of CaCl did not bring out any rushing
peristalsis.

Here the first injection of calcium brought out two waves of
rushing peristalsis. The destruction of the cord, which usually
greatly increases the activity of the intestines, supplied the stimulat-
ing factor; the intestines were very active while the calcium injec-
tion was given.

Magnesium salts and ergot. — Rushing peristalsis was rarely
brought about when in addition to ergot (and destruction of the
cord) magnesium sulphate or chloride was injected instead of
calcium. Out of five experiments only in one two incomplete
Rollbewegungen occurred. As we have shown elsewhere,! mag-
nesium salts ighibit completely the movements of the intestines pro-
duced by the injection of ergot. In the present line of experiments
the effect of ergot is apparently nearly completely lost as a stimu-
lating factor in the presence of the strongly inhibitory effect of
magnesium.

Calcium chloride and barium. — On the other hand, the inhibitory
effect of calcium is apparently a less reliable factor for the produc-
tion of peristaltic rush in the presence of such strongly stimulating
substances as BaCl. In five experiments in which calcium chloride
and barium chloride were injected intravenously, in three no peri-

1 MELTZER and AUER: This journal, 1906-1907, xvii, p. 318.

274 S. Ff. Meltzer and f. Auer.

stalsis of a rushing type made its appearance. In the two other
experiments one complete peristaltic rush occurred in each one,
besides one or two incomplete waves.

Weare here reminded of the statement of MacCallum! that “ the
peristaltic movements produced by barium chloride are usually not
stopped by the administration of calcium.”

Magnesium and barium. — Of four experiments in which barium
and magnesium chloride were alternatingly injected, in three there
were a few complete peristaltic rushes as well as incomplete waves,
and only in one all signs of Rollbewegungen were missed. In our
previous experiments ? we reported that magnesium salts are capable

f inhibiting the violent intestinal constrictions produced by barium.
While the inhibiting effect of magnesium is apparently much
greater than that of calcium, and is strong enough to overpower
temporarily the violent constrictions produced by barium, the stimu-
lating effect of the latter is, however, too strong to be completely
annihilated, even by magnesium. The result of the alternating in-
jection of the two strong antagonistic factors is therefore often a
compromise in the shape of a peristaltic rush.

In the following experiment calcium as well as magnesium was
employed : ;

Lixperiment 5.—Gray female rabbit, 1320 gm. Morphine oor... .
Abdomen opened at 11.50.

11.55 A.M. No sign of motion anywhere. Intravenous injection of 8
c.c. CaCle m/8 followed by 2 c.c. saline.

12.00 M. No effect.

12.05 P.M. Injection of 0.5 c.c. BaCl: m/8. Strong contractions of
small gut, no “running.”

12.15 P.M. 8c.c. CaCle m/8, later 0.5 c.c. BaCl, #/8.

12.20P.M. Again 0.3 c.c. BaCl: m/8. ‘‘No runs or anything ap-
proaching them.”

12.30 P.M. 4 c.c. CaCl, m/8 and 2 c.c. saline. Small gut quieter at
first, then attempt at running, but no definite wave of contraction swept
along.

12.50 P.M. 4 ¢.c. CaCl: m/8 and 2 c.c. saline. Soon after 0.1 c.c.
BaCle m/8. Same as before: tonic contractions of some parts of small
gut. ;

1.40 P.M. 0.9 c.c. MgCle m/1 and 0.1 BaCl, m/1. Shortly after a
definite run occurred over entire small gut. Later two short runs occurred.

1 MacCatuium: This journal, 1904, X, Pp. 107.
2 MELTZER and AUER: J/id., 1906-1907, xvii, p. 318.

Peristaltic Rush. 275

The several alternating injections of calcium and barium brought
no success, while the first injection of MgCl, brought on a true
Rollbewegung.

Eserin with calcium or magnesium. — We also made a few experi-
ments with eserin. In the experiments in which calcium alone was
used with eserin, there was practically no success. In one experi-
ment, however, in which at first eserin was alternated with MgSO,
no running waves occurred. Later, however, when calcium was
substituted for magnesium, the injections of calcium at first quieted
the contractions, but a few minutes after each injection a definite
complete peristaltic rush set in.

Destruction of cord. — The cord was destroyed in seven experi-
ments (below the 5th, 3d, or 2d dorsal vertebra). In four experi-
ments the abdomen was opened soon after the destruction and
watched only for fifteen to twenty minutes before an injection of
any kind was given. The intestinal activity was increased in all
four experiments, but no rushing peristalsis was seen. In three
experiments the abdomen was opened about two hours after the
destruction of the cord. In two of these experiments there were
a few spontaneous peristaltic rushes, complete and incomplete. In
the third experiment there was only one incomplete peristaltic rush,
but the intestines in this experiment have also otherwise shown very
little activity.

From these few experiments we learn at least in a general way
that the phenomenon of peristaltic rush may occur in rabbits whose
cord was so destroyed as to eliminate the influence of the splanchnics
and who otherwise did not receive any substance capable of inhibit-
ing intestinal movements.

Section of vagi. — According to van Braam Houkgeest, as will be
remembered, the post mortem Rollbewegungen did not occur when
both vagi were previously cut. We have made four experiments in
which CaCl, and ergot were given alternately after both vagi were
cut. In none of these cases did a complete Rollbewegung occur.
In three experiments there were some incomplete Rollbewegungen,
in two of which the course of these waves was short and sluggish
and marked by a very slow relaxation of the contracted part. In
one of these experiments the vagi were cut after the animal had
already received a few doses of ergot and calcium, but only one
good but incomplete Rollbewegung was produced. After cutting
the vagi and continuation of the injections, one other incomplete

276 S. Ff. Meltzer and Ff. Auer.

Rollbewegung occurred which was as good as the one before cutting
the vagi. In this case after killing the animal by asphyxia an in-
complete Rollbewegung occurred exactly like those observed while
the animal was alive.

These few observations seem indeed to justify the assumption
that the occurrence of a true complete peristaltic rush is to a great
measure dependent upon the integrity of the vagi.

We may call to mind here that cutting the vagi interferes greatly,
for some time at least, with the normal movements of the stomach
and also, as we have shown recently,! with the normal movements
of the rabbit’s caecum.

Stimulation of the vagii— Van Braam Houkgeest has also stated
that stimulation of the peripheral end of one vagus will also cause
in the living rabbit Rollbewegungen, provided both splanchnics are
previously cut. We have tested this claim in two rabbits whose
splanchnics were cut, and in three others in which the cord was
destroyed, which was equivalent to cutting the splanchnics. In no
case could we find that stimulation of the vagi brings out complete
or incomplete peristaltic rushes. Even in an experiment in which,
after destruction of the cord, rushing peristalsis occurred spon-
taneously and with readiness, stimulations of the peripheral end of
the vagi, although visibly aggravating the other motor activities of
the intestines, did not contribute to the production of a peristaltic
rush.

Our experiments have shown that the phenomenon of peristaltic
rush occurs when the animal receives injections of two groups of
substances. The first group comprises sodium sulphate, sodium
phosphate, sodium citrate, ergot, barium chloride, and eserin. The
second group comprises calcium chloride, magnesium chloride, or
magnesium sulphate. The substances of the first group are intes-
tinal stimulants; that is, by their injection the intestines are stimu-
lated to greater activity. The second group we consider as in-
hibitory substances for the intestinal movements; that is, we assume
that by their injection intestinal movements, when present, are
reduced or completely inhibited for some time. The stimulating
character of the first group of substances is well understood and
requires little discussion. Among this group the sodium salts are
the weakest stimulants, ergot acts much more strongly, and barium

1 Proceedings of the Society for Experimental Biology and Medicine, 1907,
iv, p. 37.

Peristaltic Rush. 277

and eserin have the strongest effect. With regard to the inhibitory
group a few explanatory remarks would not be out of place. The
inhibitory effect of calcium upon the intestinal movements was ob-
served by J. B. MacCallum. These movements, which were started
or aggravated by the injection of “ purgative salts’ were inhibited
by the injection of calcium chloride. This discovery was made on
the basis of J. Loeb’s well-known view of the general inhibitory
effect of calcium salts. The strong effect of barium chloride could
not be inhibited by the injection of calcium. From the observations
in the present series of experiments as well as on many other occa-
sions, we can confirm the statement that the injection of calcium
chloride causes as a rule an inhibition of intestinal movements when
normally present or brought on by the injection of stimulating
agents.

Regarding the inhibitory effect of magnesium salts upon intes-
tinal peristalsis, we have dealt recently in a special article on that
subject.1 We have found that these salts are capable of inhibiting
intestinal movements of whatever source. We may state again ex-
pressly that the violent movements of the intestines caused by
bariym or eserin can also be inhibited by the injection of magnesium
salts.

We may mention here that MacCallum also observed that intra-
venous injection of magnesium chloride inhibits the movements of
the intestines produced by sodium citrate, sulphate, etc., although
he surprisingly stated that subcutaneous injection of magnesium
sulphate has a stimulating effect upon the intestines. In our ex-
perience we found no difference between magnesium sulphate and
magnesium chloride; both inhibit the intestinal movements.

We may also state expressly that according to our experience the
inhibition exerted by magnesium salts is distinctly stronger than
that produced by calcium; the contractions are more completely in-
hibited, the effect lasts longer, and we met with no kind of move-
ments which could not be reduced or abolished by magnesium, while
calcium, according to MacCallum, cannot overcome the effect of
barium chloride.

As to the meaning of inhibition we may refer to our first paper
on the magnesium salts.2, We started from the hypothesis that
magnesium favors such an action as that of the vagus nerve. For

1 MELTZER and AUER: This journal, 1906-1907, xvii, p. 313.
2 This journal, Joc. czt.

278 S. Ff. Meltzer and f. Auer.

the intestines we may say that the effect of magnesium is similar
to the well-known inhibitory action of the splanchnics.}

Are our present results in harmony with this hypothesis? Did
we not find that by the injection of calcium or even magnesium a
most remarkable intestinal movement takes place? We must admit
that any investigator who would come across the phenomenon of
peristaltic rush occurring after an injection of calcium or mag-
nesium chloride without having much experience with these salts,
might be inclined indeed to insist that calcium and magnesium are
stimulating agents for intestinal movements.?

We shall, however, call attention to the following facts. If in-
jections of magnesium or of calcium are given in large or small
doses to an animal which previously received no stimulating salts
and whose cord was not destroyed or whose splanchnics were not
cut, no movement of the intestines ever follows those injections.
For magnesium we may state that this is the absolute rule, to which
apparently there is no exception. For calcium there is once in a
while an exception; we have occasionally seen after an injection of
calcium chloride a constriction appearing in some part of the gut,
but this was a rare occurrence, and the effect was circumscribed and
very brief. The general rule is that calcium produces no intestinal
contractions. Furthermore, if there have been slight spontaneous
movements of the intestines, or slight movements brought on by
the injection of some intestinal stimulants, an injection of mag-
nesium or calcium will invariably stop these movements for a
shorter or longer period. Finally, even when the intestines show
strong activity, “spontaneous’’ as well as those brought on by
artificial means, in the great majority of the cases an injection of

1 See also A. G. MAYER: Rhythmical pulsation in Scyphomedusa, Carnegie
Institution publications, 1906.

2 The situation with which Loeb was confronted in his studies upon the
hydromedusa Polyorchis (The stimulating and inhibitory effects of magnesium,
etc., Journal of biological chemistry, 1905-1906. i, p. 427) is of a similar misleading
character. When to a solution of NaCl in which the medusa does not show any
movements, magnesium chloride is added, the characteristic swimming movements
soon appear. This conveys the impression that magnesium acts as a stimulating
agent. The probable interpretation, however, is that the sodium chloride solution
keeps the muscle in a state of contraction, a systolic state, and that the addition of
magnesium causes a relaxation of the tonus, thereby favoring the reappearance of
rhythmic diastoles. In favor of that view LoEB mentions the fact that the mouth
and tentacles are permanently contracted in any solutions without magnesium.

Peristaltic Rush. 279

magnesium or calcium will cause at least a preliminary inhibition
of the intestinal movements.

We therefore assume that the injection of magnesium or calcium
introduces an inhibitory factor, and that the appearance of the
phenomenon of peristaltic rush occurs only as a compromise be-
tween two opposing factors, the stimulating and inhibitory elements.
When we stated above that we were enabled to produce at will
the occurrence of peristaltic rush in living rabbits, we did not
mean to claim that certain injections will invariably bring out the
phenomenon. We claim only that we are now in a position to create
a situation in which the phenomenon in all probability is likely to
occur. The peristaltic rush by no means promptly follows each
injection. On the contrary, in a prolonged experiment in which
many alternating injections were given, it frequently happened that
only one or two complete waves of rushing peristalsis made their
appearance.

The phenomenon of peristaltic rush consists, as we have analyzed
above, of two parts, — of a stimulating part, in which the circular
constriction is strong and its propagation rapid, and of an inhibit-
ing part, in which a long section of the intestine in front of the
rushing wave of constriction is completely relaxed, so as to offer
no obstacle to the swiftly driven contents. The peristaltic rush is
in its composition very similar to the normal peristalsis of the
cesophagus in which an inhibitory wave runs rapidly ahead of the
contraction to clear the path of all obstructing constrictions. By
introducing at the same time into the body stimulating and inhibit-
ing agents, conditions are created which permit the occurrence of
such specific combinations of stimulation and inhibition as to start
off the wave of peristaltic rush.

The antagonistic factors, to be favorably combined, must be
mated in proper proportions. The stimulating effect of ergot,
which is not very strong, is best combined with calcium, the in-
hibitory effect of which is also not too strong. The inhibitory effect
of magnesium is too strong and is apt to completely overpower the
stimulation of ergot. On the other hand, the strong stimulating
effect of barium is better paired with magnesium than with calcium
in order to bring out peristaltic rush.

The increased activity of the intestines after release from the
inhibitory grip of the splanchnics (1. e., after their section) is appar-
ently just of the right proportion to enter into a satisfactory com-

280 S. Ff. Meltzer and fF. Auer.

bination with the inhibitory effect of calcium. Hence the occur-
rence of peristaltic rush after an injection of calcium when the
splanchnics were previously cut.

The occurrence of peristaltic rush after destruction of the dorsal
cord would seem to require some explanation. The character of
the compromise is here not very evident, since the destruction of
the cord rather removes an inhibitory factor. More facts will have
to be collected before we could discuss this point satisfactorily.
But we may say that in our experiments on the rabbit’s caecum !
we have established the fact that simple opening of the abdomen
even in a warm saline bath is an inhibitory stimulus of a local char-
acter for the cecum. It is probably an inhibitory stimulus also for
the small intestine.

Finally, we have to recall here our observation according to which
the vagi are apparently controlling factors in the management of
peristaltic rush. In the absence of their influence no complete peri-
staltic rush takes place. Incomplete, short, sluggish runs do occur
even after the vagi are cut, and they may be of peripheral origin,
either myogenic or neurogenic. But a complete true peristaltic rush
seems to require the assistance of the central nervous system con-
veyed through the vagi.

Peristaltic rush is probably not an infrequent occurrence in vari-
ous pathological conditions, and is probably also an essential factor
in purgation. As to normal conditions, we have stated above that
we have never seen Rollbewegungen in the opened abdomen of a
normal animal. But an opened abdomen is not a normal state. We
have shown that the clearly visible movements of the stomach ? and
of the cecum? completely disappear after opening of the abdomen.
We will, however, not enter here into a discussion of that subject.

' RESUME.

Peristaltic rush (Roflbewegungen) consists of a rapidly pro-
gressing wave of contraction preceded by a completely relaxed long
section of the intestine through which fluid contents mixed with
gas bubbles is rapidly driven. A complete peristaltic rush is one

1 MELTZER and AvER: Zentralblatt fiir Physiologie, 1907, xxi, p. 71.

2 AUER: This journal, 1906-1997, xvii, p. 15.

3 MELTZER and Aver: Proceedings of the Society of Experimental Biology
and Medicine, 1907, iv, p. 37.

Peristaltic Rush. 281

which sweeps down from the duodenum to the cecum without
stopping.

Peristaltic rush was seen to occur in living animals with opened
abdomen when intravenous injections of stimulating and inhibitory
substances were given. As stimulating substances were used some
purgative salts, ergot, barium chloride, and eserin; and as inhibitory
substances, calcium chloride, magnesium chloride, and magnesium
sulphate were used. The best success was obtained by ergot and
calcium chloride.

Cutting the vagi prevents the occurrence of complete peristaltic
rush.

As a result of a general character we may consider the fact that
the simultaneous administration of stimulating and inhibitory sub-
stances did not lead to a mutual neutralization, but rather to a con-
dition in which both effects are manifest and are combined in such
a co-ordination as to favor effective motion, namely, an increase
of the motor factors and an inhibition of the antagonists, —a
condition which was designated by one of us (M.) as contrary
innervation.

Sie AC? CONTROL: OF - THE: PYLORUS.

BY Web. CANNON.
[From the Laboratory of Physiology in the Harvard Medical School]

CONTENTS.

Discussion of previous views as to the control of gastric evacuation
Mechanical agencies in the stomach .
Mechanical agencies in the intestine .
Chemical agencies in the stomach .
Chemical agencies in the duodenum .
The two factors concerned in gastric evacuation .
The facts to be explained , a
A theory of the control of he ee ele 9
Experimental evidence for the acid control of the pyleres oa
A. That acid in the stomach opens the pylorus
1. Delaying the appearance of HCl delays the initial aisgharee
2. Hastening the appearance of HCl hastens the initial discharge .
3. The appearance of acid in the antrum closely precedes the initial dis-
charge Soe
4. HCl opens the py jagns A: ‘Hie extised sGirmach
B. That acid in the duodenum keeps the pylorus closed
1. Acid in the duodenum inhibits gastric peat (Hirsch, Serdjukow.
cobler).. =. ter

2. Absence of the Seal alkaline secretions ‘froin the disdenar founda
gastric discharge. . . ys :
3. Destroying continuity hetwees domaak aud dnodeninn hastens mae
discharge . Wye Se ey
Evidence for the acid control deve a Hox previous phiehvdades on eee aigeeaoe
0S 5S er) eens fo Sees Pee ree ener ye
Carbohydrates.

Combinations of the od rues. :
Observations not accordant with the acid control .

The discharge of water

The discharge of egg-albumin -

Miteguseharse of fats. 1. - . + «

Pathological cases : :
Inferential support for the acid coagral ae sehen processes in fe naliiie region .
IG ee

Page
284
285
285
286
287
288
289
290
292
292
292,
295

296
300
301

302

302

304
306
306
309
310
311
312
313
315
316
317
319

HREE years ago I called attention to the fact that when the
different food-stuffs, uniform in amount and consistency, are

fed, they are discharged through the pylorus at different rates.

At

that time I offered in explanation of this differential discharge a

283

284 W. B. Cannon.

theory? which accorded with the facts then established. Later I
reported further observations on the discharge of the three food-
stuffs, singly and in combinations, and stated that a discussion of
these observations and an account of explanatory experiments would
be deferred to a later paper. They are presented in the following
pages.*

Opposing views have been set forth as to the manner in which
the stomach empties. The statements of Richet * and of Rossbach,®
that gastric contents are first thoroughly mixed with gastric juice
and after three or four hours or more are passed rapidly into the
duodenum, have given way before contrary results obtained under
more natural conditions. Clinical studies with the stomach tube,°
observations on the undisturbed subject by means of the X-rays,’
and investigations through duodenal fistulz,* combine to prove that
the stomach is emptied progressively during the course of gastric
digestion. The observations by means of X-rays and the investiga-
tions through duodenal openings have further demonstrated that
the chyme does not pass through the pylorus in a continuous small
stream, nor at the approach of every peristaltic wave, but occurs
occasionally, at irregular intervals.

DISCUSSION OF PREVIOUS VIEWS AS TO THE CONTROL OF
GASTRIC EVACUATION.

Both mechanical and chemical agencies have been invoked to
explain the emptying of the stomach. These agencies have been
supposed by some investigators to act in the stomach, by others to
act in the intestine. It is necessary, first, to consider these views
and the evidence adduced in their favor.

1 CANNON: This journal, 1904, x, p. xviii.

2 CANNON: /did., xii, p. 418.

8 Some of this evidence has been reported in previous communications (see
CANNON: Journal of the American Medical Association, 1905, xliv, p. 15; This
journal, 1906, xv, p. xxv).

4 RICHET: Comptes rendus, Académie des Sciences (Paris), 1877, Ixxxiv, p. 451.

® RossBACH: Deutsches Archiv fiir klinische Medicin, 1890, xlvi, pp. 296, 317,

6 Ewacp and Boas: Archiv fiir pathologische Anatomie und Physiologie und
fiir klinische Medicin, 1885, ci, p. 365.

7 CANNON: This journal, 1898, i, pp. 368, 369, 377.

8 ScHIFF, Physiologie de la digestion, Florence and Turin, 1867, ii, p. 326.
KiHNE: Lehrbuch der physiologische Chemie, Leipzig, 1868, p. 53. Also v.
MERING: Verhandlungen des Congresses fiir innere Medicin, 1897, xv, p. 433-

et et oe

The Acid Control of the Pylorus. 285

Mechanical agencies in the stomach. — The claim has been made
by those who believe that chyme is discharged only after several
hours of gastric digestion, that the pyloric sphincter, although able
to withstand the recurrent peristaltic pressure in the earlier stages
of chymification, is overcome by the more intense constrictions in
the later stages.1 As already stated, the proof is conclusive that a
delay of several hours in the discharge from the stomach is ab-
normal. The moving constriction-rings do indeed press deeper into
the gastric contents as digestion proceeds, but this late augmenta-
tion of contraction does not explain the normal gradual exit during
earlier stages of chymification. In these earlier stages I have many
times looked carefully for any relation between the moment of dis-
charge and any momentarily greater intensity of peristalsis. Wave
after wave passes with almost no perceptible variation of depth.
Yet, as the waves are passing with such notable uniformity, the
pylorus may open before the pressure of an approaching constric-
tion, and the mass in the antrum, then released, will be driven forth
into the duodenum. The next wave, and perhaps many thereafter,
of approximately the same depth, may fail to press the food on-
ward.” The occasional discharge of chyme from the stomach cannot
therefore be attributed to an occasional increase of intensity of the
peristaltic constrictions.

Mechanical agencies in the intestine. —In 1897 v. Mering® re-
ported that the introduction of a large amount of milk into a duo-
denal fistula checked the exit of water from the stomach. The
next year Marbaix* published a paper on evacuation of the stomach
as affected by a state of repletion of various parts of the intestine.
He found that in the upper half of the small intestine a state of
repletion, induced by injections through fistulz, inhibited the dis-
charge from the stomach.® In order to cause the reflex, however,
even in the first fourth of the intestine, the injected liquid had to

1 See LessHAFT: Archiv ftir pathologische Anatomie und Physiologie und fur
klinische Medicin, 1882, Ixxxvii, p. 80.

2 Cannon: This journal, 1898, i, p. 369.

8 vy. MERING: Loc. cit., p. 434.

4 MARBAIX: La cellule, 1898, xiv, p. 251.

5 An investigation of the motor functions of the stomach after pyloroplasty
(see CANNON and BLAKE: Annals of surgery, 1905, xli, p. 707) has proved that
although the upper part of the small intestine may become filled with food, there is
no cessation of peristalsis. The effect noted by v. MERING and MARBAIX would
therefore probably be due to closure of the pyloric sphincter.

286 W. B. Cannon.

occupy a considerable extent of gut. For example, filling the gut
from I0 cm. to 25 cm. beyond the pylorus caused no inhibition of
the discharge. But much less than 15 cm. of continuous content is
normally present in the upper intestinal tract. A radiograph and
tracings of X-ray shadows already published* show that the in-
testinal contents are normally disposed in separate short masses.
Under natural conditions, therefore, the extensive uninterrupted -
surface of contact required by v. Mering’s and Marbaix’s explana-
tion, in order to prevent a continuous outpouring from the stomach,
does not exist. As the continuous outpouring nevertheless does not
occur, their results do not explain the normal control of gastric
discharge. .

Von Mering’s and Marbaix’s results are confirmed by Tobler’s
observation * that the rapid inflation of a balloon in the duodenum
checks the passage of food from the stomach. This experiment,
like v. Mering’s and Marbaix’s, does not explain normal conditions,
because, as already shown,? chyme normally gathers in the duo-
denum gradually, by repeated small additions, and even when accu-
mulated lies as a slender strand which does not distend the gut.
Each strand thus formed is soon hurried forward some distance
along the tube, thus clearing the duodenum for new accumulations.

Though the passage of food from the stomach may be checked
by artificially filling a long piece of the upper intestine or by sudden
distention of the gut at one point, such conditions cannot account
for any natural control of gastric discharge from the intestinal side,
because such conditions are not normally found. The evidence,
therefore, is opposed to the conception that mechanical agencies,
acting either in the stomach or in the intestine, play an important
part in controlling the normal gastric evacuation.

Chemical agencies in the stomach. —.More than twenty years ago
Ewald and Boas found,* by use of the stomach tube on man, that
there was a considerable development of free hydrochloric acid
before the gastric contents began to be notably diminished in
amount. Where the acid may have had its effect — whether on
peristalsis or on the pyloric sphincter——was not determined.?

CANNON: This journal, 1902, vi, p. 255 ; and 1904, xii, p. 389.
ToBLeER: Zeitschrift fiir physiologische Chemie, 1905, xlv, p. 195.
CANNON: This journal, 1902, vi, p. 262.
EWALD and Boas: Loc. cit., p. 364.
> HAMMARSTEN’S statement that HCl seems to act as a stimulus to open the
pylorus (Lehrbuch der physiologischen Chemie, 3d ed., Wiesbaden, 1895, p. 246)

1
2
3
4

The Acid Control of the Pylorus. 287

Penzoldt? has studied extensively the periods during which various
common foods remain in the stomach, and has noted that foods
delaying the appearance of free hydrochloric acid remain longest.
Verhaegen,” on the other hand, has declared that it matters little
for the passage through the pylorus whether the food is acid or
neutral.

As a basis for explaining the factors in control of gastric evacu-
ation, Penzoldt’s results are inconclusive. In general the foods
used were not fairly representative food-stuffs, but were often com-
plicated mixtures. The amounts given at different times were not
equal, nor was the consistency uniform. For his judgments Pen-
zoldt was dependent on securing remnants of the gastric contents
by means of the stomach tube, — a procedure not distinguishing the
differences in the chemical reaction of the food in the two ends of
the stomach, and furnishing no data as to the progressive rate at
which the stomach is emptied. Ignorance of the effects of varying
composition of foods, varying amounts and varying consistencies,
and ignorance of the rapidity of gastric discharge as digestion pro-
ceeds, renders it difficult to arrive at exact conclusions from Pen-
zoldt’s results.

In the presence of strong opposing evidence Verhaegen’s con-
tention that neither acidity nor neutrality of the chyme has any
effect on the emptying of the stomach, may reasonably be doubted,
for his observations were made with the stomach tube on only four
individuals, two of whom were pathologic.

Chemical agencies in the duodenum. —In 1893 Hirsch observed
that solutions of inorganic acids left the stomach slowly, and he
concluded that the slow exit was due to the stimulating effect of
the acid on the mucosa of the duodenum.? Later Serdjukow, one
of Pawlow’s students, inhibited gastric evacuation by introducing
acid into the duodenum through a fistula,* thus confirming the con-
clusion of Hirsch. Tobler’s results ® also substantiate it.

was an inference from the observations of EwALp and Boas (Personal com-
munication). The statement does not occur in later editions.

1 PENZOLDT: Deutsches Archiv fiir klinische Medicin, 1893, li, p. 535; 1894,
liii, p. 230.

2 VERHAEGEN : La cellule, 1897, xii, p. 69.

8 Hrrscu: Centralblatt fiir klinische Medicin, 1893, xiv, p. 383.

4 SERDJUKOW: Reviewed in HERMANN’S Jahresbericht iiber die Fortschritte
der Physiologie, 1899, viii, p. 214.

5 TOBLER: Loc. cit., p. 198.

288 W. B. Cannon.

The main defect of all the above methods as means for determin-
ing the nature of the chemical control of gastric discharge is their
failure to distinguish between the two factors concerned in the
passage of food through the pylorus. Before pointing out how
essential the distinction is, it will be well to consider the part played
by each of the two factors.

Tue Two Factors CONCERNED IN GASTRIC EVACUATION.

One of the two factors necessary for the emptying of the stomach
is the pressure to which the food at the pylorus is subjected by
recurring peristaltic waves; the other is the action of the pyloric
sphincter. Not until the X-ray method was used to study the
mechanical processes in digestion was it possible to watch, under
normal conditions, both the movement of gastric peristalsis and the
exit of food through the pylorus. Until the application of the
X-ray method, therefore, a clear distinction between the normal
effects of these two factors could not be made.

That the natural passage of food through the pylorus is occa-
sional might be due to occasional peristaltic constrictions, or occa-
sional specially strong peristaltic constrictions, pressing the gastric
contents against an easily opened pylorus; or, on the other hand,
the occasional passage might be due to an occasional relaxation of
the pylorus in the presence of a fairly constant pressure.

It is true that some of the investigators whose work has already
been mentioned have ascribed the control of gastric discharge solely
to the action of the pyloric sphincter. Marbaix, for example, writes
of the influence of the repletion of the intestine on the closure of
the pylorus.!. His evidence for this limitation is not clear. Von
Mering, on the other hand, recognized that intestinal repletion
might check gastric discharge by stopping gastric peristalsis, and
he resected the pylorus in order to differentiate, if possible, between
the peristaltic and the pyloric factors. The failure to make this
differentiation is the essential flaw, from the point of view of this
paper, in the methods of Ewald and Boas, Penzoldt, Hirsch, Serd-
jukow, and Tobler. Their results, therefore, while significant,
cannot serve for a conclusive determination of the control of gas-
tric evacuation,

The possible confusion of the two factors is illustrated in Paw-

1 MARBAIX: Loc. cit., p. 273.

oe

The Acid Control of the Pylorus. 289

low’s report of Serdjukow’s experiments. He states, without giving
evidence, that acid chyme entering the duodenum reflexly occludes
the pyloric orifice ‘‘ and at the same time reflexly inhibits the pro-
pulsive movements of the organ (stomach). + Clearly the occlu-
sion of the pyloric orifice alone would account for Serdjukow’s
results. What is the evidence that peristalsis also is affected?

In my first paper on the stomach I stated? that gastric peri-
staltic waves in normal conditions are continuously running, so long
as food remains. Hundreds of observations made since that time
on various animals — mainly on cats, but also on dogs, guinea pigs,
and white rats — and likewise records of stomach sounds in man,?
have confirmed the view that peristalsis continues uninterruptedly
until the stomach is swept clear of its contents. In my experience,
neither ejaculation of acid chyme, nor stretching of the duodenum
with food pressed through the cut pylorus (see footnote, p. 285), has
any tendency to interrupt the sequence of waves.

The continuously passing peristaltic waves of the stomach, as
remarked in discussion of mechanical agencies in the stomach, do
not show from moment to moment notable variation of intensity.
One of the two factors concerned in gastric discharge-— the pres-
sure in the antrum — is therefore recurrently constant. The control
of the discharge, consequently, must reside with the other factor,
i. e., With the action of the pyloric sphincter. If the sphincter holds
tight, the recurring waves churn the food in the antrum; if the
sphincter relaxes, these waves press the food out into the duodenum.
The pylorus is the “keeper of the gate.”

THE Facts. TO BE EXPLAINED.

The discharge from the stomach, as already demonstrated, is
occasional. The foregoing analysis proves that this occasional dis-
charge must be due to occasional relaxations of the pyloric sphinc-
ter. To explain the action of the pylorus, therefore, it is necessary
to consider agencies which maintain an intermittent closure, —
which usually keep the passage shut, yet open it at intervals to
allow portions of the chyme to depart. None of the researches on

1 PawLow: The work of the digestive glands, English translation, London,
1902, p. 165.

2 CANNON: This journal, 1898, i, p. 367.

3 Cannon: /did., 1903, viii, p. xxii; and 1905, xiv, p. 344.

290 W. B. Cannon.

the control of gastric evacuation, discussed in the preceding pages,
were definitely concerned with this intermittent closure. Further
investigation was desirable to explain the repeated opening and
shutting of the pyloric orifice.

To explain also the differences in the rate of discharge of differ-
ent food-stuffs from the stomach further investigation was neces-
sary. In the report of the research mentioned at the beginning of
this paper I called attention to the fact that when representative
carbohydrate, proteid, and fat foods, of uniform amount and con-
sistency, are separately fed, the carbohydrates begin to leave the
stomach soon after ingestion (within ten minutes) and are passed
out rapidly; proteids commonly do not leave the stomach at all
during the first half hour and sometimes not for an hour," then they
are expelled only slowly; and fats, because of a continuous slow
exit, remain in the stomach for a long period. Since the major
portion of a diet is more likely to be composed of carbohydrate or
proteid or of the two combined, than of fat, it becomes especially
important to understand the difference in the mechanical treatment
of these two main food-stuffs. What is the pyloric mechanism
whereby carbohydrates, not digested by the gastric juice, are per-
nitted to pass quickly into the small intestine to be digested, whereas
proteids, digested in the stomach, are there retained to undergo
digestion ?

A THEORY OF THE CONTROL OF THE PYLORUS.

The investigators whose views have been presented have re-
garded factors in the stomach, or factors in the intestine, as con-
trolling gastric evacuation. An interaction of agencies in the two
situations has not been considered. The theory referred to at
the beginning of this paper, propounded to explain the differential
discharge of the different food-stuffs, is based on evidence of
opposed effects from a single stimulus acting first in the stomach
and later in the duodenum.

1 MARBAIX’S declaration that when any food is introduced into the stomach
a portion passes directly into the empty intestine, and immediately causes v.
MERING’s reflex (MARBAIX: Loc. cit., p. 296), 1 have never been able to verify.
Animals that had fasted several days were given lean beef mixed with bismuth
subnitrate, which they eagerly devoured. Radiographs taken a half-hour later
showed that in spite of continuous peristalsis there was no sign of food in the
intestine.

The Acid Control of the Pylorus. 291

The first statement in the theory is that acid coming to the
pylorus causes a relaxation of the sphincter. Thus would be ex-
plained why the initial discharge is longer delayed when proteids
are fed than when carbohydrates are fed. Both carbohydrate and
proteid stimulate gastric secretion in abundance, as researches on
dogs by Pawlow and his co-workers,’ and as clinical studies on men
have shown. Inasmuch as carbohydrates do not unite chemically
with the acid, free acid is at once present in the stomach; and
carbohydrates would therefore begin almost immediately to pass
through the pylorus. Proteids, on the other hand, join with the
acid and thus retard for some time the development of an acid
reaction;? the proteid discharge would therefore be retarded.

But acid on the stomach side of the pylorus is not the only deter-
minant of pyloric action. This is proved by feeding carbohydrate
food moistened with 0.4 per cent hydrochloric acid. The rate of
discharge is not increased. If acid in the stomach is the stimulus
relaxing the pylorus, why in this case is the rate of discharge not
increased? The observations of Hirsch and Serdjukow now have
their bearing. Since it has been shown that acid in the duodenum
does not stop gastric peristalsis, the acid reflex from the duodenum
must affect the pyloric sphincter. The second statement in the
theory naturally follows, — acid in the duodenum closes the pylorus.

It is probable that the pyloric sphincter has normally a greater or
less degree of tonic contraction, with occasional relaxations.* Cer-
tainly it has a tonic contraction persistently strong for some time
after food enters the stomach; when proteid, for example, is fed,
peristaltic constrictions may press the food against the pylorus
repeatedly for an hour (approximately 300 waves) without fotcing
food through the orifice. |

The whole theory of the acid control of the pylorus may now
be stated. The pylorus is tonically closed when food is ingested,
and remains closed against recurring pressure. The appearance
of acid at the pylorus causes the sphincter to relax. The pressing
peristaltic waves now force some of the acid chyme into the duo-
denum. The acid in the duodenum at once tightens the sphincter
against further exit. The same acid also stimulates the flow of

1 PawLow: Loc. cit., pp. 36, 100.

2 DANILEWSKY: Zeitschrift fiir physiologische Chemie, 1881, v, p. 160.

3 See BASTIANELLI: MOLESCHOTT’S Untersuchungen, 1892, xiv, p. 93; and
OseER: Zeitschrift fiir klinische Medicin, 1892, xx, p. 29I.

292 W. B. Cannon.

alkaline pancreatic juice. Since no inorganic acid is normally pres-
ent beyond the first few centimetres of the small intestine? and
since the acid reaction of the contents in this uppermost region
is replaced throughout the rest of the small intestine by practically
a neutral reaction,® the acid chyme must be neutralized soon after
its emergence from the stomach. As neutralization proceeds, the
stimulus closing the pylorus is weakened; now the acid in the
stomach is able again to relax the sphincter. Again the acid food
goes forth, and immediately closes the passage behind it until the
duodenal processes have undergone their slower change. And thus,
repeatedly, until the stomach is empty. What is the evidence for
this theory?

EXPERIMENTAL EVIDENCE FOR THE ACID CONTROL OF
THE PyYLoRUS.

As the acid of the gastric juice, according to the theory, may
have two opposing effects on the pylorus, it will be well to present
first the evidence that acid in the antrum causes the pylorus to open,
and second the evidence that acid in the duodenum causes the
pylorus to be kept closed.

A. That acid in the stomach opens the pylorus. — The evidence for
this first half of the theory will be presented under several head-
ings, as follows:

1. Delaying the appearance of hydrochloric acid delays the mitial
discharge. — Observations on the gastric discharge of different
food-stuffs, of the same amount and consistency, proved that carbo-
hydrates begin to leave the stomach early and are passed out rapidly.
In terms of the above theory this quick exit is due to the early
appearance of acid in the stomach. The appearance of acid can be
delayed if the carbohydrates are first moistened with sodium bicar-
bonate. The acid would first be neutralized by the alkaline food
near the secreting surface and in the churning antrum; and only
aiter some time would free acid appear in considerable amount. If

1 Bay Liss and STARLING: Centralblatt fiir Physiologie, 1901, xv, p. 682.

2 Moore and BERGIN: This journal, 1900, iii, p. 325.

3 Munk: Centralblatt fiir Physiologie, 1902, xvi, p. 33.

* COHNHEIM, in his summary of the factors controlling the discharge of food
from the stomach (NAGEL’s Handbuch der Physiologie des Menschen, Braun-
schweig, 1907, ii, p. 564), mentions the theory here propounded, but states that my
evidence for it is not convincing. It is fair to note that this present paper gives
for the first time the evidence in a complete and detailed form.

The Acid Control of the Pylorus. 293

the theory is correct, this postponement of the appearance of acid
should delay beyond the normal time the initial discharge of the
food.

FIGURE 1.— Curves showing the average aggregate length of the
food-masses in the small intestine at the times indicated, after

feeding potato, rice, and crackers (4 cases each) moistened with

water (continuous line), and the same moistened with 1] per cent
NaHCOs (interrupted line).

Hours 34 1 2

Crackers, rice, and mashed potatoes were chosen as representa-
tive carbohydrate foods. The rice was steamed and dried, and the
mashed potato was also dried before being used. In all cases one
per cent sodium bicarbonate was added to the dried food until a
mush was made, of the same consistency as in the standard cases.
The carbohydrates thus prepared were mixed with subnitrate of
bismuth and fed, as in the standard cases, in 25 c.c. amounts.

In the following figures are presented the average aggregate
length of the food-masses in the small intestine, as seen by the
Rontgen rays, after feeding potato, rice, and crackers moistened
with water (four cases each), and moistened with sodium bicar-
bonate (four cases each). The figures for the first two hours of
observation are given, since they are most significant in judging
the rate of discharge.

| | POTATO. CRACKERS.

With water! .

With 1% NaHCO,

The average figures for twelve cases in which the three carbohy-
drates wet with water were fed, and the twelve cases in which they
were fed wet with sodium bicarbonate, are represented graphically
a Bic. 1.

1 See CANNON: This journal, 1904, xii, p.397. For a description of the method
used, see p. 388.

294 W. B. Cannon.

Comparison of the results of feeding carbohydrate food in the
two conditions shows that at the end of a half-hour there had
emerged only about one tenth as much of the food wet with the
alkaline solution as of the same food wet with water (in six of
the twelve cases no alkaline food had left the stomach) ; at the end
of an hour, from a third to a half as much; and in two hours, from
about a half to five sixths as much. In other words, there has been
a marked retardation in the discharge of carbohydrates wet with the
alkaline solution. This result is in harmony with the observation
by Jaworski on man, that alkalinity of the contents delays the
emptying of the stomach.’

Sodium bicarbonate delays the appearance of acid.in two ways:
it checks the secretion of the gastric juice,” and for a time it unites
with the acid of the gastric juice as rapidly as it is poured out.
The evidence here presented shows that experimental conditions
delaying the appearance of hydrochloric acid delay the discharge
from the stomach.

1 JAworSKI: Zeitschrift fiir Biologie, 1883, xix, p. 444.

2 PawLow: Loc. cit., p. 95. Evidence will be presented later in this paper
that conditions not favoring gastric secretion are accompanied by low pyloric tonus,
and that under these circumstances gastric evacuation may be very rapid. That
carbohydrate foods mixed with NaHCO, may be close to the line between a reten-
tion at the pylorus till an acid reaction develops, and a swift discharge because
the pyloric tonus is low, is indicated by observations on “flaked rice” moistened
with NaHCO ,. Three cases gave the following figures :

ELQUES Dee oiiesa oc ee 1 2
0.0 11.0 19.5
Centimetres of food masses ~ U.0 5.5 145
loo 7.0 22.0

Four other cases, fed in the same manner the same amount with the same consist-
ency, yielded the following figures:

Hours . 1 1 2
210 27.0 40.5
5 PZAS 34.0 47.0
Centimetres of food masses ie 0 36.5 43.5
14.5 41.5 52.0

These latter figures were extraordinary, and can only be compared with the results
of feeding raw egg-white, which likewise does not readily excite gastric secretion

(see page 313).

eS etl rh erlhr,mhCh ree

a,

7

—e—e

The Acid Control of the Pylorus. 295

2. Hastening the appearance of hydrochloric acid hastens the
initial discharge. — Proteids normally begin to leave the stomach
only after an interval of about a half-hour after feeding, and then
continue going out at a much slower rate than do carbohydrates.
According to the theory, as already stated, the slow passage of
proteids from the stomach is due to their union with the acid of
the gastric juice, which prevents the rapid development of a marked
acid state.

FIGURE 2.— Curves showing the average aggregate length of the
food-masses in the small intestine, at the times indicated, after
feeding fibrin, fowl and lean beef (4 cases each) as natural proteid

(continuous line), and the same as acid proteid (interrupted line).

Hours 4 1 2

Evidence as to this supposition may be secured by feeding proteid
food that has previously been changed to acid proteid. Fibrin, lean
beef, and fowl, freed from fat, were chosen as representative proteid
foods. They were mixed with ten per cent hydrochloric acid and
allowed to stand until changed to acid proteid. The free acid was
dialyzed away until test showed none present. As the change to
acid proteid was accompanied by swelling of the original substance,
approximately the same proteid content was preserved by feeding
the acid proteid in twice the standard amount. Doubling the
amount of the natural proteid notably retards the outgo from the
stomach ;! if changing the natural to acid proteid has no effect on
the outgo from the stomach, doubling the amount should likewise
retard the outgo, certainly should not accelerate it.

Fibrin, fowl, and lean beef were fed as acid proteids in 50 c.c.
amounts and with the same consistency as in the standard cases.
A comparison of the rate of discharge of the natural proteid foods,
and the rate of discharge of the same foods given as acid proteids
is exhibited in the following table. The figures represent the total
length of the food-masses in the small intestine at the indicated in-
tervals after feeding. In each instance the figures are averages of
four cases.

1 See CANNON: This journal, 1904, xii, p. 409.

296 W. B. Cannon.

FIBRIN. i| LEAN BEEF.

Hours

Natural proteid

Acid proteid

In Fig. 2 are presented the curves for the average figures of the
twelve cases in which the natural proteids were fed and the twelve
cases in which these same foods were given as acid proteids.

The figures of the foregoing table show that at the end of a half-
hour the stomach had discharged from five to ten times as much
acid proteid as natural proteid; three to ten times as much at the
end of an hour; and in two hours about twice as much acid proteid
as natural proteid. Evidently the change to acid proteid and the
teeding in increased amount resulted not in slowing, but in remark-
ably accelerating the exit from the stomach. The proteid in these
cases, already united with hydrochloric acid, does not unite with
the hydrochloric acid of the gastric juice. The hydrochloric acid
of the gastric juice secreted on the acid proteid is at once free acid.
Free acid appears earlier, therefore, than when the natural proteid
is fed. The evidence given above shows that when experimental
conditions hasten the appearance of free acid, the discharge from the
stomach is correspondingly hastened.

3. The appearance of acid in the antrum closely precedes the
initial discharge. Although in the experimental conditions already
described the emergence of food from the stomach has occurred
as if acid were present to open the pylorus, this judgment is only
an inference, — there has been no demonstration that acid was
present when the first food passed into the duodenum. It is desir-
able to determine more exactly the relation between the first de-
velopment of acid and the first exit of the food. This can be done
by establishing in the antrum, close to the pylorus, a fistula.

An antrum fistula holding a simple flanged cannula with a re-
movable plug was established in several cats... The cats recovered
readily from the operation and were usually in very good health.

1 Whenever an operation is mentioned or suggested in this paper, it is under-
stood, of course, that the operation was performed under complete general
anzsthesia.

The Acid Control of the Pylorus. 297

In order that the food could be seen with the X-rays when it first:
entered the duodenum, it was always mixed with bismuth sub-
nitrate. Various methods were used to determine the first appear-
ance of acid in the antrum. The method causing least disturbance
was that used when potato was fed. Mixed with the mashed potato
(25 c.c.) were 20 drops of dimethylamidoazobenzol — an amount
staining the white potato orange and showing a clearly marked
change to pink when hydrochloric acid developed. As soon as the
potato was given (usually by stomach tube), the plug was removed
from the cylinder of the cannula and replaced by a glass syringe.
The syringe consisted of a piece of glass tubing about 25 cm. long
into which was passed a close-fitting glass rod 5 cm. longer. Half
of a short length of rubber tubing was stretched over the upper
end of the glass tube; the other half firmly encircled the projecting
rod. Thus the joint was made tight. The glass tube, which slipped
snugly into the cannula, was held in place by a ring of rubber
tubing stretched around both the cannula and the syringe. By pull-
ing up the rod the thin mushy contents of the antrum were drawn
into the glass tube. Then any change of color could be noted. If
the original orange color still persisted, the rod was pushed down
again, and thus the food was restored normally to the stomach.
Usually such observations were made every four minutes; during
the intervals X-ray observations showed whether food had yet been
passed into the duodenum. When lean beef was fed, the color
change could not be clearly seen, and it was necessary to remove
through the cannula a sample of the antrum contents in a small
pipette. The contents were tested for acid with Congo-red, dimethyl-
amidoazobenzol, and tropaolin oo. The difficulty of this method
was the liability of loss of food whenever the plug of the cannula
was removed.

In some of the following cases the conditions varied from the
normal. These cases are reported, however, because of their double
value: they not only present direct evidence that an acid reaction
in the antrum precedes the initial gastric discharge, but they also
bring to the theory of. the acid control of the pylorus the support
arising from the concomitant variation of these two processes.

A cat suffering from a severe inflammation of the nose and eyes was given
25 c.c. mashed potato mixed with bismuth subnitrate and dimethylamidoazo-
benzol, as above described. The animal was examined alternately during
fifty-five minutes after feeding, for the presence of acid and for food discharged

298 W. B. Cannon.

from the stomach. There was no sign of acid, and although usually carbo-
hydrate food begins to leave the stomach in about ten minutes, there was no
discharge."

Five days later the same cat, in much better health though not yet well, was
again examined in the same manner. Gastric peristalsis was seen five minutes
after the feeding. Nineteen minutes later no food had left the stomach, and
the previous examination for acid had revealed no sign of change. At twenty-
one minutes the potato drawn into the syringe was pink. As soon as the
X-rays could be applied, the fluoroscope showed that there was food in the
duodenum.

A few days later the digestion of some meat was interrupted in the same cat,
now in good health, by pulling out through the fistula the larger pieces of the
gastric contents, flushing out the remnants, and then washing the stomach from
a stomach tube through the fistula until there was no acid reaction in the wash
water. ‘The cat was at once given potato as before. As soon as she was looked
at with the X-rays gastric peristaltic waves were seen. Five minutes after the
feeding the potato in the syringe was pink, and a minute later the X-rays
showed that some potato had passed the pylorus.

Other observations after feeding potato have confirmed these results, — an
acid reaction of the antrum contents was always noted before the food emerged
from the stomach.

. . ° . 4
Observations after feeding lean beef gave similar results. The
following cases are significant :

A cat finished eating voluntarily 25 c.c. lean beef mixed with 5 gm. sub-
nitrate of bismuth at 2.16 p.M. At 2.49 gastric peristalsis was prominent ; but
no food had left the stomach, and the test for acid in the antrum contents was
negative. At 3.00 o’clock the condition was unchanged. ‘Then a small amount
of o.4 per cent HCl was introduced into the antrum through the fistula.?
Within a minute thereafter there were two discharges of food into the duodenum.
As there was no further emergence for some time, a little food was removed and
tested. It gave no clear acid reaction. A small amount of 0.4 per cent HCl
was again introduced into the antrum towards the pylorus, and again food
emerged. Nothing more left the stomach for ten minutes. Then the food
was once more tested, with no clear sign of acid. The introduction of more
acid caused another discharge through the pylorus. It was now 3.40. Nothing

1 When animais are thus afflicted with “distemper,” food has been observed to
stagnate all day in the stomach (see CANNON and Murpuy, Annals of surgery,
1906, xliii, p. 534).

2 In none of these injections was the amount introduced sufficient to flood
the antrum or even to dilute the gastric contents go as markedly to alter their
consistency.

The Acid Control of the Pylorus. 299

left the stomach during the next five minutes, although deep strong peri-
staltic waves had been passing continuously and pressing the gastric contents
into the cannula whenever the plug was removed. An hour and a half had
elapsed and only a very small amount of food had left the stomach, and that
had left only when acid had been experimentally introduced. This was an
unusual delay. The plug was now removed, and the peristaltic pressure per-
mitted to drive out the gastric contents through the cannula. There was no
sign of an acid reaction. It seemed as if the stomach had not been actively
secreting. Certainly the acid introduced gave only a temporary and local
acidity. ’

Twenty-five cubic centimetres of the same meat which the cat ate were now
tested for alkalinity. More than ro c.c. of 0.4 per cent HCl were added, and
the reaction was still alkaline to Congo-paper blue with o.4 per cent HCl.
Strong HCl was now added ; 20 drops were required before a neutral point
was reached. Evidently the meat for some unknown reason was strongly
alkaline.

It was later found that by mistake the meat had been boiled in a receptacle
in which some instruments had previously been boiled with sodium carbonate,
and which had not been cleaned.

Another cat with a fistula in the antrum finished eating, with evident relish,
25 c.c. boiled and shredded lean beef plus 5 gm. bismuth subnitrate at 12.05
p.M. At 12.14-15 she was fastened to the holder and examined. ‘he reac-
tion of the antrum contents was not acid, and nothing had left the stomach.
At 12.22~23 nothing had left, but the Congo test had changed from a light to
a dark red. At 12.30-31 the Congo test showed a still stronger acid change,
and at 12.36, when examined by the X-rays, the intestine contained food some
distance from the stomach.

The same cat several days later was offered the same kind of food, which
she refused to eat. The food was then given to her by spoon, but with much
difficulty, for she pushed out the food with her tongue and only received it all
finally after repeated refusals. Ten minutes after the last mouthful was swal-
lowed she was fastened to the board and examined as in the previous experi-
ment. She was restless, she mewed and frequently tossed about. For an hour
after the feeding the contents did not become acid, and although peristaltic
waves were at times clearly seen, no food passed the pylorus. That emotional
states inhibit the flow of gastric juice has been pointed out by Bickel.’ In this
experiment it is probable that there was no psychic secretion at the time of
eating, and that subsequent secretion was inhibited while the animal was fas-
tened down. The absence of an acid reaction was attended by a failure of
discharge from the stomach.

1 BICKEL: Deutsche medicinische Wochenschrift, 1905, xxxi, p. 1829.

300 W. B. Cannon.

The above cases prove that a delay in the appearance of acid in
the antrum contents, as tested through a gastric fistula close to the
pylorus, is associated with a similar delay in the passage of food
from the stomach; that this may occur in spite of vigorous gas-
tric peristalsis; that in these circumstances the introduction of a
small amount of acid near the pylorus causes immediately the exit
of food through the pylorus; and that whether potato or beef is
fed, and whether in the same animal the discharge begins at the
usual time or is much. retarded, the first delivery of food into the
duodenum is normally preceded by the development of an acid re-
action in the antrum.

These conclusions from observations on the gastric contents
through a fistula in the antrum are completely confirmed by recent
studies of the reaction of the discharged chyme. Tobler, London
and Sulima, and London and Polowzowa have tested the chyme
collected from a duodenal fistula close to the pylorus. Tobler fed
iean beef to his dogs. The repeatedly discharged gastric contents
are acid from the beginning, and continue during digestion to be
“stark sauer.”’* London and Sulima? record that when cooked
ege-albumin is fed, the discharge from the pylorus is initiated by
the pouring forth of an acid fluid. The same condition is recorded
by London and Polowzowa? after feeding white bread.

4. Hydrochloric acid opens the pylorus of the excised stomach, —
Magnus has shown® that pieces of the sinall intestine, removed
from the body and placed in continuously oxygenated Ringer’s solu-
tion, will remain alive and, so long as Auerbach’s plexus is intact,
will manifest the typical local reflex. In a recently published in-
vestigation I have given evidence that the mechanism in control of
the differential discharge through the pylorus is independent of the
central nervous system.® It seemed probable, because of the rapid-
ity of closure of the pylorus after food emerged, that the controlling

1 All these observers report that a few drops of alkaline mucus flow from the
cannula soon after observation begins. This flow does not seem to be associated
with the repeated gastric discharge.

2 TOBGER: Loc. Gif-, p. 197.

8 LonpoN and SuLimA: Zeitschrift fiir physiologische Chemie, 1905, xlvi,
po 205:

4 Lonpon and PoLtowzowa: Zeitschrift fiir physiologische Chemie, 1906,
xlix, p. 340.

5 MaGnus: Archiv fiir die gesammte Physiologie, 1904, cii, p. 362.

® CANNON: This journal, 1906, xvii, p. 429.

= ———E—

, =e *

Eee em CUTTS rs—s—SCO

The Acid Control of the Pylorus. 301

mechanism resides in the local nerve plexus, and is similar to the
typical reaction of the intestinal wall. On this supposition the fol-
lowing experiment has been repeatedly performed.

A cat, which had fasted for twenty-four or thirty-six hours, was killed by
etherization. A cut above the cardia and another just below the pylorus
separated the stomach from the rest of the alimentary canal. The stomach,
which was empty,’ was cleared of its attachments and placed in warm Ringer’s
solution (38° C.), through which oxygen continuously bubbled.

A glass tube, with a short rubber tube and a water manometer attached, was
tied into the cardiac orifice. A small amount of 0.4 per cent HCl, made blue
by the changed Congo-red, was introduced through the tube into the cardiac
end of the stomach, which was held lower than the pyloric end. The stomach
was now inflated with air until the air bubbled through the pylorus. The rub-
ber tube was next tightly clamped. When the air had ceased escaping from
the stomach, z. ¢., when the pyloric tonus withstood the intragastric pressure,
the cardiac end of the stomach was gently and slowly turned until the acid
came to the pylorus. In a moment the blue fluid poured forth into the
Ringer’s solution. The pylorus had opened.

It might be supposed that the acid coming into the antrum caused
an increased tonus of the gastric musculature and that thus the
pyloric orifice was forced open. The manometer, however, does
not show any increase of intragastric pressure. Furthermore the
stomach can be tipped so that the acid fluid enters the antrum, but
does not come to the pylorus. This does not lead to the driving out

‘of more air, — the acid does not notably stimulate contraction of the

gastric wall. The opening of the pylorus, therefore, is due to the
presence of the acid.

A one per cent sodium bicarbonate solution colored red, similarly
brought to the pylorus, does not begin to emerge for a considerably
longer time, and then usually drifts out into the Ringer’s solution
as if slowly diffusing.

It is justifiable to conclude that in the living excised stomach
free acid coming to the pylorus causes the pylorus to open.

B. That acid in the duodenum keeps the pylorus closed. — The sup-
port for this, the second half of the theory, has already been sug-
gested in part in discussing the experiments on the inhibition of

1 It is important that the stomach be taken while not digesting. In my experi-
ence, if digestion has been going on, the excised stomach exhibits peristalsis as
soon as inflated.

302 W. B. Cannon.

gastric discharge by acid in the duodenum. As other observations
to the same effect are to be described under the above heading, a
brief restatement of the previous experiments and their results will
not be out of place, and will serve to bring all the evidence together.

1. Acid in the duodenum inhibits gastric discharge. — In 1893,
Hirsch,! as already noted, found that inorganic acids left the
stomach slowly. When he isolated the stomach, however, the acids
departed as rapidly as any other fluid. He explained this difference
by assuming that the stomach is controlled by acid reflexes from
the duodenum. Serdjukow modified Hirsch’s experiment by in-
troducing through a duodenal fistula small quantities of acid solu-
tions or pure gastric juice. By repeated injections it was possible
to prevent discharge from the stomach for an unlimited time.?
Tobler’s observations were closer to the normal conditions. He
allowed a dog with duodenal fistula to eat 100 gm. lean beef. The
chyme as it emerged was caused to leave the duodenum through
the artificial opening. The stomach was thus emptied in about two
hours and fifteen to thirty minutes. The next day the dog was
given the same amount of the same kind of food, but whenever a
portion of the chyme came through the fistula from the stomach, a
similar portion of the chyme of the day before was injected through
the fistula towards the intestines. The result was that the chyme
left the stomach at considerably longer intervals and was more
thoroughly digested. The time of digestion thus became length-
ened to three hours and three hours and a half. Tobler’s observa-
tions have been completely confirmed by Lang.*

The evidence of Hirsch, Serdjukow, Tobler, and Lang proves
definitely that acid chyme in the duodenum checks the outgo from
the stomach. Since gastric peristalsis, as previously shown in this
paper, is not stopped by the discharge of acid chyme, the effect must
be due to the action of the pyloric sphincter. Acid in the duodenum
causes pyloric contraction.

2. Absence of the normal alkaline secretions from the duodenum
retards gastric discharge. — Pawlow records that the passage of
acid solutions out of the stomach is remarkably slower in dogs with

1 Hirscu: Loc. cit., pp. 378, 383.

2 SERDJUKOW: Loc. cit. Also PAWLOW: Loc. cit., p. 164.
$ TOBLER =, Loc. tti.,.p: 197-

4 LanG: Biochemische Zeitschrift, 1906, ii, p. 225.

4

The Acid Control of the Pylorus. 303

a pancreatic fistula than in those without one.!_ In order to test
whether the discharge of normal gastric contents was likewise re-
tarded by a similar condition in the duodenum, the following ex-
periment was performed. The larger pancreatic duct and also the
bile duct were tied so as to prevent the flow of the secretions into
the intestine. After several days the animals were given the stan-
dard amount of mashed potato and bismuth subnitrate, with the
usual consistency. The outgo from the stomach was determined
as before by measuring the length of the food-masses in the small
intestine. The figures of the following table give the total length
of these masses at the times indicated, in normal conditions (four
cases) and also after tying the larger pancreatic and the bile ducts
(four cases).

POTATO.

Normal conditions

Pancreatic and bile ducts tied | 0.0

The observations recorded in this table are represented graphi-
cally in Fig. 3. These observations were made six and twelve days

FicurRE 3.— Curves showing the average aggregate length of the
food-masses in the small intestine, at the times indicated, after
feeding potato (4 cases) in normal conditions (continuous line),
and the same after tying pancreatic and bile ducts (interrupted

line).

after the operation. It is obvious that there has been a very marked
checking of the normally rapid outgo of the potato from the
stomach; nothing out in a half-hour, a fourth the normal amount
in an hour, and a third the normal at the end of two hours.

Why there should be no exit of the food during the first half-hour

1 PAWLOW: Loc. cit., p. 164.

304 W. B. Cannon.

is not clear, but the very slow increase of the intestinal contents
thereafter — from 7.5 to 14.5 cm. in the second hour of digestion,
compared with the increase from 10 to 31.5 cm. in the second half-
hour in the normal state — is in harmony with the observation that
acid in the duodenum closes the pylorus.

Under normal conditions acid in the duodenum stimulates the
secretion of pancreatic juice and bile. These alkaline fluids must
neutralize the acid chyme, for an acid reaction is not found beyond
the first few centimetres of the small intestine.’ The neutralizing
of the acid removes the stimulus keeping the pylorus closed. If the
alkaline fluids are prevented from entering the intestine, the acid
is necessarily neutralized more slowly, the pylorus is kept closed
during longer periods, and the emptying of the stomach therefore
occurs at a slower rate.

FIGURE 4.— Curves showing the average aggregate length of the
food-masses in the small intestine, at the times indicated, after
feeding lean beef (4 cases) in normal conditions (continuous
line), and the same after setting aside the duodenum (inter-
rupted line).

Hours } 1 2

3. Destroying continuity between stomach and duodenum hastens
gastric discharge. — Additional evidence as to the relations between
the duodenum and the pylorus in the control of gastric evacuation
may be secured by setting aside the duodenum and causing the
stomach to empty into a lower part of the gut. The intestine was
cut through about 1.5 cm. beyond the pyloric furrow, and again
about 30 cm. beyond. The upper end of this separated portion was
turned in and closed with stitches; the lower end was joined to the
gut near the ileocolic opening by an end-to-side junction. The
upper end of the main part of the intestine was now united to the
small remnant of duodenum contiguous to the pylorus by the simple
and effective F. G. Connell suture.2, Thus the stomach emptied not
into the duodenum, but into a piece of the intestine formerly 30 cm.
beyond.

1 See page 292.
2 See F. G. CONNELL: Journal of the American Medical Association, 1go1,

XXXVil, p. 952. My thanks are due to Dr. F. T. Murpuny for indispensable aid in
this operation. ,

The Acid Control of the Pylorus. 305

The following figures give the average aggregate length of the
food-masses in the small intestine, at the times indicated, after
feeding shredded lean beef of standard amount and consistency,
in four normal cases and in four cases with duodenum set aside.
From two to nineteen days had elapsed since the operation.

Hours

Normal conditions .

Duodenum set aside

These results are represented graphically in Fig. 4. Reference
to the table and to Fig. 4 shows at once the difference between the
factor which acts inside the stomach and the factor which acts in
the duodenum to control the pylorus. Both conditions reported in
the above table display the typical retardation of the initial dis-
charge characteristic of proteids. Setting aside the duodenum evi-
dently did not change that. That retardation, according to the
conclusions already reported, is an affair of the stomach alone.
And the figures in the above table serve to confirm those conclusions.

When the food begins to emerge, the results are suddenly quite
different. Instead of 3 cm. at the end of an hour, 16 cm.; and twice
the normal amount at the end of two hours — such is the effect of
destroying the continuity between stomach and duodenum. After
the first delay (in one case no food left the stomach for an hour)
proteid is poured forth at a remarkably rapid rate.

The above results were secured before the completion of the
research on the passage of different food-stuffs from the stomach
deprived of its extrinsic nerves, to which reference has been made.*
That research proved that the differential discharge from the stomach
is under local control. As the investigations of Magnus had shown
that the local control of intestinal reflexes resides in Auerbach’s
plexus, it seemed probable that merely cutting a ring around the
intestine as close as possible to the pylorus, and deep enough to sever
both muscular coats, would yield information as to the path of in-
fluence from duodenum to stomach. A ring was cut as above
described, and the separated edges of the muscular coats were then

1 See CANNON: This journal, 1906, xvii, p. 429.

306 W. B. Cannon.

held together by only the mucosa and the submucous connective
tissue. When proteid was fed, there was again the initial delay —
nothing out at the end of a half-hour —and this was followed by
an exit almost as rapid as when the duodenum was set aside. The
conclusion may be drawn that the influence from duodenum to
pylorus runs through a local reflex, mediated by the myenteric
plexus. As Bayliss and Starling have shown ‘that reflex augmenta-
tion of intestinal contraction may occur from 1 to 6 cm. above a
stimulated point,? it is clear that acid chyme may be effective
through a considerable extent of the duodenum in causing reflex
pyloric contraction.

EvIDENCE FOR THE ACID CONTROL DERIVED FROM PREVIOUS
OBSERVATIONS ON GastTrRIC DISCHARGE.

As already explained (p. 287), the data as to the discharge from
the stomach, secured by use of the stomach tube, are of little service
for the present investigation, because they do not indicate the rate
of gastric evacuation from time to time during digestion. The
method used in my earlier research on the discharge of the different
food-stuffs from the stomach? gave characteristic curves of the
rates at which proteids, carbohydrates, and fats pass the pylorus.
The present investigation was undertaken to explain these character-
istic rates of discharge. To what extent do the results of the earlier
research agree with the other evidence that acid in the stomach
signals the opening of the pylorus?

Proteids. — In the earlier research above referred to, X-ray ob-
servations showed that proteids frequently did not begin to leave
the stomach during the first half-hour, and that after they began to
leave they departed slowly (see curve of natural proteids, Fig. 2).
Although in comparing results of different investigations the factor
of food consistency cannot be closely estimated, yet observations
through duodenal fistula on the passage of proteid chyme from the
stomach support in general the X-ray observations. Thus, for
example, Moritz? noted that the exit of the gastric contents began
about three quarters of an hour after his dog finished eating 200 gm.

raw meat. And Lang reports that the first slight discharges of the

1 BAYLiss and STARLING: Journal of physiology, 1899, xxiv, p. 112.
? CANNON: This journal, 1904, xii, p. 387.
8 Moritz: Zeitschrift fiir Biologie, 1go1, xlii, p. 574.

The Acid Control of the Pylorus. 307

gastric contents did not occur for at least fifteen minutes after
feeding his dogs 200 gm. fibrin.’ Peristalsis starts almost immedi-
ately after the ingestion of food. The dog’s stomach has about
four waves per minute.” It is clear that in these cases of duodenal
fistule the food has been churned by numerous peristaltic waves,
and these waves have repeatedly pressed food upon the pylorus,
before the sphincter has relaxed and permitted an exit into the
duodenum. The evidence that the exit does not occur until the con-
tents of the antrum are acid has already been given. ‘The first acid
secreted unites with the proteid. The relatively long delay of the
initial discharge of proteid from the stomach is thus accounted for
by a relatively slow development of a marked acid reaction in the
food which is pressed up to the pylorus.

Doubling the amount of proteid food strikingly delays the initial
discharge of the proteid from the stomach *—a result explicable
on the ground that the increased amount of proteid to become acidi-
fied in the antrum necessarily delays the proper acid reaction for
opening the pylorus.

The continued comparatively slow outgo of proteid into the in-
testine can also be explained. As proved in my first paper on the
stomach, mixing currents do not run throughout the cavity. The
mixing occurs only in the pyloric end; the centre of the mass in
the cardiac end long remains unchanged in reaction.* Since the
antrum does not secrete acid, all the acidity of its contents is due to
acid pressed in from the cardiac end. But unchanged proteid, stored
in the cardiac end, is also continuously being pressed into the antrum.
There is thus continuous utilization of the imported acid. Since it
is altogether probable that a certain degree of acidity is necessary
for opening the pylorus, the fresh proteid masses, by uniting with
the acid and thus reducing the acid reaction, would naturally dimin-
ish the rate of exit from the stomach. That this factor is important
in checking the rapid outgo of proteid food is indicated by the fact
that acid proteids, not demanding large amounts of acid, pass the
pylorus with almost carbohydrate rapidity (cf. Figs. 1 and 2).

Doubtless also the proteid discharge continues to be slow because

1 LanG: Loc. cit., p. 229.

2 Roux and BALTHAZARD: Archives de physiologie, 1898, xxx, p. 89.

8 CANNON: This journal, 1904, xii, p. 410.

* Cannon: /ézd., 1898, i, p. 378. See also GRUTZNER: Archiv fiir die gesammte
Physiologie, 1905, cvi, p. 463.

308 W. B. Cannon.

proteid chyme presents a greater amount of acid for neutralization
than does carbohydrate chyme. Tobler and Lang have shown that
acid proteid in the duodenum will check gastric evacuation.’
Khigine’s results prove that when 200 gm. flesh are fed to a dog,
50 per cent more gastric juice is secreted during the first four hours
of digestion than is secreted in the same time when the same amount
of bread is fed.2 The neutralizing of the larger amount of acid in
the duodenum would necessarily require a longer time, and would
result in a slower rate of discharge than would be expected when
bread is fed.?

Penzoldt’s observation that, with due regard to the effects of vari-
ation in quantity, those flesh foods with which there is an earlier
appearance of acid remain a shorter time in the stomach than those
with which the acid appears later; and likewise his observation that
of the vegetables legumes, which are richest in proteid and can
unite with much acid, are longest delayed in the stomach, — these
observations * made on man are in agreement with the supposition
that the pylorus remains closed until acid appears. The statements
of Roux and Balthazard,® and of Moritz,® based on animal experi-
mentation, that raw meat remains in the stomach a long time and
leaves. slowly, confirm Penzoldt’s results. On the other hand,
Cahn’s* contention that emptying of the stomach begins with pep-
tonization, and Roux’s § declaration that concentrated peptone seems
to accelerate evacuation, do not oppose the idea of the acid control
of the pylorus in normal conditions, for when proteid is fed, acid
would be present by the time peptonization had occurred in any
considerable amount. To suppose that peptone is required for
pyloric relaxation is manifestly unwarranted, — the pylorus opens

1 ToBLER: Loc. ctt., pp. 197, 198; LANG: Loc. cét., p. 240.

2 KHIGINE: Archives des sciences biologiques, St. Petersburg, 1895, iii, p. 461;
see also PAWLOW: Loc. cit., p. 35.

3 KHIGINE’Ss figures show slight gastric secretion continuing for ten hours after
the ingestion of 200 gm. bread, but continuing only eight hours after the ingestion
of 200 gm. flesh. I cannot believe that the bread was longer in the stomach than
the meat; all observations on the evacuation of the stomach indicate that carbo-
hydrate departs much earlier than the same amount of proteid.

* PENZOLDT: Deutsches Archiv fiir klinische Medicin, 1894, liii, p. 217.

5 Roux and BALTHAZARD: Loc. cit., p. QI.

® Moritz: Loc. czt., p. 569.

CAHN: Zeitschrift fiir klinische Medicin, 1887, xii, p. 4t.
Roux: Comptes rendus, Société de Biologie, 1901, liii, p. 846.

a

The Acid Control of the Pylorus. 309

early when carbohydrates are fed, yet under the circumstances no
peptone can have been formed.

Carbohydrates. — With reference to the departure of carbo-
hydrates from the stomach it is of interest to recall Marbaix’s sug-
gestion. He noted that potatoes leave the human stomach rapidly,
and that gastric juice cannot attack them to any extent; he pointed
out that an important question lay here.1. The testimony that the
delay in the discharge of carbohydrates from the stomach is usually
not great,” that in conditions of diminished acidity they are more
easily borne than meat,® that test meals mainly carbohydrate leave
the stomach earlier than those containing meat, all points to the
possibility that acid on the gastric side of the pylorus signals the
relaxation of the sphincter.

A portion of the curve showing the rate of discharge of typical
carbohydrate foods (the curve for the first two hours of digestion)
is reproduced in Fig. 1. The original curve, published in 1904,
corresponds remarkably to that recently published by London and
Polowzowa,° representing the hourly percentage volume of gastric
evacuation collected from a duodenal fistula after feeding white
bread. The two curves almost exactly coincide during the first two
hours of digestion; thereafter my curve, although similar in shape,
is naturally somewhat higher, since it represents not merely the dis-
charge, but the accumulation of the discharge in the small intestine,
minus the amount absorbed or passed into the large intestine. The
method of study used by London and Polowzowa completely cor-
roborates the X-ray method which I used. Both methods show that
carbohydrate foods begin to leave the stomach soon after ingestion.
I noted them in the duodenum ten minutes after ingestion; eight
to twelve minutes is the time given by London and Polowzowa
before the bread chyme emerges, acid in reaction. The carbo-
hydrate foods, once started, pass out rapidly; indeed, they remain
in the stomach only about half as long as the same amount of proteid
food having the same consistency.

As previously intimated, this early and rapid exit is in accord
with the other evidence that acid in the antrum relaxes the pyloric

1 MARBAIX: Loc. cit., p. 299.

2 See PENZOLDT: Archiv fiir klinische Medicin, 1893, li, pp. 549, 559-
8 ScHULE: Therapeutisches Monatschrift, 1899, xiii, p. 601.

4 CANNON: This journal, 1904, xii, p. 398.

5 LonpDon and PoLowzowa: Loc. cit., p. 364.

310 W. B. Cannon.

sphincter. It is well known that carbohydrate food stimulates the
flow of gastric juice.t But carbohydrates do not unite with the
acid. Hydrochloric acid is consequently at once present to open
the pylorus. And the acid is secreted as rapidly as the duodenum
can receive the chyme, for the giving of carbohydrate already mixed
with acid, does not increase the rate of the passage into the intestine.

Combinations of the food-stuffs. —In the research on the passage
of different food-stuffs from the stomach it was found that when
carbohydrate was fed first and proteid second, the proteid, filling
the cardiac end of the stomach, did not materially check the depar-
ture of carbohydrate food, lying in the antrum; but proteid in the
antrum, when proteid food is fed first and carbohydrate second,
results in the characteristic slow discharge. The proteid holds back
the carbohydrate occupying chiefly the cardiac end. In the former
case the carbohydrate content of the antrum did not retard the
development there of an acid reaction; in the latter case the proteid
did retard that development. This observation indicates that the
acid, which opens the pylorus, acts close to the pylorus, — a conclu-
sion which is sustained by the effects of acid in the excised stomach
(see p. 301).

When carbohydrates and proteids were mixed in equal parts, the
mixed food did not leave the stomach so slowly as the proteids, nor
so rapidly as the carbohydrates; the discharge was intermediate
in rapidity. This result was to be expected, for a large proportion
of proteid was present to unite with the acid secreted, and this would
tend to retard the discharge in the manner already discussed
(Seer p2307))t

In a mixture of fats and proteids in equal parts the presence of
the fat caused the mixture to leave the stomach even more slowly
than the proteid alone. This result also is in accord with the sup-
position that acid opens the pylorus, for fat alone inhibits, and fat
mixed with proteid notably retards and diminishes, the flow of
gastric juice.” Moreover the development of an ‘acid reaction is
checked by the union of the acid with the proteid. It is quite natural
that this combination of food-stuffs should be slowest of all to pass
from the stomach.

Fat mixed with carbohydrate in equal amounts caused the carbo-

1 PAwWLow: Loe. cit., pp: 36, 100.
2 PAWLow: Loc. cit, pp. 97, 103. Also Fermi: Archiv fiir Physiologie,

Supplement-Band, rgot, p. 76.

—o-

The Acid Control of the Pylorus. 311

hydrates.to pass the pylorus at a rate slower than their normal.
In this case the fats again undoubtedly retarded and diminished the
gastric secretion, but the carbohydrates, unlike the proteids, did
not further hinder the appearance of an acid reaction. The checking
of the outgo can therefore be explained solely by the effect of the
fats in diminishing the secretion of gastric juice.

A review of the evidence presented in this section shows that,
observations on the rate of discharge of proteids, carbohydrates,
fats, and combinations of these food-stuffs, can be readily explained
on the assumption that acid in the stomach opens the pylorus and
acid in the duodenum closes it. This fitness of the theory to explain
the peculiar differences in the gastric discharge of the different food-
stuffs, makes it still more probable as a statement of the normal
mechanism of the pyloric passage.

OBSERVATIONS NOT ACCORDANT WITH THE ACID CONTROL.

The argument has been suggested against an acid control of the
pylorus, that other sphincters of the alimentary canal are not thus
controlled, and therefore it is unnecessary to assume such control
for the pyloric sphincter. The pyloric sphincter, however, is in
several respects unlike any other in the alimentary canal: (1)
peristaltic waves are rhythmically pushing food against it (five or
six times per minute in the cat), sometimes for half or three quarters
of an hour, without causing relaxation; (2) the pylorus has on
either side a secretion of opposite reaction — acid above, alkaline
below —a condition unlike any other alimentary sphincter except
the cardia.1 Because of these peculiarities it is unjustifiable to
argue from the control of other sphincters as to the control of the
pylorus, especially since the local changes in chemical reaction can
explain the normal pyloric functioning.

It has also been urged that much of the gastric contents must
escape before it is conceivable that any great proportion of acid
is present. This objection to the acid control fails to take into
account the important difference between the two ends of the

stomach. What may be true of the bulk of the food lying in the

cardiac end may not be at all true of food in the antrum. A small
amount of food in the antrum may be thoroughly mixed with acid

1 A research yet unpublished shows that the cardia may be kept ene closed
during gastric digestion by the acid gastric contents.

312 W. B. Cannon.

when merely the surface of the mass in the cardiac end has been
slightly acidified. As has been previously shown in this paper (see
pp. 301, 310), it is the acid reaction of the food at the pylorus that
is significant in causing the sphincter to relax.

The discharge of water. — Another objection to the idea that acid
on the stomach side opens the pylorus is the fact that water is very
rapidly discharged. For example, Moritz, in studying by means
of a duodenal fistula the emptying of the stomach, noted that water
begins to enter the intestine almost as soon as it enters the stomach;
it may pass out in single gushes or continuously. In thirty minutes
500 c.c. of water may go from the stomach into the intestine.? Simi-
lar results have also been reported by other observers who have
studied the exit of water from the stomach. Not only water, but
likewise physiological salt solution, may go out rapidly.*

It should be noted, in the first place, that water and salt solution
are very different in consistency from the foods ordinarily taken
into the stomach. Furthermore, water and salt solution produce
only a very slight, if any, secretion of gastric juice. When only
100 or 150 c.c. of water are injected, very often not the least trace
of secretion occurs. “It is only a prolonged and widely spread
contact of the water with the gastric mucous membrane, which gives
a constant and positive result (secretion).”° The rapid exit of
water from the stomach would preclude the conditions which make
it even a feeble stimulant of gastric secretion. The failure of water
to excite any noteworthy amount of gastric juice favors a rapid exit,
so far as the duodenal reflex is concerned, for the acid stimulus
closing the pylorus is thereby absent. Within the stomach water
certainly has an effect on the pyloric sphincter very different from
foods which evoke an abundant flow of gastric juice. When such
foods are given, scores of peristaltic waves may sweep up to the
pylorus before the sphincter relaxes; but when water is given, it
begins to leave the stomach at once.

It seems probable that a state of increased pyloric tonus accom-
panies the conditions favoring secretion of gastric juice. Pawlow has

1 CANNON: This journal, 1898, i, pp. 378, 379-

2 Moritz: Zeitschrift fiir Biologie, 1go1, xlii, p. 584.

8 See GLEY and RONDEAU: Comptes rendus, Société de Biologie, 1893, xlv,
p- 517- Roux and BALTHAZARD: Archives de physiologie, 1898, xxx, p. go.

4 Moritz: Loc. cit., p. 589; also CARNOT and CHASSEVANT: Comptes rendus,
Société de Biologie, 1906, lx, p. 866.

5 PAWLOow: Loc. cit., p. 94.

The Acid Control of the Pylorus. g13

shown that the psychic secretion of gastric juice is due to impulses
coming to the stomach by way of the vagi.1 Vagus stimulation also
produces an augmentation of the contraction of the pyloric sphinc-
ter.2 Vagus impulses, therefore, cause the initial flow of gastric
juice —the psychic secretion—and they also cause increased
pyloric tonus. Water does not present the conditions for psychic
secretion: it is not chewed with a relish; it is swallowed rapidly;
it does not satisfy appetite; and once in the stomach, it begins to
pass immeciately through the pylorus, as if the sphincter were in
a relaxed state. The fact that water may pour through the pylorus
not rhythmically but fairly continuously (see p. 312) points definitely
to a diminished pyloric tonus. This fact and the failure to stimu-
late gastric secretion seem related to each other. In these facts
may be found a probable explanation of the rapid discharge of
water from the stomach.?

The discharge of egg-albumin. — In the same class with water is
raw egg-white. In my observations on the rate of discharge of
different food-stuffs from the stomach, I pointed out that egg-
albumin formed an exception to the general rule that proteid passes
out from the stomach slowly.* Since then this observation has been
confirmed by London and Sulima in a study on dogs with duodenal
fistule. They found that raw egg-albumin begins to pass the py-
lorus immediately after ingestion; it emerges in large gushes at
intervals of four or five seconds. These gushes are therefore too
frequent to correspond to the occurrence of peristaltic waves. For
about twenty minutes the egg-white issues from the stomach with
an alkaline reaction; then the reaction becomes acid, and the dis-
charge naturally is more seldom (one to three minute intervals) and

1 PawLow: Loc. cit., p. 51.

? OPENCHOWSKI: Centralblatt fiir Physiologie, 1899, iii, p. 4. OSER (Joc. cit.
p. 288) states that vagus stimulation completely closes the open pylorus. See also
MAy, Journal of physiology, 1904, xxxi, p. 270.

8 That water does not go rapidly from the stomach merely because of its fluid
consistency is shown by observations of Moritz (/oc. ctt., pp. 589, 590). Weak HCl,
he states, passes out more slowly than water, and beer passes out with even greater
retardation. The slow exit of weak HCl may be explained solely by its effect in
closing the pylorus from the duodenal side. Beer stimulates gastric secretion not
only by its alcohol content, but because it is bitter (see PAWLow: Loc. cét., pp.
138, 139, and CHITTENDEN, MENDEL, and JACKSON: This journal, 1898, i, p. 207).
Beer, therefore, must go out slowly, because of the acid control of the pylorus.

* CANNON: This journal, 1904, xii, p. 399.

314 W. B. Cannon.

less abundant.1. In this connection it is of interest that Pawlow
found fluid egg-white no more effective in exciting gastric secretion
than an equal volume of water.? Like water, fluid egg-white does
not offer the conditions for arousing psychic secretion; and attend-
ing that condition there is a state of diminished pyloric tonus, as
evidenced by discharges through the pylorus much more frequent
than the peristaltic waves in the dog’s stomach. The rapid passage
of fluid egg-white from the stomach would therefore be explained
in the same manner that the rapid outgo of water is explained.*

According to the results of my earlier investigations, however,
egg-white coagulated by heat also left the stomach at a rapid rate.
This observation likewise has been confirmed by London and Sulima.
They found, however, that, unlike fluid egg-white, the coagulated
form did not begin to leave the stomach immediately, but several min-
utes after ingestion. When the gastric discharge began, its reaction
was acid. First the discharge had only fine particles of the egg-albu-
min, but later these were much larger. These unchanged particles
are significant, for they indicate that the acid has been secreted more
rapidly than it could unite with the compact coagulum of the egg-
albumin.® This failure of the acid to unite with albumin as soon as
secreted brings about the same condition that prevails when carbo-
hydrates are fed, — there is an early appearance of free acid in the
stomach. London and Sulima report large amounts of free hydro-
chloric acid in the chyme of coagulated egg-white.* Moritz, Tobler,
and Lang, on the other hand, declare that although the chyme of
beef and fibrin is acid in reaction, it does not contain free hydro-
chloric acid.*. This difference in the rapidity of union with the acid
as it is secreted would account for the difference in the rate of dis-
charge of these proteids. The slow union of acid with coagulated
egg-white and the resultant early appearance of free acid in the
stomach explains the rapid departure of this food.

1 Lonpon and SuLtma: Zeitschrift fiir physiologische Chemie, 1905, xlvi,
Pp. 233-

2 PawLow: Loc. cit., p. 96.

8 The very rapid exit of a rice preparation moistened with NaHCO, (which
hinders gastric secretion) may be similarly explained (see p. 294). .

* LONDON and SuLIMA: Loc. cit., pp. 215, 220.

5 See FERMI: Loc. cét., p. 59-

6 LONDON and SuULIMA: Loc. cit., p. 212.

7 Moritz: Loc. cit., p. 569; ToBLER: Loc. cit., p. 197; LANG: Loc. cit.,

p. 240.

The Acid Control of the Pylorus. 315

The discharge of fats. —In the research on the passage of the
different food-stuffs from the stomach, X-ray observations showed
that fats remain long in the stomach. The discharge begins slowly
and continues at about the rate at which the fat leaves the small
intestine by absorption and by passage into the large intestine.
These results of X-ray examination are in entire accord with those
of Zawilski,! Frank,? Matthes and Marquadsen,? Boldireff,* Carnot
and Chassevant,® and Levites,® who used various other methods.

In attempting to understand the discharge of fats, it is necessary,
first, to consider their effects both in the stomach and in the duo-
denum. Observations by Lobassow‘ and Fermi ® have shown that
fat in the stomach does not stimulate the flow of gastric juice. On
the other hand, according to Lintwarew,”® fat in the duodenum, like
acid, may cause a prolonged checking of the gastric discharge.

As already shown, there is reason to believe that the taking of
food not causing a flow of gastric juice is accompanied by a state
of low pyloric tonus. This condition seems to be as true of fats
as of water and fluid egg-white, for Boldireff has reported that
when fats are fed in considerable amount, a mixture of pancreatic
juice, bile, and intestinal secretion flows back into the stomach.'®
This result could not occur unless at times the pyloric sphincter
were in a relaxed state, and unless at times the pressure in the
stomach were less than that in the duodenum. In this connection
it is of interest to recall that of the three food-stuffs fats produce
the slowest rate of gastric peristalsis 4 and commonly the weakest
(1. e., the shallowest) waves.

As noted earlier in this paper, fats differ from carbohydrates and
proteids in very seldom constituting the chief elements of a diet.

1 ZAWILSKI: Arbeiten aus dem physiologischen Anstalt zur Leipzig, 1876,
Bp. 156.

2 FRANK: Archiv fiir Physiologie, 1892, p. 501.

3 MATTHES and MARQUADSEN: Verhandlungen des Congresses fiir innere
Medicin, 1898, xvi, p. 364.

* BOLDIREFF: Centralblatt tiir Physiologie, 1904, xviii, p. 457.

5 CaRNOT and CHASSEVANT: Loc. cét., p. 866.

6 LEvITES: Zeitschrift tiir physiologische Chemie, 1906, xlix, p. 276.

See PAWLow: Loc. cit., p. 97.

8 FERMI: Loc. cit., p. 76.

9 LINTWAREW: Biochemisches Centralblatt, 1903, 1, p. 96.

10 BOLDIREFF: Loc. cit., p. 458.

11 CANNON: This journal, 1904, xii, p. 392.

316 W. B. Cannon.

They differ also in not arousing gastric secretion. They are fur-
ther peculiar in acting by themselves in the duodenum, not only
to-inhibit gastric evacuation, but also to stimulate the flow of pan-
creatic juice... Clearly fats do not require the secretion of gastric
juice for changes in the stomach, or for the control of their exit
into the intestine, or for the stimulation of a pancreatic secretion
specially favorable to their digestion.

It may be that fats have also a special relation to the pyloric
mechanism. But the alternative possibility of an acid control, even
when fats alone are fed, should not be overlooked. Fatty acid may
be set free in considerable amount in the stomach by gastric lipase,
if the fat is fed as an emulsion.” A separation of fatty acid also
occurs when in the early stages of fat digestion pancreatic juice
enters the stomach.® If the failure of fats to excite gastric secretion
would place them at first with fluid egg-white as substances readily
passed through an easily opened pylorus, the later development of
acid in the fats contained in the stomach might cause them to con-
trol their own discharge like other foods developing an acid reaction
of the gastric contents.

In the duodenum it is noteworthy that fats are changed with
an effect quite unlike that of the other food-stuffs. Fats cause the
pancreatic juice to flow, but the pancreatic juice, instead of diminish-
ing the acidity of the duodenal contents, increases the acidity by
separating a still greater amount of fatty acid* Even when dis-
solved in bile the fatty acids give the solution an acid reaction.®
To this increasing acidity of the contents of the upper intestine,
and also to the weak and sluggish gastric peristalsis which fats
evoke, may reasonably be attributed the fact that fats pass from the
stomach only as fast as they are absorbed or carried into the large
intestine.

Pathological cases. — It may be urged that certain cases of patho-
logical secretion of gastric juice — cases of hypo- and hyperchlor-
hydria, and of achylia gastrica, for example — do not yield results
accordant with the acid control of the pylorus. Thus in achylia

1 Dottnsky: Archives des sciences biologiques, St. Petersburg, 1895, iii,
p- 424.

2 VOLHARD: Zeitschrift fiir klinische Medicin, 1go1, xlii, p. 429.

® See-LEVITES: Lye. G77, p-.276:

4 See LEVITES: Loc. ctt., p. 279.

5 Moore and Rockwoop: Journal of physiology, 1897, xxi, p. 64.

The Acid Control of the Pylorus. 317

gastrica the absence of acid does not lead to a retention of food
in the stomach. But gastric evacuation in the absence of an acid
reaction is only one problem to be settled in achylia gastrica, —
pancreatic secretion without the natural acid stimulus in the duo-
denum must also be investigated and explained. As shown by
these examples, the discovery of natural relations at once reveals
the character of disturbed relations. If in spite of disturbed rela-
tions the processes concerned continue to be serviceable to the
organism as a whole, an adaptation to the new conditions must have
occurred. The ability of organs to adapt their functioning gradu-
ally to pathological states is well known in many instances. This
adaptation, however, must be studied by itself as a special subject.
Thus, after the normal physiology of the pylorus is made clear, it
becomes of interest to know to what extent and in what manner dis-
turbances in the stomach and duodenum are attended by changes
in pyloric reflexes which are compensatory. Undoubtedly other
factors may modify the usual mechanism. Already it has been
shown that the pylorus may remain tightly closed against persistent
peristalsis for five or six hours, if serious injury is done to the duo-
denum.! But this is a pathological state. It is with the normal
physiology of the pylorus that the present research is primarily
concerned.

In the foregoing consideration of evidence not in accord with the
acid control of the pylorus, it has been necessary to assume that
the ingestion of material not stimulating a flow of gastric juice
is attended by a weaker tonus of the pyloric sphincter than that
- prevailing when food is eaten with relish. It has also been neces-
sary to assume that if acid is secreted on proteid more rapidly than
the proteid can change to acid proteid, the free acid will then cause
a rapid emergence of the food from the stomach. For both these
assumptions evidence is presented. If these assumptions are granted,
the conclusion remains valid that acid in the stomach opens, and in
the duodenum closes, the pyloric sphincter. .

INFERENTIAL SUPPORT FOR THE ACID CONTROL FROM OTHER
PROCESSES IN THE PyLoriIc REGION.

The evidence that the appearance of acid at the pylorus is the
signal for the relaxation of the pyloric sphincter receives strong

1 CANNON and Murpuy: Annals of surgery, 1906, xliii, p. 515.

318 W. B. Cannon.

support from the relation of this process to other processes in the
automatic mechanisms of the stomach and duodenum.

As is well known, the acid of the gastric juice is secreted only
in the cardiac end of the stomach. Edkins has reported experiments
showing that the condition for the continuance of gastric secretion
after the initial psychic secretion,.and thus the condition for the
continuance of gastric digestion, lies in a chemical stimulation of
the glands through the blood stream.t| The chemical stimulant
is produced, not by the mucous membrane of the cardiac end of the
stomach, but by that of the pyloric end. It is not produced by the
mucosa alone, but by the action upon the mucosa of acid, peptone,
or sugar solutions. Evidently on this basis, if the pylorus opened
as soon as food entered the stomach, the food would pass from the
antrum without presenting to the mucosa of the antrum the acid
requisite for maintaining the secretion of gastric juice. That the
processes in the stomach may advance in an orderly manner, there-
fore, it 1s necessary that the food be retained until the BoEHey in
the antrum is acid.

The processes in the duodenum likewise require that food shall
be checked at the pylorus until acid in reaction. If the food were
allowed to depart before becoming acid,? it could not stimulate
chemically the duodenal reflex. The pylorus consequently would
not be held closed, and the upper small intestine would be crowded
full of food through an uncontrolled pyloric sphincter. Furthermore,
the chyme, unless held back until acid, would not, on entering the
duodenum, excite the flow of pancreatic juice and bile. Thus, if
the pylorus relaxed at the approach of the first peristaltic wave
(after meat had been fed, for example), the food would not only
go out from the antrum wholly undigested by gastric juice, but
would bear no provision for being digested by the pancreatic juice.
In order that the pancreatic juice may be caused to flow and may
have time to become mixed thoroughly with the chyme without
being overwhelmed by fresh discharges from the stomach, food
must be retained in the antrum until acid in reaction.

If it is granted that the antrum contents must be acid before
being permitted to pass the pylorus, it is of interest to note how
favorably the stomach is arranged for the utilization of its secretions.
Evidence has previously been presented that in order to open the

1 EpKINS: Journal of physiology, 1906, xxxiv, p. 133.
2 In this discussion the somewhat variant case of the fats is not regarded.

The Acid Gane of the Pylorus. 319

sphincter the acid must be at the pylorus. Clearly, if the antrum
secreted acid, the acid would at once open the pylorus and let out
the food (meat, for example) before the gastric juice had had oppor-
tunity to digest it. But the antrum, in which the acid stimulus acts,
does not itself secrete acid. The acid and the food with an acid
reaction must be brought from the cardiac end of the stomach and
thoroughly mixed with the contents of the antrum before the pylorus
relaxes. The necessity of importing the acid into the antrum insures
a thorough mixing of the food with the gastric juice before the
food departs, and provides time for gastric digestion.

The acid control of the pylorus is therefore an arrangement
whereby the food is held in the stomach until provision is made for
_ the continuance of gastric secretion, until the gastric juice has
had time to act, and until the food can bear with it the acid needed
for processes in the duodenum. In the duodenum the acid chyme
stimulates the flow of pancreatic juice and bile, and holds the pylorus
closed until this chyme has been thoroughly mixed with these di-
gestive fluids. This thorough mixing stops gastric digestion, in-
jurious to the action of the pancreatic ferments, by neutralizing
the acid. As the acid is neutralized, the stimulus holding the pylorus
closed is weakened, and then the acid in the stomach is again effec-
tive in causing the pylorus to open.

The acid control of the pylorus here described receives further
inferential support from the fact that the acid affects the pyloric
sphincter just as a stimulus in the intestine affects the intestinal
muscle. Bayliss and Starling! have shown that in the intestinal
wall is a local reflex, such that a stimulus causes a contraction above
the stimulated point and a relaxation below. The action of acid on
the two sides of the pylorus is in exact agreement with this so-called
“law of the intestine’’; the acid when above causes a relaxation
of the sphincter which is below, and the acid when below causes a
contraction of the sphincter which is above. As already noted (see
footnote, p. 311), the cardia also obeys this law. It is not impossible
that throughout the portion of the alimentary canal consisting of
smooth muscle, this reflex is the mechanism for orderly action. ~

SuM MARY.

The stomach is emptied progressively during the course of gastric
digestion; by occasional discharges through the pylorus.

1 BAyLiss and STARLING: Journal of physiology, 1899, xxiv, p. 142.

320 W. B. Cannon.

Mechanical agencies, either in the stomach or in the intestine, play
an unimportant part in controlling gastric evacuation; for (1) the
occasional discharges through the pylorus are not the result of
momentarily deepened peristalsis, and (2) the upper intestine in
normal conditions is not sufficiently filled or distended to check the
outgo from the stomach.

Observations on chemical conditions in the stomach have hitherto
been defective for judging the mechanism of the pylorus, because
the food given at different times has not been identical in amount
nor uniform in consistency, and the difference in the chemical re-
action of the two ends of the stomach has not been distinguished.
Furthermore, these studies, like the observations of Hirsch, Serd-
jukow, and Tobler, that acid in the duodenum checks gastric dis-
charge, have failed to distinguish between two factors always
concerned in the passage of food through the pylorus.

The two factors are (1) the pressure at the pylorus due to re-
current peristalsis, and (2) the action of the pyloric sphincter.
The X-ray method shows that during gastric digestion peristaltic
Waves are passing, not occasionally, but continuously. Since the
discharge from the stomach is not continuous, but occasional, the
control must rest with the pyloric sphincter.

It is necessary to explain the intermittent closure of the pylorus;
the usual closure, and the occasional opening. It is also necessary
to explain why, for example, carbohydrates begin to leave the
stomach early and depart rapidly, whereas proteids of the same
amount and consistency begin to leave the stomach only after some
time, and then depart slowly.

These facts can be explained on the theory that acid in the antrum
opens the pylorus, acid in the duodenum closes it. Because the
acid in the duodenum is soon neutralized, the closure of the pylorus
is intermittent.

That acid in the antrum signals the opening of the pylorus is
indicated by the following evidence: (1) moistening carbohy-
drates with NaHCOsg retards their normally rapid exit from the
stomach; (2) feeding proteids as acid proteids remarkably hastens
their normally slow exit; (3) observations through a fistula in
the antrum show that an acid reaction closely precedes the initial
passage of food through the pylorus, ‘that the introduction of acid
causes pyloric opening, and that delaying the acid reaction causes
retention of the food in the stomach in spite of strong peristalsis;

The Acid Control of the Pylorus. 321

(4) when the stomach is excised and kept alive in oxygenated
Ringer’s solution, the pylorus is opened by acid on the gastric side.

That acid in the duodenum keeps the pylorus closed is shown by
the following evidence: (1) acid in the duodenum inhibits gastric
discharge (observations of Hirsch, Serdjukow, and Tobler), and as
shown above, the effect is not due to stoppage of peristalsis, but
to closure of the pylorus; (2) the stomach empties more slowly than
normally when the tying of pancreatic and bile ducts prevents
alkaline fluids from neutralizing the acid chyme in the duodenum;
(3) the discharge of proteid becomes rapid if the pylorus is sutured
to the intestine below the duodenum, or if a ring is cut through the
muscular coats immediately beyond the pylorus. The effect from
the duodenum is thus a local reflex mediated, like movements of the
small intestine, by Auerbach’s plexus.

Evidence for the acid control of the pylorus is also found in
the application of the theory to previous observations on gastric
discharge. Proteids leave the stomach only after considerable delay,
and then emerge slowly; this fact can be explained (1) by the slow
development of a marked acid reaction in the stomach due to the
preliminary union of acid with proteid, and (2) by the large amount
of acid borne into the duodenum by proteid chyme. Carbohydrates
leave the stomach early and rapidly, — a result to be expected, since
the acid secreted upon them does not unite with them, and is at once
present to open the pylorus. The peculiar rates of discharge of
combinations of these food-stuffs are also readily explained on the
theory above stated. This fitness of the theory to explain established
facts gives it additional support.

The rapid exit of water through the pylorus without change
of reaction, and the similar rapid exit of raw egg-white — facts
not in accord with the acid control — are accounted for on the as-
sumption that conditions not favoring gastric secretion are attended
by a low pyloric tonus, and vice versa. Reasons are given for this
assumption. The rapid exit of coagulated egg-white, exceptional
among proteids, is explained by its slow union with the secreted
acid. Fats leave the stomach very slowly. Like water and raw
egg-white, they do not stimulate gastric secretion; but they may
become acid in the stomach by the separation of fatty acid. Their
very slow exit can probably be accounted for largely by the fact
that when fats are fed, the pancreatic juice, instead of decreasing,
increases the acidity of the duodenal contents.

322 W. B. Cannon.

Strong support for the acid control is found in its relation to
other processes in the stomach and duodenum. The retention of
food in the stomach until the antrum contents are acid is necessary
(1) for the proper continuance of gastric secretion and (2) for
the accomplishment of gastric digestion. Such retention is also
necessary in order (3) that the chyme emerging into the duodenum
may bear with it the acid required to cause the flow of pancreatic
juice and bile, and (4) that the pylorus may be held closed until
these important secretions are thoroughly mixed with the acid

chyme.
The facts presented bring the pyloric mechanism under the “ law
of the intestine,’ —the acid when above (in the antrum) causes

a relaxation of the sphincter which is below, and the acid when
below (in the duodenum) causes a contraction of the sphincter
which is above.

A FURTHER STUDY OF THE ACTION OF MAGNESIUM
SULPHATE ON THE HEART.

By WM. peEB. MACNIDER anp S. A. MATTHEWS.

[From the Laboratory of Experimental Therapeutics, University of Chicago.|

ie a previous number of this Journal! it was shown by one of us
that the salts of magnesium, either the sulphate or the chloride,
when injected into the circulation of dogs, in doses of from 2 to 4 c.c.
of a 2 m/1 solution, had a very depressing effect upon the heart,
characterized by a progressive slowing of the heart rhythm and a
simultaneous decrease in the amplitude of the contractions, which
soon led to a complete standstill from which the heart could be re-
covered by artificial stimulation.”

In this investigation we have attempted to study more closely the
condition of the heart poisoned by magnesium sulphate, and, if pos-
sible, to arrive at some conclusion as to how and where it affects the
heart tissue.

The animals used in this study were medium-sized dogs (6-8 kilos).
The magnesium sulphate solution (#/1 MgSO,) was injected into the
saphenous vein, and the tracings were taken direct from the heart
by means of a Roy-Adams myocardiograph ; ® simultaneous tracings
being taken from the right auricle and right ventricle.

It was found that doses of from 8 to I2c.c. of an m/t MgSO, solution,
if injected in the space of from-two to three minutes, were sufficient to
bring the heart to a standstill in the condition already described.

A study of the heart thus weakened by the administration of mag-
nesium sulphate showed, even after it had been brought to a complete
standstill, that it possessed the following properties :

(1) Light induction shocks about one per second sent through the
right ventricle caused contractions, which followed the law of con-

1 Matruews, S. A.,and D.C. Jackson: This journal, 1907, xix, p. 5.

2 MATTHEWS, S. A., and D. C. Jackson: /dzd, 1907, xix, p. 9.

8 Roy and ADAMI: Philosophical transactions, 1853, clxxxili, p. 207.

525

324 Wm. deB. Macnider and S. A. Matthews.

traction of heart muscle. The contractions which were thus induced
in the ventricle were taken up by the auricle, — apparently propagated
to the auricle. After the heart had been brought to a complete rest
it generally required from five to eight minutes of continuous stimu-
lation before it resumed its spontaneous rhythm, but after the spon-
taneous beat had once been
established the contractions
were even more complete than
before the administration of
magnesium sulphate (Fig. 1).
(2) After the heart had been
somewhat slowed, the ampli-
tude of the contractions being
very much diminished, and the
UN Ai hth heart dilated tosuch a degreeas
Ne! to place it almost on the verge of
BIGUR Se ng seat 1 eas aaa tear standstill, stimulation of the
One half the original size. At beginning of
this tracing 8 c.c. m/1 MgSO, was injeeted. accelerators had the same bene-
Shows complete standstill and effect of ac-_ ficial effect in re-establishing
celerator stimulation both on auricle and the spontaneous beat as did
ventricle. ; ; ¢
stimulation of the heart di-
rectly, with the exception that when the heart muscle was stimulated
directly, it responded immediately, while stimulation of the accelera-
tors induced contractions after a certain latent period.! Stimulation
of the motor nerves had the same effect on the irritability of the heart
muscle as direct stimulation. The influence of the accelerator stim-
ulation continued for some time after cessation of the stimuli; not
only the rate but the force of the heart beat was augmented.” The
beat having once been obtained by a certain strength of electrical
stimulus, the current may be gradually decreased and yet the heart
muscle will respond by a tontraction to each stimulus. In fact, if the
heart responds to a certain strength of current by a normal contrac-
tion, unless the current be gradually weakened there is danger of
throwing the heart into delirium cordis. This is in accord with other
well-known observations that electric stimulation of the heart muscle
has a tendency to increase its irritability.®

1 Hunt, REID: This journal, 1899, ii, p. 469.

2 Hunt, REID: Jdzd.

8 CusHny and MATTHEWS: Journal of physiology, 1897, xxi, p. 213. W. H,
GASKELL: SCHAFER’S Text Book of Physiology, 1898, ii, p. 217.

Lhe Action of Magnesium Sulphate on the Heart. 325

(3) After the heart has been brought under the influence of mag-
nesium sulphate, section of the vagi tended to increase the diminished
tone, and thus to slightly antagonize the magnesium action (Fig. 2).
It was noticed also that as the magnesium suiphate action increased,
the inhibitory action of the vagi lessened, but stimulation of these
nerves always gave some re-
sponse until just before the | |
heart action ceased.) At this
point vagus stimulation would
first show some slight inhibi-
tion followed by quite a marked |
acceleration. This would indi-
cate, as is probably the case,
the presence of accelerator
fibres in the vagi. The dimin-

Stein maul

like result was obtained after
FIGURE 2.— Myocardiograph tracing. Dog.

the administration of 2 mg.
atropine. Two thirds the original size. Upper ven-

As soon as the increased ac- tricle, lower auricle. Vagi sectioned at YY,
celerator influence manifested Increase in contraction resembling accelera-
. : tor stimulation.
itself, whether it was due to
electric stimulation, to the use of certain chemicals, or to the throw-
ing out of the vagi by sectioning or by atropine, the heart would
withstand from two to three times the ordinary lethal dose of magne-
sium sulphate.

Electric stimulation of the auricle, the current being passed be-
tween the points of contact of a Roy-Adams myocardiograph, always
caused a normal contraction which was generally propagated to the
ventricle apparently in a normal manner. In two cases there were
partial blocks, in one of which the ventricle responded to every third
beat of the auricle and in the other to every fourth auricular contrac-
tion. In another instance there was a complete block between the
auricle and the ventricle; the auricle responded, but the ventricle
remained inactive and intensely dilated.

While in all probability there is more or less of a delayed passage

|

|

1 It was stated by one of us in the previous article, already referred to, that
the vagi remained unaffected. This is not the case, as stated in this paper.

326 Wm. deB. Macnider and S. A. Matthews.

of impulses from the auricle to the ventricle in all stages of magne-
sium sulphate action upon the heart, the partial and complete block-
ing only became apparent late in the experiment, and after the animal
had been subjected to large doses. The blocks would apparently in-
dicate a lowering of the conducting power of the elements in the
auriculo-ventricular junction, namely, the “‘neuro-muscular bundles
(is); 2

It will readily be seen from the foregoing description that the heart
is placed in a rather interesting condition in that it will respond to
all external stimuli with the exception of a lowered vagus irritability,
but is bereft of its power to originate a beat.

At the present time there are two theories as to the origin and
maintenance of the heart beat. Each has some supporters. One
school maintains that the heart, consisting as it does of contractile
tissue and therefore bearing an analogy to other tissues of this type,
can be thrown into activity only when the normal physiological stim-
ulus reaches it through the intervention of some nervous mechanism.

The other school insists that the heart muscle has the distinguish-
ing property of automatic rhythmicity independent of any nervous
element, and that the nervous mechanism is for the purpose of regu-
lating and not inaugurating a beat.

Without any attempt at a solution of these theories, we can say
that after the heart has been robbed of automaticity by magnesium
sulphate and rendered quiescent, it can be made to beat rhythmically
and apparently in a normal manner by stimuli reaching it through a
nervous mechanism (accelerator nerves). This would suggest that
impulses capable of originating auricular contractions might normally
reach the heart over the accelerator nerves.

In considering the probable points upon which magnesium sulphate
might act to bring about these changes in the heart we hardly need
to consider its central action, although a central depression of the
accelerator centre might give rise to a lessened tone of the heart.
But if the vagus centre should also be depressed, as it probably is,
an apparent increase of accelerator action might appear. Also vagus
depression either centrally or peripherally would naturally tend to
increase the heart tone. Although the accelerators are active, this
does not preclude the possibility that magnesium has some effect
through this medium, for these nerves might be made to functionate
in a normal manner by strong electrical stimulation; yet at the same

1 ERLANGER, J.: This journal, 1906, xv, p. 170.

The Action of Magnestum Sulphate on the Heart. 327

time they might be so depressed that the normal physiological stim-
ulus would have little or no effect. This assumption is strengthened
by the known fact that removal of the vagus action by sectioning or
by the use of atropine increases the functional activity of the accel-
erators and will enable them to overcome in some degree the magne-
sium sulphate depression. Also direct stimulation (electrical) of the
heart muscle increases its irritability and contractility, and renders it
more responsive to impulses reaching it over the accelerators and
seems to enable it to more readily resume its normal rhythm after
magnesium sulphate poisoning. Further, certain chemicals, as cal-
cium chloride, barium chloride, and strophanthus, all of which have
as their predominate action an increased tone and irritability of the
heart, exert such a powerful action on the heart muscle or on its
motor nerves that a tracing may be obtained after magnesium sul-
phate poisoning almost identical to one produced when the accelera-
tors are stimulated electrically. After the use of these drugs, just as
after electrical stimulation, the heart will continue to beat in a normal
manner, but with increased efficiency, even when under the influence
of from four to five times the lethal dose of magnesium sulphate.

The logical conclusion arrived at from such reasoning would seem
to be that magnesium depresses the nervous mechanism in the heart
(both accelerator and inhibitory, the latter more than the former) to
such an extent that it will not transmit to the contractile tissue im-
pulses of the same degree as are sent in by the usual physiological
stimulus, whatever that may be, but that it will transmit the stronger
electrical stimulus or the normal motor impulse after the heart muscle
has been rendered more irritable.

This conclusion necessitates the conception that the efficiency of a
stimulus depends not only upon its strength, but in a large measure
upon the receptive power of the object stimulated, — in this case the
muscle fibres.

_ The physiology of the intrinsic nervous mechanism of the heart is
so imperfectly understood that it is impossible to say just what réle
it plays in maintaining or restoring the tone of the heart muscle. As
before stated, stimulation of the accelerator nerves will restore the
heart tone and often cause it to beat after it has been brought to a
standstill by magnesium sulphate. Likewise direct stimulation of the
heart muscle will do the same. Our experiments seem to suggest a
dual action. The inhibitory (vagi) nerves are evidently depressed.
The accelerator nerves may be and probably are depressed so as to

328 Wm. deB. Macnider and S. A. Matthews.

destroy, in a measure, their tone-giving influence over the heart.
The mechanism by which impulses are propagated from auricle to
ventricle is evidently weakened and may in some cases be wholly
paralyzed. Finally, the muscle itself may be rendered non-receptive
to the influence of stimuli, as in most cases of poisoning certain tis-

sues seem to be more vulner-

able than others. Here the

| depression seems to be more

! or less localized in the struc-
tures which transmit impulses

from the auricle to the ven-
tricle. In several cases we
noticed when the heart had
come to a complete standstill,

| | both the auricle and ventricle
il were irritable to electric stim-
| ulation, yet the auricular con-
FiGuRE 3.—Myocardiograph tracing. Dog. Two traction was not propagated
thirds the original size. Upper ventricle,lower to the ventricle (Fig. 2).
auricle. Shows complete block. Auricle stim- While this 16 evidently one
ulated electrically. Induction shocks, . ; :
point upon which magnesium
sulphate acts, the whole phenomenon of its action cannot be explained
in terms of ‘‘heart-block.” The standstill of both auricle and ventricle
can only be accounted for on the assumption that there is present in
the cardiac tissue some element capable of being stimulated, and that
such stimulation is transmitted to the muscle, resulting in a contrac-
tion. In the heart of the Limulus, Carlson! found such an element
1 A. J. CARLSON: This journal, 1904, xii, p. 72.
Note. Barium chloride used in the dose of 1-2 c.c. of a 7/50 solution acts very
powerfully on the heart, causing violent and complete contractions.
Calcium chloride in 5 c.c. doses of wz/8 solution has a like effect, but to a less

degree, and seems to prevent the excessive action of the barium. For this reason
we have used in these experiments the following combination :

Calcium chloride 77/8 . . 230 C.c.
Barium chloride #/50. . 20 C.Cc.
250 C.C.

The average dose of this combination was Io c.c.

Strophanthin, while not so efficient as the above combination in antagonizing
the magnesium sulphate action, has a very decided effect, causing forcible contrac-
tions and relieving dilatation. The average dose was I-3 mg.

After the use of barium and calcium and to a less extent strophanthin, very
large doses of magnesium sulphate may be given (15-20 c.c.) with little damage
to the heart.

The Action of Magnesium Sulphate on the Heart. 329

in the median nerve cord. Should such a nervous mechanism exist
in the mammalian heart, and magnesium sulphate have a like action
upon it, as it has upon the tissues which propagate the auricular im-
pulse to the ventricle, it would account for the lost automaticity of
the heart while the irritability of the heart muscle still remains.

Whether such chemicals as calcium chloride and barium chloride,
or strophanthus, which antagonize magnesium sulphate, do so by
acting on the structures which have to do with the propagation of
the cardiac impulses or by increasing the irritability of the cardiac
muscle, cannot be stated definitely. We can say, however, they re-
lieve the partial heart block produced by the magnesium sulphate
and give rise to a heart tracing similar to stimulation of the
accelerators.

ON THE SWELLING -OF- IBRIN:

By MARTIN H. FISCHER anp GERTRUDE MOORE:
[From the Frank B. Yoakum Laboratory of the Oakland School of Medicine.)

INTRODUCTION.

| Base forces active in the absorption and secretion of water by
animal and vegetable cells have from the earliest periods of
modern physiology been the object of active study and research.
This is scarcely to be regarded with surprise when the multitude
of physiological reactions in which a storage or a movement of
water plays a leading role is considered. We need, by way of illus-
tration, cite only the maintenance of turgor in plant cells, the often
enormous pressures exerted by plants during growth, many of the
phenomena of glandular activity, and cedema.* Following the
original attempts to explain the observed phenomena through spe-
cific properties of living protoplasm there arose those in which fil-
tration and (in the higher forms) the pressure of circulating liquids
were assumed to play a predominant part. In the course of time
the inadequacy of these physical agencies became apparent, so that
the more modern attempts to explain the migration of water through
changes in osmotic pressure — especially as rendered more gener-
ally applicable to biological phenomena through the fundamental
work of Overton and Meyer on the lipoids — came as a welcome
aid to those workers who seek the solution of physiological prob-
lems along purely physical and chemical lines. But the impossibility
of explaining more than a comparatively small part of such phe-
nomena as have been alluded to in this paragraph through the laws
of osmotic pressure has become more and more apparent as careful
physico-chemical observations on living matter have multiplied. It
would seem in order, therefore, to cast about for further forces

1 See FISCHER, MARTIN H.: The physiology of alimentation, New York,
1907, pp. 182-187 and 267-269.
33°

On the Swelling of Frbrin. 331

)

capable of influencing the migration of water which we may im-
agine active in protoplasm. Of such, that which we must call for
the present and until it is analyzed physico-chemically the variable
affmity of colloids for water, seems to stand in the very front rank.
Toward a physico-chemical analysis of this affinity of colloids for
water, great strides have already been made, and when it is com-
pleted we shall, no doubt, be able to deduce from a few simple prop-
erties of a colloid its behavior toward water. At present we have
still to content ourselves with studies of isolated colloids and their
behavior in the presence of water under various external conditions.

The following paragraphs deal with the affinity of beef fibrin

for water as influenced by various acids, various salts, and a few
non-electrolytes.

METHOpDs.

As our experiments are comparative in character, it was neces-
sary to have at our disposal sufficient quantities of a uniform
material. To this end we employed the dried fibrin from blood,
prepared by Merck, which we treated as follows. After powdering
in a mortar the fibrin was allowed to swell in a 1/10 normal hy-
drochloric acid solution. When the maximum amount of swelling
had been reached, the acid was poured off and distilled water sub-
stituted. This was frequently changed until a neutral reaction to
phenolphthalein was obtained. To facilitate the washing, an air
current was drawn through the wash bottle in such a way as to
keep the fibrin in constant motion. After being allowed to dry on
filter paper the fibrin was desiccated in a thermostat at 37° C. This
dried fibrin was ground a second time in a mortar, and weighed
amounts of the resulting powder were used in the experiments.
Unless otherwise noted, the remarks which follow refer only to
fibrin prepared in this way. This fibrin is fairly, but not entirely,
free from salts, and swells under proper conditions into a colorless,
jelly-like mass.

Our stock acids were one fifth normal (phenolphthalein being used
as an indicator). From these were prepared the dilutions men-
tioned in the experiments.

Our experiments consisted in introducing a weighed quantity
(0.25 gm.) of powdered fibrin into a measured quantity (20 c.c.)
of various solutions contained in test tubes having a uniform diame-
ter (1.7 cm.), and-‘measuring the height of the swelled fibrin column

432 Martin HH]. Fischer and Gertrude Moore.

in each of the tubes. After a few preliminary experiments we
found these quantities of fibrin and solution to give the most satis-
factory results. While weighing the fibrin would seem to promise
more accurate results than measuring its height, we found that the
latter method, except in a few special instances, was less subject
to experimental error than the former. The maximum amount of
swelling in any of the solutions is usually attained within an hour.
When an experiment is long continued, the fibrin slowly goes into
solution, and the height of the column may in consequence decrease.

The following paragraphs give a survey of the experiments we
have thus far performed.

EXPERIMENTS.

1. Fibrin, when prepared as above, swells more in any acid solu-
tion than it does in distilled water, but the amount of this swelling
is greater in some acids than in others. The following table shows
the order in which the acids are effective in this regard, and indi-
cates at the same time the striking difference in the height of the
fibrin columns between the end members of the series.

TABLE, i.

Abide Height of fibrin column in mm. after

All 1/10 normal. 30 min. 4 hrs.

Hydrochloric . . . 28 28
Phosphoric .

Lactic .

Formic

Oxalic .

Nitric .

Acetic .

Citric):

Sulphuric

Water.

A mere glance at the table is sufficient to show that we are not
dealing with the simple effect of hydrogen ions as determined by

Ox the Swelling of Fibrin. 333

their relative concentrations in these acids, for while a “ strong”

acid (hydrochloric) stands first and another (sulphuric) last in the
series, several “ weak” organic acids are found between these
extremes.

The arrangement of the acids in the series (Table I) remains
the same no matter what the concentration of the acids. We have
employed the above acids in concentrations ranging from one fifth
normal to one two-hundredth normal. The arrangement in the
series with all of these is the same. While the differences between
the end members of the series is very great (except in the more
dilute solutions), that between members following each other may
be so slight (see Table II) that it falls within the limits of ex-
perimental error. We were not surprised in consequence to find
neighboring members in the series occasionally changing places.
Thus, formic would at times stand ahead of lactic, or phosphoric
below it.

The amount of the swelling 1s determined by the concentration
of the acid, and is the greater, the greater the concentration of the
acid. This can be seen in the following table in which is indicated
the maximum amount of swelling (after about six hours) in acids
of different concentrations.

TABLE II.

Acid. 1/10 N. | 1/20 N.|1/40 N.| 1/80 N. 1/160 N.

Hydrochloric.
Tactic. .
Formic.

@xalic .
swelling
as in water ;
in water
Acetic . (?)
Citric

Sulphuric .

9
7
7
6
INERICE Geatier cy Gs 9(?) Sel lnsos same as
8(?
6
6
6

Water .

2. We next studied the effect of different salts on the swelling
of fibrin in acid solutions. With the exceptions to be noted, the
addition of any salt to a pure acid solution decreases the amount

334 Martin H. Fischer and Gertrude Moore.

of water absorbed by fibrin in that solution. The exceptions are
formed by those salts which are capable of reacting with the acids
used in the experiments. The addition of barium chloride to a
sulphuric acid solution does not decrease, but increases the amount
of swelling of the fibrin. This is because barium sulphate is pre-
cipitated, while hydrochloric acid is produced, and fibrin swells
more in a hydrochloric acid solution a in a sulphuric acid solu-
tion of the same concentration.

The greater the concentration of the salt in an acid solution the
less does fibrin swell in that solution. Table III illustrates this
statement and what has been said in the previous paragraph:

TABECE tit.

0.25 gm. fibrin in each tube : 2 | (NH4)eSO,4 KI

30 c.c. 1/10 N. hydrochloric acid.

15 c.c. 1/5 N. hydrochloric acid
+15 c.c. 1/2 M. salt solution

15 c.c. 1/5 N. hydrochloric acid
+15 c.c. 1/4 M. salt solution .

15

. 1/5 N. hydrochloric acid
c

Gc
+15 c.c. 1/8 M. salt solution

15 c.c. 1/5 N. hydrochloric acid
+15 C.c. 1/16 M. salt solution. . 5c 10 (?) 10 (?)

15 c.c. 1/5 N. hydrochloric acid
+ 15 cc. 1/32 M. salt solution. . sc 9

The effect of any salt upon the swelling of fibrin in an acid solu-
tion seems to be made up of the sum of the effects of its constituent
ions. To determine this we studied the effect of different equimo-
lecular salt solutions having either a common cation or a common
anion upon the swelling of fibrin in hydrochloric acid solutions of
different concentrations. With any given base the general arrange-
ment of the anions is the same; with any given anion that of the
cations is the same. While the difference between the end mem-
bers of such two series is very marked (5 to 12 mm. of fibrin
column in our experiments), that between members following each
other may be only a fraction of a millimetre. As such an amount
is within the limits of experimental error, it is not improbable that
the placing of any individual ion in the following two series, which

On the Swelling of Frbrin. 335

we have constructed, may not be entirely correct, but the general
grouping is. Anions arrange themselves in the following order.
Fibrin will swell most in an acid solution in which the first named
is present, and progressively less if one of the succeeding anions
is present:

Chloride, bromide, nitrate, acetate, tartrate, citrate, sulphate,
todide, ferrocyanide, sulphocyanate.

Cations arrange themselves in the following order: Fibrin
swells most in an acid solution if the first named is present, and
progressively less if one of the succeeding members in the series
is present:

Potassium, ammonium, sodium, calcium, magnesium, strontium,
barium, copper (2), uranium.

If the series of anions given above is compared with the series
of acids (p. 332) arranged according to their ability to make fibrin
swell, it is found that the order of the anions is the same in the
two. This, and the fact that the acids do not arrange themselves
according to the degrees of their electrolytic dissociation, has led
us to suspect that the action of any pure acid on fibrin is an ex-
pression of the concentration of the hydrogen ions minus the effect
of the anion in decreasing the amount of water absorbed.

3. The difference between the amount our ordinary fibrin swells
im distilled water and in pure salt solutions is not great. It is
measurable in our experiments by 3 to 5 mm. A definite difference,
however, exists, though it is impossible from our present series
of experiments to construct accurate tables for the different salts.
It can be said in general that these appear to be very similar to
the two tables of ions given above for the action of salts on the
swelling of fibrin in acid solutions. We have determined that
fibrin swells more in pure potassium solutions than in equimolecular
sodium solutions, and in these more than in equimolecular calcium
solutions. When fibrin is introduced into equimolecular solutions
of salts having a common cation, it is found to swell more in a
chloride than in a nitrate, and more in this than in a sulphate. These
conclusions have been drawn from measurements of the height of
the fibrin column in test tubes containing equal amounts of fibrin
(0.25 gm.) in equal volumes (20.0 c.c.) of equimolecular salt
solutions varying in the different series from 1/1 molecular to
1/16 molecular; and-from weighings of equal amounts of fibrin
(2.0 gm.) after having swelled in equal volumes (100 c.c.) of these
same salt solutions.

336 Martin H. Fischer and Gertrude Moore.

4. When fibrin has swelled to its maximum in a pure acid solu-
tion, it will swell still more if distilled water is substituted for the
acid After 0.25 gm. fibrin had been left for an hour in each of
three test tubes containing 20.0 c.c. 1/10 N. hydrochloric acid,
the protein columns were each 16 mm. in height. One tube was
kept as a control. From the second the hydrochloric acid was
poured off, and after several rinsings water was substituted. From
the third the hydrochloric acid was poured off, and a 1/1o N.
sulphuric acid solution was substituted after the fibrin had first been
rinsed in this fluid. At the end of an hour the fibrin column in
the hydrochloric acid solution still stood at 16 mm.; that in the
distilled water had risen to 25 mm.; that in the sulphuric acid had
contracted to g mm. In-a second similarly arranged experiment
the fibrin columns measured 15, 14, and 16 mm., after standing an
hour in pure hydrochloric acid solutions. Water was substituted
in the first tube, sulphuric acid in the second, while the third was
kept as a control. At the end of an hour in these solutions the
columns measured 26, 9, and 16 mm. respectively.

5. The taking up and giving off of water by fibrin represents,
to a large extent, a reversible process. In the foregoing paragraph.
the shrinkage of fibrin, when sulphuric acid is substituted for hy-
drochloric, has been illustrated. The final height of the protein
column is the same as would have been attained if the dry fibrin
had originally been allowed to swell in a pure sulphuric acid solu-
tion. Similar reversions can be obtained with salts. The fibrin
column described in the previous paragraph, which measured
16 mm. in a pure 1/10 N. hydrochloric acid solution and-25 mm.
in distilled water, shrank to 10 mm. when a mixture of equal parts

1 These remarks refer only to fibrin treated as described in the introduction.
As already stated, this fibrin is fairly free from salts, but not entirely so. When
special pains are taken to still further reduce the amount of salts present in the
fibrin, quite different results are obtained. We allowed some of our ordinary fibrin
to swell a second time in 1/10 N. hydrochloric acid, then washed it twenty-four to
seventy-two hours in distilled water (kept in constant motion through a stream of
air) until the fibrin was neutral in reaction toward phenolphthalein, dried the fibrin
on filter paper, first in air and then in a thermostat kept at 37°, powdered the fibrin
a second time in a mortar, and then repeated the whole process a third and even a
fourth time. Fvbrin when thus prepared swells more in distilled water than in
acid. This observation leads us to believe that the following is most probably true.
Absolutely pure fibrin (free from salts) swells most in pure water. Every electro-
lyte reduces the amount of swelling, but the acids are much less powerful in this
regard than the salts (see paragraph 7, p. 339).

On the Swelling of Fibrin. 337

1/5 N. hydrochloric acid and 1/4 M. sodium chloride solution
was substituted for water. This, too, represents the height to
which the fibrin would have swelled if placed directly in this
mixture.

Two more series of experiments may be introduced here to fur-
ther illustrate what has been said. 0.25 gm. of fibrin was introduced
into each of seven test tubes containing 20.0 c.c. 1/10 N. hydro-
chloric acid. After fifty minutes the fibrin columns measured

22,20. 20 22022. 22. 24 Tom.

respectively in the series of tubes. The supernatant pure acid solu-
tion was now carefully pipetted off each of the fibrin columns,
and these were covered with 20.0 c.c. of a mixture of equal parts
of a 1/5 N. hydrochloric acid and a 1/4 M. solution of the fol-
lowing salts (— a 1/8 molecular salt solution in a 1/10 N. hydro-
chloric acid solution): KCl, NH,Cl, NaCl, CaCl,, MgCl,, SrC'g,
BaCl,. After an hour in these acid-salt mixtures the fibrin columns
stood,
18.5, 18, 17-5, 17, 15-5) 15-5) 14 mm.

On the following day the columns stood slightly lower (about
I mm.), but their order remained the same.

In a similarly conducted experiment 0.25 gm. fibrin introduced
into each of four test tubes containing 20.0 c.c. 1/10 N. hydro-
chloric acid had swelled to measure after thirty-five minutes,

20, 20, 20, 21 mm.

After pipetting off the pure acid solution, and substituting 20.0 c.c.
of mixtures of equal parts of 1/5 N. hydrochloric acid with 1/4 M.
solutions of BaCl,, BaBr,, Ba(C,H,O,’., Bala, the columns stood,
at the end of fifty minutes,

r1, 8, 7, 5 mm.
and twenty-four hours later,
9, 7-5, 6, 4-5 mm.

It will be seen that the order of the cations and of the anions,
as outlined in this paragraph, is identical with that given on

page 335-

338 Martin Hl. Fischer and Gertrude Moore.

6. The taking up and giving off of water by fibrin does not,
however, represent a completely reversible process in the sense that
if fibrin has swelled to a certain degree in one fluid, and to a greater
or less degree in a second, it will (in the time allowed in our ex-
periments) return entirely to the original degree of swelling when
again placed under the original external conditions. In other words,
a more or less lasting impression is made upon the fibrin by the
conditions through Which it has previously passed. This well-
recognized property of certain colloids is illustrated by the fol-
lowing experiment. 0.25 gm. of fibrin is introduced into each of
three test tubes containing 20.0 c.c. 1/10 N. hydrochloric acid.
After seventeen hours the fibrin measures 21 mm. in each of the
tubes. One tube is now refilled with 1/10 N. hydrochloric acid
and kept as a control. The acid is poured off the second, and after
the fibrin is washed in water, it is covered with this fluid. The
third tube has the acid poured out of it, and after first rinsing in
the fluid, the fibrin is covered with 20.0 c.c. of a mixture of equal
parts of a 1/5 N. hydrochloric acid with a 1/4 M. sodium chloride
solution. After twenty-four hours the fibrin in the pure acid still
measures 21 mm.; that in the water 51 mm.; that in the acid-salt
mixture 16 mm. The solutions are all changed again, this time
back to a pure 1/10 N. hydrochloric acid solution. At the end
of another twenty-four hours the control tube still measures 21 mm. ;
the fibrin from the water has shrunk to 27 mm.; and that from
the acid-salt mixture has swelled to 18 mm. The solutions are now
changed a third time. The control is put into another change of
1/10 N. hydrochloric acid, and the other two tubes have water and
acid-salt mixture, respectively, once more introduced into them.
When another twenty-four hours have elapsed, the control measures
22 mm., the fibrin column in the water 65 mm., and that in the
acid-salt mixture only 14 mm.

In explanation of these facts it is probably correct to assume that
each time fibrin is removed from an acid solution to distilled water
a part of the salts existing as impurities in the fibrin diffuse out,
so that each time the process is repeated the fibrin approaches more
nearly a salt-free condition, and in consequence gets into the state
in which it is capable of absorbing the largest amounts of water
(see p. 336, footnote). The reverse is true where fibrin is removed
from a pure acid solution and placed in one containing salts in
addition. With each change the fibrin loads itself with the salt,

On the Swelling of frbrin. 339

and as colloids part only with difficulty from a salt which they have
(mechanically?) bound, as much salt is not lost in the pure acid
solution as was taken up in the acid-salt mixture. The effect of
a new transference into an acid-salt mixture is therefore added to
what has remained of the effect previously produced.

7. We have made a few experiments with non-clectrolytes. We
found the thrice washed fibrin, described on p. 336 (footnote), to
swell as much in a pure 1/2 M. solution of cane sugar, glycerine,
or dextrose, as in pure water. 0.25 gm. of the fibrin swelled to
a height of 27 mm. in each of these fluids. In a 1/1o N. hydro-
chloric acid solution prepared as a control the same amount of this
fibrin swelled only to a height of 24 mm. and in a 1/5 M. sodium
chloride solution to 12 mm. Ina 1/2 M. urea the fibrin column
stood at 20 mm. Weare at a loss to understand why urea should
be an exception in the series of non-electrolytes, but believe it due
to the fact that urea solutions on standing give rise to (electrolytes)
ammonium compounds.

Neither do non-electrolytes share with electrolytes the power of
decreasing by their presence the amount fibrin will swell in any
acid solution. 0.25 gm. of thrice washed fibrin was placed into
20.0 c.c. of each of the following solutions. The height of the
protein column in each of the tubes at the end of two hours is
appended.

Distilled watery. (<4 se> o/iv )- |: 3G0nm.
1/20 N. hydrochloric acid ....,.. 28 “
Equal parts of 1/5 N. HCl and
Wt, Me OIVGEnING s- (5 3 fn 30. yt
Eee WiceSUCEOsetc: es, . oes. (2G “eS
Dre iwiitedmee ss: ee mw 2g ee
Dm Wieidexerases 8. iss 65 3 a ey.

66

1/5 MM. sodimm chloride... 23

The same fact is brought out if, instead of putting dry fibrin into
mixtures of hydrochloric acid and these non-electrolytes, we first
allow fibrin to swell in a pure hydrochloric acid solution, and then
substitute for this an acid non-electrolyte mixture. After 0.25 gm.
of fibrin had swelled to its maximum in 20.0 c.c. of a 1/10 N.
hydrochloric acid solution, the protein columns in the series of tubes
measured

26,-26,,26, 26, 25, 25, 23 mm.
The pure hydrochloric acid was now pipetted off the fibrin, and this
covered with 20.0 c.c. of a mixture of equal parts of 1/5 N. hydro-

340 Martin H. Fischer and Gertrude Moore.

chloric acid with the following solutions: 1/2 M. sucrose, 1/2 M.
dextrose, 1/2 M. urea, 1/4 M. uranium nitrate, 1/4 M. copper
sulphate, 1/4 M. sodium sulphate, 1/4 M. sodium nitrate. After
remaining seventy minutes in these solutions the fibrin columns
stood

28,27, 27, D2,/E5;-13, 2onmme:

CONCLUDING REMARKS.

We believe that the experimental results outlined above have
a bearing upon a number of problems in physiology. One of these
concerns the digestion of proteins in the presence of pepsin under
the influence of different acids. While the results obtained by dif-
ferent authors are still in a sense contradictory, all are agreed that
different acids affect pepsin proteolysis very differently, and such
careful studies as those of Pfleiderer* have shown that the mere
degree of dissociation of the acids does not determine the order
of their arrangement. While the exact placing of the individual
acids in a series is not the same with different authors, —a fact not
strange when the difficulties of making quantitative determinations
of catalysis are considered, — the general grouping is. Sulphuric
acid, for example, is usually the end member of the series in spite
of its great dissociation, while various organic acids, nitric acid,
and hydrochloric acid stand above it when arranged according to
the degree in which these acids favor pepsin proteolysis. We wish
to emphasize here what has already been pointed out elsewhere,”
that the general arrangement of the acids according to the way in
which they favor proteolysis under the influence of pepsin is in
large measure identical with the arrangement of the same acids
according to their power of making fibrin swell. This urges upon
one the belief that the medium in which a proteolytic ferment acts
serves to determine the rate of proteolysis not only through tts effect
upon the ferment, as is generally belicved, but also through its effect
upon the substance undergoing proteolysis.

Not only can experimental facts from the literature be adduced
to support such a conclusion, but a series of preliminary fermenta-
tion experiments which we have ourselves carried out do so also.

1 PFLEIDERER: Archiv fiir die gesammte Physiologie, 1907, lxvi, p. 605.
2? See FISCHER, MARTIN H.: Physiology of alimentation, New York, 1907,
pp. 117-118.

On the Swelling of Frbrin. 341

Even a superficial glance at the careful determinations of Chitten-
den! and his co-workers, Allen and Hutchinson, suffices to show
how the addition of every salt decreases the amount of fibrin di-
gestion in a pepsin-hydrochloric acid mixture. The results obtained
by these authors with the use of different salts cannot be directly
compared with each other, as they worked with percentage solutions.
But a little calculation soon brings out the fact that those salts which .
were found above to be most effective in diminishing the amount of
swelling of fibrin in hydrochloric acid are those which most power-
fully retard the digestion of this substance im pepsin-hydrochloric
acid mixtures.

Our own fermentation experiments consisted in studies of the
rate of digestion of fibrin in various pure acid solutions, and in
hydrochloric acid solutions containing various amounts of different
equimolecular salt solutions. In order to obtain comparable results
we followed, in a somewhat modified manner, Griitzner’s? method
of determining the rate of proteolysis. We stained our dried fibrin,
prepared as described in the introduction, as deeply as possible in
a 1/4 per cent carmine solution. When at the end of several days
the supernatant carmine solution no longer lost in color, and the
fibrin was stained uniformly red, we washed the fibrin in distilled
water, and after drying on filter paper in the air, desiccated it in
a thermostat kept at 37° C. Weighed amounts of this dried colored
fibrin, which control experiments showed to swell just as readily
as our ordinary fibrin, were then introduced into our different fer-
mentation mixtures.

We found the arrangement of the acids and the salts to be prac-
tically the same in the fermentation experiments as in those dealing
simply with the swelling of fibrin. Acetic acid and the acetates
constituted with us the most notable exceptions. While our colored
fibrin swelled readily enough in these solutions, the degree of di-
gestion in pure acetic acid, or in pepsin-hydrochloric acid solutions
containing acetates, was not so great as might have been expected
from the degree of swelling. We are now trying to discover the
cause for these exceptions. ;

1 CHITTENDEN, R. H., and ALLEN, S. E.: Studies from the Laboratory of
Physiological Chemistry of Yale College, 1885, i, p. 76; CHITTENDEN, R. H., and
HutTcuinson, M. T.: /did., 1887, ii, p. 55.

2 GRUTZNER, P.: Archiv fiir die gesammte Physiologie, 1874, viii, p. 452, and
Korn, A.: Ueber Methoden Pepsin quantitativ zu bestimmen. Inaugural Disser-
tation, Tiibingen, 1902.

342 Martin H. Fischer and Gertrude Moore.

As we know that living matter is made up in the main of colloidal
material, the experiments on the swelling of fibrin are of interest
because they give us an insight into the means by which cells and
tissues are able to take up and give off large quantities of water.
A colloid such as fibrin absorbs under proper conditions from
twenty to fifty times its weight of water. This more than covers
the largest amount of water that is ever absorbed by any of the
ordinary living tissues. When we consider that the amount held
by a colloid can be enormously affected through agencies which we
can imagine active in the living organism, such as slight changes
in the absolute or relative concentrations of any electrolytes (acids,
bases, salts) which may be present; and that these variations in
the amount of water held may occur independently of any changes
in osmotic pressure, we are perhaps justified in believing that in
the variable affinity of the colloids for water lies the explanation of
much that is obscure in those physiological phenomena that are
associated with the taking up or giving off of liquids.

The absorption of water by the gastrocnemius muscle of the frog
is entirely analogous to the absorption of water by our ordinary
(not entirely salt-free) fibrin. Such fibrin swells more in any acid
solution than it does in distilled water or any pure salt solution.
The same holds true of frog muscle. When fibrin has attained its
maximum degree of swelling in an acid solution, it will swell still
more if placed in distilled water. So will a muscle. Those non-
electrolytes which are incapable of altering the amount of water
absorbed by fibrin, are also incapable of altering the amount of
water absorbed by frog muscle. Equimolecular salt solutions affect
the swelling of fibrin to different degrees. Fibrin swells more in
a solution of a potassium salt than in one of a sodium salt, and
more in this than in an equimolecular solution containing a calcium
salt. The same holds true of frog muscle. The analogy between
the absorption of water by muscle and by soaps was pointed out
years ago by Jacques Loeb. Since then we have learned that the
soaps are colloidal solutions. We seem in consequence to have
in this fact and in the results of our experiments on fibrin adequate
reasons for believing that the absorption of water by muscles is
governed by the same laws which govern the absorption of water
by colloids in general.

Slr USE: OF BONE. ASH WITH. THE: DIET, IN
. METABOLISM EXPERIMENTS ON DOGS.

DYNAN RED Wes LEE AND WILEIAM J; GIES.

[From the Laboratory of Biological Chemistry of Columbia University, at
the College of Physicians and Surgeons, New York.]

HEN healthy dogs are fed a diet consisting of moderate
quantities of hashed meat, cracker meal, lard, and water,
very little of the food remains undigested and nearly all of it is
absorbed. On such a diet, or one that may be equally well digested
and assimilated, the fecal discharges from a dog are comparatively
infrequent, but the consistency of the excrementitious matter that is
eliminated is, as a rule, markedly diarrheal. When an animal that
receives such food is confined in a cage and the urine is collected
automatically in a receiver at the bottom, the elimination of thin
stools is very apt to result in mingling of urine and feces. If the
animal is used under these conditions for an investigation in which
urine and feces are to be analyzed separately, such mixture of the
excreta, contingent on the watery consistency of the feces, may
entirely vitiate the experiment, or at least will add greatly to its
difficulties, and to the annoyance attending accurate analytic work.
In the first experiment in which the senior author made such
observations, a bitch weighing 10.9 kilos received daily, in two
equal portions, at 8 a. Mm. and 6 P. M., a total of 250 gm. of hashed
meat, 70 gm. of cracker meal, 40 gm. of lard, and 500 c.c. of
water. After a preparatory period which ended with defecation,
the bitch eliminated brown to black diarrheal stools on the
following days only, during the period of twenty-seven days she_
was under observation: Ninth, fourteenth, eighteenth, twenty-
third, and twenty-seventh days... The thoroughness of the as-
similation of the food given this particular animal, in spite of

1 CHITTENDEN and GIEs: This journal, 1898, i, p. 1; also Studies in physi-
ological chemistry, Yale University (1897-1900), 1901, p. 1; and Gres and col-
laborators : Biochemical researches, 1903, i, Reprint No. 16.

343

344 Matthew Steel and Wilham F. Ges.

dosage with 5 gm. of borax daily from the tenth day to the
eighteenth day inclusive, was shown by the fact that the total
amount of nitrogen in the feces during this experiment lasting
twenty-seven days was only 9.66 gm., whereas the food during the
same time contained 268.33 gm., or an average of 9.94 gm. daily.
The nitrogen of the feces discharged during the entire experiment
of twenty-seven days was less, therefore, than that in one day’s
ration. When one considers that the nitrogen of feces consists not
only of food residues but also of gastro-intestinal secretory contribu-
tions, it is obvious that the food was very economically utilized by
the dog of this experiment. The fecal matter alluded to was always
thin, brownish black in color, extremely offensive in odor, and very
difficult to remove completely from the cage for quantitative analytic
purposes.

Shortly after the completion of the experiment that gave the data
just alluded to, two additional experiments of similar character
yielded results that were practically identical with those already
mentioned. Thus, in the longer experiment of the two referred to,
a bitch weighing 10 kilos received daily, in two portions, at 8 A. M.
and 6 p. M., 160 gm. of hashed meat, 40 gm. of cracker meal, 30 gm.
of lard, and 430 c.c. of water. The experiment lasted fifty-six days,
during twenty-four of which the animal received large doses of
borax or boric acid. The total amount of nitrogen in the food
during the fifty-six days was 358.82 gm., or a daily average of
6.41 gm., whereas only 13.55 gm. of nitrogen were contained in the
total mass of feces eliminated during that time, in spite of the
dosage to which the animal had been subjected. Defecation
occurred oftener than in the first of these three experiments, but,
since the stools were watery, their more frequent elimination in-
creased the difficulties and annoyances attending their complete re-
moval from the cage and their preparation for accurate quantitative
analysis.

The senior author and his co-workers made many similar observa-
tions in this laboratory on the thorough assimilability in dogs of
such food as that referred to above and also on the fluidity, offen-
siveness, and troublesomeness of feces eliminated by dogs on diets
of the same or similar character.1 From every other standpoint,
however, the use of such food in experiments on dogs was always

1 See MEAD and Gigs: This journal, 1901, v, p. 104; also GiEs and col-
laborators : Biochemical researches, 1903, i, Reprint No. 21.

Fale

The Use of Bone Ash with the Diet. 345

eminently satisfactory. A special method for the preparation and
preservation of meat for use in metabolism experiments was ac-
cordingly devised,’ and the food mixture referred to above, viz.,
prepared meat, cracker meal, lard and water, has been given regu-
larly since, in all the experiments in this laboratory on dogs, be-
cause of the assimilability of that diet as well as the convenience
of handling it, and, until recently, in spite of the annoying fecal con-
ditions attending its employment. As the number of workers in
metabolic investigation increased here, however, the difficulties and
annoyances attending the elimination of diarrheal stools and their
preparation for quantitative analysis became more and more promi-
nent, and developed to an extent that concentrated attention upon
them.

About five years ago, when the need for relief from this situation
became pressing, the senior author, in considering the possible
ways and means to that end, recalled his observation that dogs,
which had received food containing bone, eliminated solid feces
that were gray in color and practically odorless, and which dried
quickly at ordinary room temperature to brittle masses that could be
very easily crushed to a light calcareous powder. This condition
of the feces appeared to be due to the preponderance in them of bone
salts. It seemed apparent that in such cases the ingested bone was
thoroughly digested and the organic products absorbed, but that
the inorganic constituents of the bone, mainly the calcium *com-
pounds (unlike the organic constituents), were almost, if not wholly,
eliminated per rectum. Might not the addition of bone to the food
regularly given our dogs, as stated above, overcome the annoying
fecal conditions attending the use of that particularly assimilable
diet?

The desirability of adding indigestible matter to the diet of our
animals, in order to give solidity to their feces, had frequently come
to mind, of course, because of the favorable experiences of others
with such materials. It seemed unwise, nevertheless, to subject
dogs in any carefully conducted experiments to the continuous me-
chanical effects of materials that are never contained in food to
which the dog is ordinarily accustomed. The above-mentioned
observations of the effect of bone, in the diet of dogs, on the char-
acter of the feces, made it seem probable that bone could be em-

1 Gries: This journal, rgor, v, p. 235; also Gres and collaborators: Biochem-
ical researches, 1903, i, Reprint No. 1.

346 . Matthew Steel and William F. Gres.

ployed to advantage to regulate the consistency of the feces. It
was obvious that no objection could be raised to such a use of bone
on the ground that it was foreign matter which the dog’s digestive
apparatus could not handle without probable detriment to the ani-
mal, for it is too well known that bone is a satisfactory food-stuff
for dogs.

There were, however, certain practical objections to the use of
bone for the purpose indicated. It would be impossible to feed
bone in daily masses of uniform composition, unless it were given
in the form of bone dust! from large, uniformly mixed, reserve
supplies of the latter. But such pulverized bone would contain con-
siderable organic matter whose removal by digestion from the com-
pact particles might not be uniform, in which event the bone dust
would introduce undesirable inequalities into the daily metabolic
changes, and would make irregular contributions to the feces. It
seemed wise not to run this risk if it could be avoided. From this
standpoint, crude bone black 2 appeared to be better adapted for the
purpose, because it presumably contains no soluble or digestible
organic matter. Nevertheless, poorly burned commercial bone
black, possibly with noxious products in it, could not well be dis-
tinguished from the best commercial samples, and on that account,
even if there were no objection to the mechanical effects of the
contained carbon, it seemed wise to reject bone black also as the
carrier of the desired inorganic bone matters. ‘As in the case of
bone dust, it was preferable to avoid unnecessary risks.

For many reasons the pure white, thoroughly roasted, pulverulent
bone ash of commerce appeared to be best adapted for the purpose.
The material is cheap and easily obtained in large quantities. It
seemed probable that a desired consistency of the feces could be
obtained with less bone ash than either bone dust or bone black.
There was no reason to think that incineration of bone causes un-
desirable transformation of its inorganic constituents, although
carbonate is increased. The absence from bone ash of organic mat-
ter and of half-burned noxious products increased the likelihood
of its greater availability than the other products mentioned.

Even in the use of bone ash, however, there seemed to be certain
undesirable conditions. The somewhat larger proportion of car-
bonate in bone ash than in either bone dust or bone black suggested
possible disturbances from immoderate liberation of carbon dioxide

1 Pulverized bone. 2 Pulverized bone charcoal.

The Use of Bone Ash with the Diet. 347

in the stomach. It also appeared likely that the constituents of bone
ash would be more prone to exhibit direct and rapid chemical influ-
ence in the stomach (where introduction of the free bone ash would
place the salts in immediate contact with the gastric juice and gas-
tric mucosa) than would practically the same inorganic matters,
when given in the form of slowly digestible bone. Any such in-
creased chemical effects, due to augmented mass action of the free
osseous inorganic matters, might be expected to induce deleterious
local results as well as to cause general metabolic influences.

A few preliminary trials with bone ash in the usual food of several
dogs made it obvious, however, that all the expected advantages
were gained from its use, but none of the suspected disadvantages
appeared to follow its ingestion. The first mention of the satis-
factory addition of bone ash to the diet of dogs in this laboratory
was made casually four years ago,’ in a report on the investigation
in which its systematic use was begun, after the preliminary experi-
ments that assured its successful employment. Shortly afterward
we presented a more detailed account of the advantages that accrue
from the use of bone ash in the diet of dogs, especially in me-
tabolism experiments.’

During the past five years bone ash has been added regularly to
the food of practically all the dogs used in this laboratory, and has
been employed with signal satisfaction to all concerned. It has
even been given advantageously in milk to kittens, to overcome
proneness to diarrhea. Diarrheal tendencies, which commonly
follow the use, for food, of milk, rice, dog biscuits, etc., can be
almost immediately reduced to a minimum, or prevented entirely
by the addition to the diet of proper amounts of bone ash. We
have never witnessed a single exhibition of unfavorable symptoms
that could be ascribed to the ingestion of reasonable amounts of bone
ash with food. Even when excessive quantities of bone ash were
added to the diet of healthy dogs, the only unfavorable result ever
noted was a certain degree of distress as defecation was inaugurated.
We have ordinarily given about 10 gm. of bone ash daily in the
food of dogs of average size. 1 gm. per kilo has been an effective

1 TALTAVALL and Gres: Proceedings of the American Physiological Society,
This journal, 1903, ix, p. xvi; also GiEs and collaborators : Biochemical researches,
1993, i, p- 59.

2 Gigs: Proceedings of the American Physiological Society, This journal, 1904,
x): xxii.

348 Matthew Steel and Witham F. Gites.

average quantity with the food used in this laboratory. (See foot-
MOte; “p: 350!)

Besides preventing diarrhea in dogs subsisting on various kinds
of food that usually encourage the elimination of watery stools, the
addition of bone ash to a diet consisting of hashed meat, cracker
meal, lard and water, considerably increases the bulk of the fecal
matter, and makes its discharge comparatively frequent and quite
regular. A dog such as the one referred to on p. 343, if given
daily 10-15 gm. of bone ash with the food, would ordinarily defecate
about once a day. The feces under the influence of the bone ash
treatment are usually almost odorless, and have the typical consist-
ency and appearance of the fecal elimination commonly passed from
dogs after bones have been eaten by them. Intestinal putrefaction is
kept at a fairly constant low level when bone ash is administered.

Occasionally caged dogs eat their own feces, especially, it seems,
when their discharges are particularly odoriferous. Such an event
in a metabolism experiment cannot occur without very disturbing
consequences. Since we began the administration of bone ash to
dogs five years ago, we have seen a manifestation of this tendency
by only one of several hundred dogs that have been under con-
tinuous observation. The exceptional dog referred to was receiving
daily an excessive amount of protein matter, and the proportion of
bone ash had not been kept as high as it should have been. The
feces were therefore quite soft, and were ill smelling besides; and
the dog disposed of them in the manner indicated. As a rule the
feces that are passed after the administration of bone ash are so
nearly devoid of offensive odor, and are so chalky in appearance and
consistency on drying, that they appear to excite in caged dogs no in-
clination to ingest the excrement.

The eliminations from dogs on the diet referred to above are
passed in the form of light-brown lumps. These lumps do not
adhere to the cage, but may, as a rule, be easily lifted and quickly
removed completely, without disintegration. They dry quickly at
room temperature to fairly hard, yellowish white, brittle masses,
and in this condition may be rendered perfectly pulverulent in a
mortar with a few strokes of a pestle. Such a powder can be
quickly made as homogeneous as any other. Consequently, desic-
cation on a water bath or by any special means is entirely unneces-
sary as a prelude to pulverization; and loss of volatile products,
such as ammonia, is greatly reduced. The powdered feces from

The Use of Bone Ash with the Dret. 349

dogs on the diet indicated are light, yellowish white, and much like
the original bone ash in every way. Passing the pulverized product
gently through a very minutely meshed sieve is a simple matter, and
the isolation of hair and any foreign particles can thus be quickly
and satisfactorily attended to. Feces of this character are easily
handled, may be readily prepared for analysis without any annoy-
ance to the operator, and give no trouble in analytic procedures.

For some time seven animals in as many cages have been under
daily observation in this laboratory in a number of researches now
in progress. It takes Mr. Christian Seifert, our efficient laboratory
assistant, a little less than three hours to weigh the food for, and
feed, all these dogs; to collect and measure all their eliminated
urines, and determine the reactions and specific gravities before
bottling them; to remove completely the feces to, and break up the
larger masses in, flat weighed evaporation dishes preparatory to
deposit on shelves for spontaneous desiccation; to brush up and
bottle the hair and scurf from the drip pans; to put each cage
into condition for the collection of excreta during the succeeding
twenty-four hours; to co-operate in meeting all ordinary emer-
gencies; and to keep an accurate record of every detail on these
matters pertaining to each animal. If the feces were diarrheal, it
would be impossible even for Mr. Seifert, with all his experience
and ability, or for any one else, daily to accomplish so much in
so short a time. The great advantages accruing from the use of
bone ash with the diet of dogs are appreciated most where large
numbers of dogs require daily attention in quantitative experiments.
In this laboratory we regard the bone ash treatment as a particularly
fortunate method for the advancement of metabolism work.

If it is desired to mark off as distinctly as possible the feces of
any particular day, the addition to the food or bone ash for that
day of a trifling quantity of pure, finely powdered, thoroughly
washed charcoal imparts an unmistakable gray color to the feces
eliminated during the succeeding twenty-four hours. This color
is about the same as that of a mixture of the amount of charcoal
taken and the daily portion of bone ash. If charcoal is withheld
from the next day’s food, the gray color disappears from the fecal
mass of the succeeding day, or is present, sharply defined, in only
a small portion of it. We have found that if a gram or two of
powdered charcoal is given to a dog of ordinary size, fed regu-
larly each morning on the diet commonly used by us in this labora-

350 Matthew Steel and Witham F. Ges.

tory,’ the black material put into the food on one morning appears
almost wholly in the fecal discharge during the succeeding twenty-
four hours. The fecal elimination during a period of twenty-four
hours after a meal, under the usual conditions of our experiments,
represents, with a fair degree of completeness (75 per cent or more),
the fecal formation during the same time. These facts favor very
satisfactory division of an experiment into periods of fairly definite
excretory conditions.

The fondness of dogs for bones and the digestibility of bone in
dogs are well known. Bone ash does not seem to introduce into
a diet anything that is injurious to dogs, but appears to behave in
them much like an equivalent amount of bone. That bone ash has
little, if any, effect on the taste of the usual diet, or on the animal
itself, is evident from the total indifference manifested by dogs
toward it. When excessive amounts are given, little attention is
paid to the bone ash even by well-nourished animals with the usual
appetite.

Thus a normal dog, weighing 17 kilos, took daily for a week or
more as much as 100 gm. of bone ash admixed with 250 gm. of
hashed meat, 70 gm. of cracker meal, 30 gm. of lard, and 500 c.c.
of water, without showing any observable effects whatever, except
more frequent and abundant defecation. Although such a soupy
mixture, containing a great excess of the bone ash, looks much like
thick milk of lime, the dog ate it with evident relish. This animal
was a well-nourished one, with only an ordinary appetite.

Moderate amounts of bone ash in the food appear to be devoid
of harmful effects on both digestion and absorption. We expect
before long to inquire especially into the possible effects on fat
absorption. So far as we can now judge there is no hindering
influence.

The nature and extent of the chemical changes that bone ash
undergoes in transit through the gastro-intestinal tract have not
yet been determined, but that very little of such bone ash is absorbed
was shown by the results of the following special experiment on the
influence of bone ash on the excretion of calcium and phosphate
in the urine.

A well-nourished dog, weighing 6.5 kilos, and confined in one
of Gies’ cages,? was given daily 97 gm. of prepared meat,? 26 gm.

1 Usual amounts fer £ilo: prepared meat, 15 gm.; cracker meal, 4 gm.; lard,

3 gm.; bone ash, I gm.; water, 35 c.c.
? GIES: This journal, 1905, xiv, p. 403.
8 Gres: /bid., 1901, v, p. 235; also GiEs and collaborators: Biochemical re-

searches, 1903, I, Reprint No. 1.

The Use of Bone Ash with the Diet. he

of cracker meal, 20 gm. of lard, and 225 c.c. of water. This diet
was fed daily at 10 A.M. After a preparatory period of five days,
during which the dog’s weight fell to 6.42 kilos, our analytic work
was begun, and was continued on the urines of twenty-seven days.
Samples of three-day volumes were used for the determinations of
calcium and phosphorus.

Phosphorus was determined in 25 c.c. portions by the usual alkali
fusion process. It was weighed as magnesium pyrophosphate and
expressed in the records as P,O;.

Calcium was determined, by the advice of Professor Sherman,
as follows: 50 c.c. of urine were strongly acidified with acetic acid,
then nearly neutralized with ammonium hydroxid, and treated with
a moderate excess of ammonium oxalate. After standing not less
than four hours, the precipitate was filtered off, washed with water,
and dissolved in 20 c.c. of hydrochloric acid of 1.04 sp. gr. This
solution was then diluted with water to a volume of 150 c.c., heated
to 70° C., and, at that temperature, titrated with 7% solution of
potassium permanganate. The calcium found was expressed in the
records as CaO.

Our results may be seen at a glance in the table on page 352.

The summary of analytic data on page 352 makes it evident
that administration of bone ash had no influence on the urinary
excretion of calcium. It is apparent from the same data, also, that
the elimination of phosphorus (presumably also of phosphate) was
not increased by the ingested bone ash.

The observed gradual diminution of eliminated phosphorus may
have been due chiefly, if not wholly, to the metabolic changes re-
sponsible for the steady decline of body weight and decrease of
urine volume. It is probable that the animal did not receive food
enough upon which to maintain weight equilibrium and that, under
these conditions, phosphorus catabolism was less pronounced at the
close of the experiment than at the beginning. The low tide of
phosphorus elimination at the time when most bone ash was given
may be more than a coincidence, however, and additional experi-
ments may show that urinary excretion of phosphorus (phosphate)
is somewhat diminished by the administration of bone ash. If this
should prove to be the case, the decrease in the amount of phos-
phate in the urine might be due to diminished absorption of alkali
phosphate. Such a result might ensue from interaction, in the in-
testine particularly, between calcium and phosphate (each being

352 Matthew Steel and Witham F. Gres.

TABLE I.

RESULTS SHOWING EFFECTS OF BONE ASH IN THE DIET ON THE URINARY
ELIMINATION OF PHOSPHATE AND CALCIUM FROM A Doc.

Volume. P,Os. CaO.

Three-day Three-day Three-day

Spe- Z 2 é
iP periods. periods. periods.

cific
te Dail Dail
ity. aily aily Daily

Total.| aver- | Total.| aver- | Total.| aver- Dry
age. age. age.

weight.

09

SCMAANOCOCCOE

Cicr % 5 < gm. &

1.6063

IOAN MN fh WD
DAwMmonNO MN PB

8 12
9

0.4673

pA
oO

ne
oO
—

0.3096

a
oO

0.3180

0.2613

0.2749

0.7317 | 0.2439

present in both the food and the administered bone ash), with the
production of less soluble or precipitated products. Possibly such
interactions would occur especially between alkali phosphate and

The Use of Bone Ash with the Diet. 302

that portion of the available calcium that had been converted in
the stomach into chlorid from carbonate, and which, as chlorid,
would be prone in the intestine to combine with phosphate in in-
creasing proportion as the mixture containing them became less
acid or perhaps alkaline in reaction. Changes of this sort, compara-
tively slight at best with the quantity of bone ash we use, would
doubtless be quite regular and uniform, daily, in the average experi-
ment, and would probably affect none of the essential conclusions
on metabolism. The total amount of phosphorus (phosphate) in
the excreta would doubtless be unaffected, for the phosphate with-
held from absorption would be passed within a few hours, as cal-
cium phosphate, into the feces.

Additional experiments in this connection, as well as on the effect
of bone ash on absorption, will shortly be undertaken.

When we first began the use of bone ash in the diet of dogs, it
occurred to us that great difficulty might be encountered in any
effort to measure accurately the metabolic changes in which phos-
phorus was involved. It seemed probable that the large amount of
preformed and non-significant phosphate in the feces (derived di-
rectly from the bone ash) would be particularly difficult to collect
to such a degree of quantitative completeness that inevitable me-
chanical losses and variations would not be regularly great enough
to hide metabolic fluctuations. We have found, however, that fecal
conditions after the administration of bone ash are so favorable
for easy quantitative assemblage of the daily portions of solid
excreta, and their preparation for analysis without material loss,
that true phosphorus balances may be determined as satisfactorily
as those of nitrogen or sulphur. All that is required is constant
care and watchfulness at every stage of the process of collecting
the excreta. |

The above-mentioned facts were first shown in the hemorrhage
experiments in this laboratory described by Hawk and Gies.1 Their
experience has been repeated here in a number of metabolism in-
vestigations, e. g., in the radium research by Berg and Welker.”

The summaries on pages 354 and 355 give conclusive data bear-
ing on these points. _

It often happens in metabolism work on animals, when the urine
is not removed artificially, that urine fractions come in contact with

1 Hawk and Gries: This journal, 1904, xi, p. I7!.
2 BERG and WELKER: Journal of biological chemistry, 1906, ii, p. 371.

0s°0
8r'0
tr'0

on)

c0'0-
c0'0—-
£0'0—

60°0—
c2'0—-
Lo0-

cS'6

PLL
SLL

£L'6
ESZt
IST

65°0
$S'0
8t0

‘

€ ‘1a7VM fwd g ‘Yse dUOg {WS CT ‘prey f'wS ZG ‘feo AayORIN {-wS OOZ Yeout poredaid

Tho
so'oO—

co-
Is'0-

cS'e
8e°9
68'S

685 -) 7." % * ™ pisauysamyy |G
68'S ; . . .

* aSeysiowary | ¢

*|euon | 6

‘Poo “SOT ZT “Sop ay) Jo WTI M4

09°0
8r'0
cS'0
Ly'0
0s'0
+50
80
“wa

OsT'O—
$90'0+
ec0'0r
£9TO+
Sl0'O+
4 Os
$h0'0+
“m3

£L£0-
SF O+
erOr
0) Aa
sTO+
61+
ssOr

‘OSRIDAL
Ayieq

citadel

‘aSVISAR

Aweq TROL

‘aouLleg

Matthew Steel and Witham F. Gees.

“SNOUOHdSOHG

er'0

es 91
98°0S
+8°62
Cb’ +7
oe
€6'SZ
“wd

“payaso

“Xf

L i 0£6
86 91
62'S
bO'TE
LOVE
OL 6E
Ly 6%
“wa

HEA

*‘pajsas
“uy

c8'0
cL0
+L0
L9°0
£0
6S°0
mg

Sre'0—
622:0—
Ost'0—
+10°0—
sovo+
£20°0—
sevot
“wa

(ofa
Oe)
Is ¢—
sl0-
SO'T+
6£°0—
1I9'I+
“ma

*OSVIDAV
Ape

“AIVULIL

‘QSRIOAR

Agpeq. | Pel

“20uRleEq

+e'F
659
+9'02
SOUL
6E°L
LOet
09

S3°C
66'F
£891
Lei
+8 aes auoye visayjsxuy

(%€67) asvyssowsy ysT
SSZI | pur ‘uoneisdo “sayjsxuy
99'6 Cave VOM Af

m3

*(%9¢'¢) eBeyssowasy yp
* (%Is'¢) aBeysroway pe
* (YeeE) eBeyssoway pz

* — paurquio9
uoneiado pur visayjsxuy

* [eWION

*payaio
“X47

‘poqsas
“Uy

‘SUOT}IPUOD

MASI,

‘yNHATAg

“Saollad

"9'9 QOS “tayeM SWS OT ‘yse suo fwd O¢ ‘prey fw OZ ‘feaut ray9vIN {WS ECZ Geow paredaid :poog “sopry ZT ‘Sop 941 JoYSIOM LV

354

‘“LAId AHL OL GAddV SVM HSY ANOG HOIHM NI SSO0q NO SLNUWIYAIXY NI SHONVIVG SNYOHdSOHG GNV UNHdTNS

At Wrava

355

The Use of Bone Ash with the Diet.

‘90h “dl 722 207 ‘UAMIA AA pur OAAT 9g ‘TOb “dl “729 207 “MANTA pur OUAT g "b6E “d “717 07 :YANTAN pur ova
‘88 ‘d “729 207 :UANTAM pue OUAG "L1G ‘d 719 207 i sary pure AMVY ‘207 “dl “722 207 : Sait) pue MMVI] _

020 Pi Otee | eect S28 8f°6 sto 1o';O— | 600—- | 96T LST ~ “ prog untipey | §
£20 LOO+ | 9FO+ | HLL 02'8 sl0 LO0— 9" 8010S) cadet +9'T oo ee SLIT 7

‘99 SZe “1am Fwd ¢ “Yse auOq f-wS OT ‘par, fwd OZ ‘Jeaw s9yovI {WS OOT ‘yeout poredord *sO[Py 9°¢ ‘Sop ayy Jo WY BIO A

£70 S907) OLe+ S62 Il‘9 £20 LOO | 2007 | SET orl Se SS a aks ESR a ate pull
020 ORO | ott SPel L9'+1 8l0 COO || SLOE |. ze Ore "oe" " pruoiq wnipey | ZT
910 9S°0+ Isct+ | O€¢ IT9 L£10 COO | LO Sct ort ee eee Se ROO Niel: ais

‘9D phe “19M Fwd ¢ “Yyse auog {WS 6] ‘purl {WS OP ‘eau ayOvsO ws OLT Seow poiedaid :poog ‘sopry 9°9 ‘Bop ayy Jo 1yS19 Ay

L270 LOO | LZT0+ Ih +1 SSF +20 LOG) RODS i W625 I8’€ ae Prwoszq wnweg | el
0£'0 +0'0+ SCOT | 92'8 Is'8 Sc 0 ae 10‘;0— Ie? CoS Sn SO hace Maan UNO NT ey,

"9'9 Q9Z ‘197eM Sus ¢ ‘Yse oUOG fw QZ ‘parry f-WS zp row Jayovs19 fwd fT] Jeouw posvdaid : poo ‘sorry sg ‘Sop ayy JO JYSIOA

020 CO0— | -ShO= |. GEST bob 1rA0) +0°0— SEN Ses OL "7 * * prwoiq untieg | 6T :

c

ALY) in TloO+ | LO% 80°L cz 0 cO0-— | 920-— | Sz COS SE en” ORC UMO NT 6

‘9°9 OIG ‘A9VM fwd ¢ ‘Yse suo fwd QZ ‘pry {WS ce ‘frau tayovss Sw 06 “vow paivdaid : poog ‘soply gp ‘Sop ayy JOVWSIOM gD

356 Matthew Steel and William F. Gites.

fecal masses. If the animals are confined in cages on suitable wire
supports that suspend the feces, this mingling tendency is reduced
to a minimum, because the urine fractions immediately run through
the wire netting as they are voided. Even in the best cages with
wire net supports, however, fecal fragments occasionally pass
through the meshes, and are apt to be washed by urine fractions that
pass over them to the receiver, and perhaps flush them into the latter.
Bone ash in the food of dogs makes their fecal discharges solid, as
we have said, and if such eliminations are not removed from the
cage soon after passage (and during the night this cannot be done
regularly), they speedily dry, also become somewhat brittle as
a consequence, and, if the dog tramples on any of the masses
(though this happens only occasionally, for the animal usually
avoids the material), fragments pass through the wire support, no
matter how small the meshes may be. What is the effect of urine
on such fecal matter? Is calcium or phosphate dissolved from it
by the normally acid urine, and are appreciable changes brought
about in the composition of the latter when dog urine washes fecal
fragments? Does acid urine decompose any calcium carbonate, for
example, in such feces?

We have given this matter attention repeatedly, and have found
that acid urine is practically devoid of solvent action on fecal con-
stituents derived from bone ash. The following observation on
bone ash itself is conclusive: |

Freshly voided strongly acid urine from a well-nourished dog,
and which had not been in contact with any fecal matter, was _
selected for a determination of the solvent action of normal urine
on the calcium in bone ash. Two equal portions of the urine,
150 c.c. each, were transferred to covered beakers of equal size.
To one portion, 2 gm. of bone ash were added. The other portion
was kept as a control. Each was thoroughly stirred repeatedly
during a period of four hours, after which both were filtered. Cal-
cium was determined in 50 c.c. samples of each filtrate, by the method
referred to on page 351, with the following results:

50 c.c. of filtrate from urine without bone ash contained 0.0067 gm. of CaO.

50 c.c. of filtrate from urine with 2 gm. of bone ash, contained 0.0070 gm.
of CaO.

As a rule no greater changes would occur in the course of a
metabolism experiment. Because of the greater bulk of the “ bone

The Use of Bone Ash wrth the Dret. a7

ash feces,” and the consequent attenuation of the soluble substances
in them, it is probable that any removal of such substances by urine
is ordinarily both absolutely and relatively less from “bone ash
feces” than from the excrementitious matter of dogs whose diet
does not contain bone ash.

A SIMPLE ELECTRICAL ANNUNCIATOR FORRRe.
IN METABOLISM EXPERIMENTS, AND IN CON-
NECTION WITH FILTRATION, DISTILLATION a?
SIMILAR OPERATIONS.

By WILLIAM H. WELKER.

[From the Laboratory of Biological Chemistry of Columbia University, at the
College of Physicians and Surgeons, New York.]

N the paper describing his cage for metabolism experiments,
Dr. Gies? referred to the advantages of the “ sliding shelf”

devised as a holder for the urine receiver, and in that connection
made the following remark: ‘“ The shelf also favors the use of
electrical apparatus to ring out the time of elimination of urine
fractions, in experiments in which fractions of the urine must be
examined separately and immediately after their natural excretion ”’
(p. 407). This remark alluded to one of the several additional de-
vices Dr. Gies had contemplated perfecting for use with the cage
described.

In his preliminary communication regarding our annunciator
Dr. Gies* stated that, “in order that an annunciator might be of
the greatest service in metabolism work in the way already indi-
cated, and also to insure its usefulness for filtration, distillation,
and other operations in which the weight of a product above a
certain maximum amount could be relied upon to close an electrical
circuit and announce the delivery of the material, it was necessary
that it should be delicately responsive to the weight of several grams,
and yet be readily adjustable within relatively wide limits in that

1 This apparatus was exhibited before the Society for Experimental Biology
and Medicine, May, 1906, and was shown at the Scientific Exhibition of the
New York Academy of Sciences, in the American Museum of Natural History,
December, 1906.

2 GieEs: This journal, 1905, xiv, p. 403.

8 WELKER and GIEs: Proceedings of the Society for Experimental Biolog
and Medicine, 1906, iii, p. 77.

358

filectrical Annunctiator. 359

respect; also that it should be light in weight and of small compass,
but durable, and resistant to derangement from any cause; and that
it should hold, without risk to, or modification of the contents, any
suitable vessel placed upon it.”

At the request of Dr. Gies, and with his advice, I proceeded to
design a working form of apparatus to meet these requirements.
The simplest possible device appeared to be two square pieces of
board hinged together along one side, and held apart by an adjust-
able spring on the opposite side, with a bell and appropriate elec-
trical fittings. In the first model the pieces of board, which were of
wood, warped somewhat; consequently hard rubber was substituted
in the form finally adopted. In the first model a piano hinge was
used to connect the boards. In the second, the piano hinge was
replaced by a pair of small steel-pivoted brass hinges, Friction
in these pivoted hinges was much less than in the long piano hinge,
and the substitution of the former for the latter increased the sensi-
bility of the apparatus.

As finally constructed, our annunciator consists of (1) a pair
of square pieces of hard rubber (436 X 436 X 5/16 in.) united by
(2) a pair of small steel-pivoted brass hinges. (3) A post 1s
attached to the middle of the bottom side of the lower rubber board.
This post is threaded with a fine screw thread, It also has a key way
running its full length. (4) The nut regulating the pressure is
a thin circular disk milled on the outside. On top of the disk 1s
(5) a seat for (6) the spring, with a key for the key way in the
post. The lower end of the spring is slipped over the thin end of
the seat to its rim. To the upper board is fastened (7) a brass
plate with a hole in it through which the post attached to the lower
_board can pass freely. On the under side of this plate is a seat for
the top of the spring. When all these parts are in position, a small
circular (8) nut is screwed on the post above the plate. With the
aid of this nut the distance between the boards may be readily
adjusted. Electrical contact is made by means of (9) a platinum
plate in the top board and three (10) platinum points in the bottom
board. The platinum plate is connected by an (11) inlaid copper
wire running to the upper portion of one of the hinges. The lower
side of this hinge is connected with one of (12) the binding
posts at the side of the bottom board. The points are connected by
means of an inlaid copper wire extending to the other binding post.

360 William H. Welker.

(13) Four small steel points are fixed at the corners of the bottom
board to add stability to the annunciator. (14) The receiver, for
urine, distillates, etc., which rests on the upper board, is held in
place by means of (15) a large circular washer of rubber packing,
which is bound to the board by (16) a circular aluminium washer
fastened with four small (17) screws, and 1s of such size as to leave
the edges of the rubber free to oppose any movement of the receiver.

(18) A small dry cell furnishes

the current, and (19) a little bell

sed completes the mechanism. The
‘we accompanying illustrations make
ae ~ . A : =

clear the relationships of the parts

mentioned.

The apparatus is very compact.
With the dry cell and the bell
included, it may be placed on a
surface 5 <X 8% in. The small
bell we have used has a delicate
musical sound, and answers all
ordinary purposes. When used
under a metabolism cage to an-
nounce the elimination of excreta
or vomit, the ringing of such a
bell does not disturb the animal.

Electrical annunciator. It is obvious, of course, that the
apparatus may be connected with
a bell in a room some distance from that in which the main part

of the mechanism 1s placed.

In its present form the apparatus can be adjusted so sharply that
it will announce deposits of liquid as slight in weight as that of
1 c.c. of water. The deposit of fragments of feces, or vomit, or
anything equal to that weight, is duly announced. The apparatus
may be so delicately sensitized that, when resting on the lightly
suspended “ sliding shelf” of Gies’s metabolism cage, vibration of
the cage due to any special movement of the dog effects rapid
intermittent contact between the opposing platinum plate and points
in the annunciator, thus making and breaking the current, and pro-
ducing a discontinuous ringing of the bell until the movements of
the animal cease and the force of the vibrations falls below the

Electrical Annunciator. 361

effective impetus. Such announcements obviously help a busy
observer to keep close watch over the animal at significant intervals.’

On account of its simplicity, and the ease with which it may be
adjusted to ring out different amounts, this apparatus can be used
very advantageously in distillation work when certain fractional
volumes are desired. It can also be used in filtrations when the
filtrates are to be partitioned into certain volumes. The adjustment
for any volume is made by placing the weights equivalent to the same
on the upper board beside the receiver, and then turning the adjusting
screw at the bottom of the post, thus increasing or diminishing the
tension on the spring until the contact is just made, as indicated
by the ringing of the bell. The weights then being removed, an
equal weight of liquid must collect in order that contact may again
bé made and announced by the bell.

This apparatus has been used in this laboratory with particular
success on the “ sliding shelf’ of our metabolism cages to announce
the elimination of urine, vomit, or watery stools. It promises to
add materially to the equipment of any laboratory whose workers
appreciate the time that may be saved, and the accuracy that may
be insured, in many operations by the employment of such an
automatic device.

1 HAWK recently made mistaken allusions to GIiEs’s metabolism cage, which
BERG and I endeavored immediately to correct (Journal of biological chemistry,
1906, ii, p. 410). Hawk did not comprehend the very many advantages accruing
from the use of detached receivers, and particularly failed to appreciate the special
serviceability of the “ sliding shelf ’’ of the cages in this laboratory as a support for
a readily removable receiver. The use of an annunciator in the advantageous
ways indicated in this paper is made particularly satisfactory by the construction
of our cages and the character of the urine receiver used with them.

AN IMPROVED ANIMAL HORDE RS

BY GUSTAVE, MM; ANGE GE.

[rom the Laboratory of Biological Chemistry, of Columbia University, at the
College of Physicians and Surgeons, New York.|

HILE engaged recently in a study of the elimination of vari-

ous anilin dyes? from a dog with a biliary fistula, it was
desirable occasionally to suspend the animal in its normal upright
position for extended periods. At the same time provision had to
be made for the separate collection of the excreta and the bile.
As no apparatus possessing all these features was immediately
available, a suitable holder was improvised, which served excel-
lently the particular purpose for which it was designed.

Profiting by the suggestions of Professor William J. Gies, I
extended the plan of the apparatus so as to make it a dog holder
of general serviceability. The finished holder has been employed
in this laboratory by Drs. Carter, Cussler, Salant, and by Mr. Welker
and myself, in researches now in progress, and has always given
satisfaction. As its use during the past year has demonstrated its
wide adaptability and usefulness, a description of it is herewith
presented.

The particular piece of apparatus referred to will hold easily any
dog weighing from about 5 to 13 kilos. The dimensions given below
can readily be made to suit if smaller or larger holders are desired.
The support constructed for use in this laboratory not only answers
practically all purposes for dogs of average size, but can be placed
conveniently under an ordinary table out of the way when not
in use.

The apparatus consists of four movable parts on a frame: (1)
a belt to be fastened around the chest of the animal, with an at-

1 This apparatus was shown at the Scientific Exhibition under the auspices of
the New York Academy of Sciences, held at the American Museum of Natural

History, December, 1906.
2 MEYER: Journal of the American Chemical Society, 1907, xxix, p. 906.
262
362

An Improved Animal Folder. 363

tached band to be buckled around the neck, in order to restrain the
movements of the animal and to keep the head in a forward position ;
(2) a support for the hind legs, consisting essentially of two suitable
padded rings into each of which a leg can be placed; (3) an adjust-
able horizontal bar on which parts 1 and 2 may be fastened at
various positions and in different degrees of tension; (4) a shelf
holding a tray, with wire grating and having an outlet, for the separa-
tion and retention of feces and the delivery of urine into a suitable
retainer.

The accompanying illustrations give a general idea of the con-
struction of the holder and its appearance when in use. The extreme

Improved animal holder.

dimensions, not including the tray, are: length, 89 centimetres;
width, 41 centimetres; height, 50 centimetres. The cross beams
(A) are 67 centimetres apart. The horizontal bars (B) are 89
centimetres long, 61% centimetres wide, and 2% centimetres thick,
with bevelled edges. One of the horizontal bars is firmly fixed on
the cross beams, at one side. The other can be moved upon the
cross beams, and has a play of 20 centimetres. Two brass window
catches, attached to the movable bar, and running in grooves on
the cross beams, are used to fasten the adjustable bar in any per-
missible position. The grooves in the cross beams are reinforced
by angle brass (% X %&% inch).

Free to move between the legs which support the cross beams
is a hanging shelf (C). This shelf may be raised or lowered into
any position by means of a rope and pulley attachment. The shelf
is designed to hold the zinc-lined tray, which is also movable, and
is held in position on the shelf by means of a window catch. The

304 Gustave M. Meyer.

tray is provided with a readily removable wire screen. ‘The parts
of the tray may be quickly separated and thoroughly cleansed.

The device for keeping the animal in the holder consists of canvas
harness specially designed for use in this connection, and referred
to briefly above. The canvas (20 oz.), taken double, offers sufh-
cient firmness, with pliability, to be particularly well adapted
for this purpose. Leather could of course be substituted. The
chest belt (D) is 47 centimetres long and 13 centimetres wide. Two
flaps, each 20 centimetres long, are attached, one 10 centimetres,
the other 15 centimetres, from the ends of the belt. Three straps
and buckles fastened on the belt serve to tighten it around the chest
of the animal. Extending forward from the chest belt is a strap.
This is looped around the chest belt so that it may be readily centred
into position or removed at desire. The strap is 53 centimetres long
and 6 centimetres wide. On this strap the neck band (F) is fitted
so as to slide back and forth. The neck band is 30 centimetres
long, and is also provided with a set of straps and buckles.

Another attachment, not directly affixed to the harness, but which
is placed on the support directly in front of the forelegs, is a plain
strap (/’) 7 centimetres wide and 50 centimetres long. Its purpose
is to prevent the dog from placing his forefeet on the wooden
support, as well as to offer a comfortable resting-place for the neck
and head.

No particular difficulty was experienced in designing suitable
harness for the head and shoulders of the animal. On the other
hand, it was not an easy matter to hit upon a suitable arrangement
for the hind quarters. It was desirable to have as little attachment
as possible, but sufficient to keep the animal properly adjusted and
supported. A pair of padded canvas rings, 45 centimetres in diam-
eter, suspended on a band (G) 2% centimetres wide at its central
portion and 7 centimetres wide for some distance at its ends, and
having a length of 50 centimetres, was found to serve the purpose
desired. The rings are detachable, the ends being provided with
clasp buttons for this purpose.

To attach the various sections of the harness to the horizontal
bars, brass grommets ' were affixed, 5 centimetres apart from centre
to centre, at the ends of the canvas pieces. Several rows of eyelets

' Brass grommets (size number one) and coach buttons to match may be
obtained at any carriage supply house. A grommet punch is required to attach
the grommets. These articles are handy accessories in any laboratory.

An Improved Animal Holder. 365

make it easy properly to adjust the harness. The horizontal bars
are provided along their outer edges with brass coach buttons, with
which the eyelets may be engaged.

To fasten the animal in the holder it is best first to buckle the
chest and neck bands into position on the animal. The chest belt
should be fastened firmly, but of course should not make the animal
uncomfortable. The neck band should not be buckled tightly, for
the head movements need be only moderately restricted. When
the buckles on the straps referred to have been properly adjusted,
it is a simple matter to sling the ends over the bars and fasten the
grommets on the buttons. The rings for the hind legs can now
be put into position. There still remains the adjustment of the
horizontal bars, which, when properly effected, tends to give the
whole supporting apparatus a proper tension, and holds the animal
securely and comfortably. Finally the shelf should be so adjusted
that the paws of the animal barely touch the wire screen on the
tray.

As already mentioned, the holder is adjustable to dogs of various
sizes. The harness encircling the chest and neck is not apt to be-
come soiled if proper care is taken. The holder has served very
well for experiments on dogs with gastro-intestinal or biliary
fistulas. No undue pressure is exerted in the proximity of the
wounds in such animals, and there is no provocation of special
exudation. In experiments on dogs with esophageal fistulas the
neck band may be discarded, and soiling of the chest strap can be
prevented by spreading several layers of cloth on it.

Besides being of use in experiments on animals with fistulas,
the holder has proved constantly of service for restraining animals
for many other purposes, for example, during injections, feeding
by means of a stomach tube, introduction of enemata, etc. The
holder is especially convenient for supporting animals firmly in
vertical positions. The holder with the animal in it may be turned
on either end, in which positions the animal is unable to throw the
apparatus over or to change its own posture in it.

THE ELIMINATION OF RADIUM FROM NORMAL AND
NEPHRECTOMIZED ANIMALS|!

By WILLIAM SALANT#? ann GUSTAVE M. MEYER.?

[From the Laboratory of Biological Chemistry of Columbia University, at the
College of Physicians and Surgeons, New Vork.|

INTRODUCTION.

Bese the past fifty years numerous studies of the fate of
foreign substances after their introduction into the body have

shown conclusively that the kidney is not the sole organ of elimina-
tion for such substances. Metallic and other elements, as well as
complex organic compounds, if given by mouth or injected into the
circulation, pass from the organism not only by way of the kidneys,
but may be excreted by the digestive organs, by the skin, in the
Jachrymal or other secretions, or by various additional routes.

Thus the large number of investigations of the fate of iron in man
and other animals, no matter in what form or by what channel it may
have been administered, have shown that the large intestine is prac-
tically the only channel of elimination for this element. The same
was shown for manganese and certain other heavy metals whose
elimination has been studied. Somewhat similar data have been
obtained for the alkali earth metals. Thus, calcium, strontium, and
magnesium, although eliminated in the urine after their injection,
are found in much larger quantities in the contents of the intestines.
Various other elements, which are excreted chiefly by the kidneys,

' The fifth paper in a series of researches with radium inaugurated by Dr.
WILLIAM J. Gies in this laboratory several years ago. See (1) BERG and
WELKER: Journal of biological chemistry, 1906, i, p. 371; (2) BuRTON-OPITz
and MEYER: Journal of experimental medicine, 1906, viii, p. 245; (3) MEYER:
Journal of biological chemistry, 1906, ii, p. 461 ; (4) HussaAKoF: Medical record,
1907, Ixxii, p. 89. See also a recent report of an unfinished research in the series,
by GAGER: Science, 1907, xxv, p. 462.

* Fellow of the Rockefeller Institute for Medical Research.

® A preliminary report appeared in Science, 1907, xxv, p. 459.
366

Ltlimination of Radium from Animals. 367

are, to some extent, also eliminated by the intestinal epithelium.
The work of Good! has shown that lithium chloride injected
hypodermically into cats and dogs, although largely excreted in the
urine, is also found in the saliva as well as in the stomach and in
the. intestinal contents. Hanford? found that cesium is mainly elim-
inated by the kidneys, but it was also detected, after subcutaneous
as well as after intravenous injections, in the saliva, in the contents
of the stomach, in the contents of the intestine throughout its entire
length, and also in the bile. Likewise, rubidium, which leaves the
body mainly in the urine, was found by Mendel and Closson® in small
quantities in the feces. Such examples might be multiplied.

That the gastro-intestinal wall also serves as an organ for the elim-
ination of complex organic bodies has been demonstrated in the cases
of various substances, such as morphin, quinin, atropin, curarin, snake
venom, and other toxic compounds of biological origin. These partic-
ular substances seem, according to several investigators, to be elimi-
nated by the gastric epithelium as well as by the kidney.

That the bile is an important channel of elimination of various
substances introduced into the body has been shown by several
workers. After the intravenous injection of albuminate of zinc into
cats, dogs, and rabbits, Michaelis* found zinc in the bile of these
animals. Mosler® detected iodin in the bile of dogs eight hours after
the administration of potassium iodide. He obtained the same results
when he gave potassium iodide to a patient with a biliary fistula.
Wichert® studied the elimination in the bile of a large number of
elements when introduced into the body. Potassium iodide or potas-
sium bromide was given by mouth to cats and dogs. After a short
time the bile of these animals showed the presence of iodin or bromin
respectively. Similar results were obtained with antimony, nickel,
bismuth, lead, silver, and arsenic. In this connection may be men-
tioned the experiments of Mead and Gies‘* in this laboratory with
tellurium. According to these investigators this slement is elimi-
nated not only by the kidneys, but also by the liver, aiid even by the
~ 1 Goop: American journal of the medical sciences, 1903, Cxxv, p. 273.

? HANFORD: This journal, 1903, ix, p. 235.

8 MENDEL and CLosson : This journal, 1906, xvi, p. 152.

* MICHAELIs: Archiv fiir physiologische Heilkunde, 1851, x, p. 127.

§ MOSLER: VIRCHOW’S Archiv, 1858, xiii, p. 29.

® WICHERT: Dissertation, Dorpat, 1860, xiii, p. 29.

7

MEAD and GIeEs: This journal, 1991, v, p. 104. Also Gies and collabora-
tors: Biochemical Researches, 1903, i, Reprint No. 21.

368 Wilham Salant and Gustave M. Meyer.

lungs. Organic bodies of complex chemical constitution may like-
wise be eliminated in the bile. According to Brauer,! methylene
blue, when given by mouth, is excreted into the bile. Meyer? has
lately carried out in this laboratory a large number of experiments
with various anilin dyes. All of them were eliminated in the bile.

It is evident, therefore, from the foregoing very general résumé
that foreign substances, if introduced into the body, may be ejected
through many exits.

The results of the recent work here with barium by Berg and
Welker,’ and with radium and anilin dyes by Meyer,‘ led us all in
this laboratory to believe that various conclusions pertaining to the
paths of elimination of certain substances, inorganic compounds par-
ticularly, have been jumped at too hastily by a number of workers.
Thus, Meyer® has lately drawn attention to what appears to be an
unwarranted deduction by Mendel and Sicher® on the excretion of
barium in the urine,— unwarranted because the experimental data
offered by Mendel and Sicher failed to support it. With regard to the
elimination of foreign substances through the intestinal mucosa, it
seems that in many cases the experimental evidence did not support
the announced positive conclusion to that effect. In some of the
instances alluded to, such a conclusion appears to have hinged on
the assumption that the presence of a given substance in the intes-
tinal contents indicates the excretion of that substance through the
intestinal wall, —a deduction that is assuredly fallacious in view of
the fact that metallic and other elements have been detected in such
gastro-intestinal streams as saliva, bile, and pancreatic juice. Hence
the excretory function of the intestine can be accurately determined
in a particular case only after the passage of all such secretions into
the intestines has been prevented. Meyer is at present engaged in
such an investigation of the excretion of barium, and the companion
study described in this paper is a further outcome of our feeling here
that these questions need more thorough experimental study than
many of them have received.

Although evidence is not wanting that the intestines are capable

' BRAUER: Zeitschrift fiir physiologische Chemie, 1903, xl, p. 182.

2? MEYER: Journal of the American Chemical Society, 1907, xxix, p. 892.

® BerG and WELKER: Journal of biological chemistry, 1906, i, p. 371.

* MeverR: Loc. cit. Also, Journal of biological chemistry, 1906, ii, p. 46t.
°> MEYER: Journal of biological chemistry, 1906, ii, p. 474.

* MENDEL and SICHER: This journal, 1906, xvi, p. 147.

Liimination of Radium from Animats. 369

of eliminating various foreign substances introduced into the body,
the need of additional data obtained in the way suggested above is
very desirable. Moreover, since therapeutic measures, based on the
assumption that the gastro-intestinal mucosa eliminates foreign sub-
stances that are of no use to the organism, are frequently resorted to
by clinicians, a study of elimination through the stomach as well as
through the bile was also desirable. On account of the facility
with which very minute amounts may be sharply detected in the
tissues and fluids of the body, Professor Gies suggested the use of
radium for the study of our problem.

EXPERIMENTAL.

Methods. — Our experiments were carried out on dogs and rabbits
under ether narcosis. A permanent biliary fistula was established in
one dog (Experiment 2). In all the other animals bile was obtained
from temporary fistula. To prevent the passage of saliva into the
stomach and intestines during an experiment, a ligature was placed
at the cardiac end of the stomach. Ligatures were likewise placed
at the pylorus, and below the opening of the duct of Wirsung, also at
the junction of the large and small intestines in the dog. In each
rabbit the upper and lower ends of the cecum were closed by ligatures.

In each experiment radium bromide (1000 activity), in aqueous
solution, was injected subcutaneously. Bile and urine were usually
collected. The contents of the different sections of the gastro-
intestinal canal were carefully removed at the end of an experiment,
and examined for radium by the quadrant electrometer. A detailed
description of methods for the detection of radium in animal tissues
and fluids has already been described by Meyer.!. The same methods
were employed in this investigation.

Experiment 1. Dog; weight, about 1o kilos. ro mg. of radium bromide were
injected. About two hours later the dog was killed. The bile collected
from the gall bladder was radioactive.

Experiment 2. Dog; weight, about ro kilos. A permanent and complete
gall bladder fistula was established on June r1, at 4 P.M. The dog made
a very good recovery from the operation. On June 18th the stitches were
removed, and the fistula was apparently well healed. During the col-
lection of bile the dog was placed in the holder described in the
preceding paper.

1 Meyer: Journal of biological chemistry, 1906, ii, p. 462.

370 William Salant and Gustave M. Meyer.

June 25th, 2.30 P.M. 30mg. of radium bromide were injected into
the left leg. The bile obtained several hours later was radioactive.’
Urine and feces collected separately were likewise radioactive.

June 27th, 4 p.m. The dog was killed by chloroform narcosis. ‘The
stomach was found empty. Small pieces of the gastric wall were removed
for examination. The contents of the intestines as well as small pieces:
of the wall of the intestines were also separately tested for radium. The
results of the examination showed that neither the stomach nor the in-
testines were radioactive. The contents of the intestines, however, were
radioactive.

Experiment 3. Female dog; weight, 8 kilos. 10.30 Pp. M.: ether anesthesia.
The stomach was ligated at the cardiac and pyloric ends. The small
intestine was ligated at some distance below the opening of the common
duct. The large intestine was ligated at the junction of cecum and appen-
dix. The neck of the gall bladder was clamped and a cannula placed in
the common bile duct. 11.15 A. M.: 20 mg. of radium bromide were in-
jected into the right leg. 11.15-1.15 P.M.: 10 c.c. bile were collected.
I.15-5 P.M.: 7 c.c. bile were collected. Urine was removed from the
bladder at 5 p.m. The dog was killed at 6 P. M.

Both samples of bile as well as the urine were radioactive. The stomach
and contents separately examined failed to show the presence of radium.
The contents of the small intestine included between the pylorus and the
opening of the duct of Wirsung were active ; the wall of this part of the
intestine was inactive. The record of the test for radioactivity of the
remaining portion of the small intestine was unfortunately lost. The con-
tents of this portion proved to be radioactive. ‘The large intestine as well
as the contents separately examined were both inactive. The blood was
also examined and proved to be radioactive.

These experiments show, therefore, that in the dog the liver as
well as the kidney is able to eliminate radium. The behavior of the
gastro-intestinal tract is especially worthy of remark in this connec-
tion. Neither the stomach nor the large intestine excreted radium.
Elimination of this element seemed to take place, however, all along
the small intestine.

The advisability of studying elimination in the rabbit now sug-
gested itself, since it has been shown by Noel Paton and Bergmann
that herbivora behave differentiy from carnivora in respect to elimi-
nation. Thus Paton’s! experiments with phosphoric acid have
shown that none of it is excreted in the urine of goats, whereas

* Paton, N., DunLop and AITCHISON: Journal of physiology, 1900, xxv.
Laas, J? ? b
p: 212.

Elimination of Radium from Animals. Paes
Bergmann ! made the interesting observation that in dogs phosphoric
acid is almost entirely excreted by the kidney, while in sheep
elimination takes place through the intestines.

Experiment 4. Female rabbit; weight, 2.2 kilos. rz A.M.: ether narcosis.
A cannula was placed in the common bile duct.
as follows: Above the cardiac and immediately below the pyloric
ends of the stomach, below the duct of Wirsung, above and below the
At 12 M. 10 mg. of radium bromide were injected into the

Bile was collected during the following periods :

Ligatures were attached

cecum.
right leg.

[ 12-1 Dy Ie. me Cre,
Il I-4 6c“ 15 (73
IIT 4-8.30 ‘“ ra
IV 8.30-9 A.M. I5
Wrine obtained at rT PoM 10 *°
4 ee ime) ce

At 9 A.M. the next morning the rabbit was found dead. 15 c.c. of
bile were secreted after 8.30 p. M. the previous day. ‘The stomach and
intestines were then removed from the body of the animal and carefully
washed free from alladherent blood. ‘The contents of the various portions
of the gastro-intestinal canal were washed into clean casseroles, and in each
case thoroughly mixed before portions of them were placed in trays in
preparation for the tests for radium. ‘The results of the examination were

the following :

Bile, sample I, Radioactive
ca WL ke
a (a OCS de
a SW, sf
Urine; sample I, Not’ -*

‘* II, Radioactive

Stomach, Radioactive

Contents of “ , Not radioactive

The part of the small intestine
the pylorus and beyond the duct of Wirsung was not radioactive.
contents of this part were radioactive.
well as the contents separately examined were radioactive.

included between ligatures placed at

The
The rest of the small intestine as
Neither
The large intestine was inactive,

cecum nor contents were radioactive.
while the contents were radioactive.

17BERGMANN : Archiv fiir experimentelle Pathologie und Pharmakologie, 1gor,
xlyii, p. 77.

372 Willam Salant and Gustave M. Meyer.

Experiment 5. Male rabbit (gray) ; weight, 2 kilos. Ether narcosis. Cannula .

in the common bile duct. Ligatures were placed as follows: At the
cardiac and pyloric ends of the stomach, and above and below the
cecum. ‘The bladder was emptied. 15 mg. radium bromide were then
injected into the left leg at1 p.m. Bile was collected as follows:

I=? PN co (C2Ge
aan oR ans eas

At 9 A.M. next day the rabbit was found dead. 12 c.c. of bile were
secreted after 5.15 p. M. of the previous day. The stomach and intestines
were removed from the body of the animal and treated as described
above. .

The bile as well as the urine collected at the end of the first hour after
the injection of radium bromide was only slightly radioactive. Radio-
activity of the bile and urine secreted during the next three hours was,
however, very marked. Another sample of each of these secretions
obtained before the death of the animal was only slightly radioactive.
The stomach as well as its contents, separately examined, failed to show
any radioactivity. Examination of the small intestine for radium likewise
proved negative, but its contents were markedly radioactive. Neither
cecum nor large intestine showed radioactivity, while the contents in
both cases were only slightly radioactive. The blood which was cbtained
from the heart was also tested. ‘The results were negative.

Analysis of the data obtained in our experiments with rabbits indi-
cates that in these animals elimination of radium invariably takes
place through the liver, the kidneys, and the small intestines, after its
introduction into the circulation. Moreover it is worthy of remark in
this connection, that the liver and the kidneys are apparently equally
efficient as organs for the excretion of radium, since, as was shown
in experiment 5, the elimination of this element began in both organs
at about the same time and probably continued for equal periods.
Quite different was the behavior of the stomach, the cecum, and the
large intestine with regard to the excretion of radium. In neither of
the two rabbits experimented on was there any indication that radium
had been eliminated through the stomach. Although the wall of the
stomach was radioactive, the contents of the stomach failed to show
the presence of this element. The cecum presented some variation
in this respect. Radioactivity of the contents in one rabbit was slight,
but was altogether absent in the other, —an indication that this part of
the intestine, at least in normal rabbits, is variable in its activity as
an organ for the elimination of radium. The large intestine of the

Elimination of Radium from Animats. aya

rabbit seems likewise to vary somewhat in this regard in different indi-

viduals. Thus, while radioactivity of the contents of this part of the
intestines was very marked in one rabbit, it was slight in the other,
suggesting that this part of the gut does not eliminate radium with
equal facility in all rabbits. If the results of our observations on dogs
and rabbits are now compared, the following interesting resemblances
and differences in the manner of elimination of radium in these ani-
mals may be seen.

By referring to the table on page 376 we see that the liver as well
as the kidney participated in the elimination of radium in the dogs and
rabbits. The gastro-intestinal canal, however, exhibited differences
in this connection. While the stomach in each of the animals
- failed to excrete radium, elimination invariably took place through
the wall of the small intestine. The large intestine, on the other
hand, acted dissimilarly in the dog and rabbits. Thus, in the dog
the portion of the intestine which, it will be remembered, is the
organ for elimination of iron, manganese, and certain other ele-
ments, did not excrete radium. In the rabbits, on the contrary, elimi-
nation through the large intestine did take place, although the rate
of elimination seemed to vary in different individuals. If deductions
based on the evidence presented on the foregoing pages are warranted,
we may conclude that the elimination of radium takes place chiefly
through the liver, the kidneys, and the small intestine, and, to a
lesser extent also, through the large intestine in some of the herbivora.

At this stage of the investigation it occurred to us that a study of
the elimination of radium from nephrectomized rabbits would be very
desirable. Although considerable experimental evidence has accu-
mulated to indicate that even when the function of the kidney is un-
impaired, the elimination of foreign materials through the digestive
tract and the fluids entering it, may take place, the study of the pas-
sage of substances through these channels in disease of the kidney
or when the urinary passages have been blocked, has made but little
progress. The few available experimental data on this subject suggest,
however, that organs cther than the kidney may by way of compensa-
tion eliminate foreign substances from the body under such conditions.
Claude Bernard! in his experiments with potassium ferrocyanide has
shown that this substance which is normally excreted through the
kidney may pass into the saliva if the renal arteries or the ureter

1 BERNARD, CLAUDE (quoted by ACHARD et Lopez): Séances et memoires
de la Société Biologique, March 15, 1902.

374 Willan Salant and Gustave M. Meyer.

have been ligated. In this connection may also be mentioned the
fact that in advanced cases of nephritis, urea may be eliminated
through the digestive tract, the lungs, the skin, and in the lachrymal
and other secretions. It was of interest, therefore, to study whether
those parts of the gastro-intestinal canal, such as the stomach and
cecum, which fail to excrete radium in normal rabbits, might eliminate
this element if the kidneys were removed. Our experiments were car-
ried out on two healthy, full-grown rabbits on which the operation
for double nephrectomy was performed under ether narcosis and
elimination of radium studied in the way indicated in the previous
experiments.

Experiment 6. Male rabbit ; weight, 2.24 kilos. Both kidneys were removed
by the abdominal route, and a cannula was introduced into the common
bile duct immediately afterward. Ligatures were placed at the cardiac
end of the stomach, around the duodenum, immediately below the pylorus,
at the junction of the cecum and small intestine, and at the junction of
cecum and large intestine. to mg. of radium bromide were then in-
jected into the left leg. The bile collected during a period of thirty
minutes after the injection of radium was not radioactive. The next
sample collected during the succeeding hour was active. A third sample,
obtained nine hours after the administration of radium bromide, was also
active. ‘The rabbit died ten hours after double nephréctomy had been
performed. The stomach and the various parts of the intestines were
carefully taken from the body of the animal, washed free from adherent
blood, the contents removed and put into zinc trays, which were treated
as previously described. In each case the organ and its contents were
separately examined for radium.

The tests of the various parts of the digestive tract for radium have
shown the following results: Stomach as well as contents were slightly
radioactive. Both the small intestine and its contents were also radio-
active, whereas the cecum as well as its contents failed to show the presence
of radium. The large intestine and contents were both radioactive. The
results obtained show therefore that elimination through the liver is not
accelerated as a result of removal of both kidneys, for the bile obtained
during a period of thirty minutes after radium was administered was not
yet radioactive. ‘The results obtained with stomach contents, on the
other hand, show that there is a tendency to vicarious elimination, for the
contents were slightly radioactive. In this regard the. intestines behaved
the same as those of normal rabbits. The contents of both the large and
small intestines showed the presence of radium, whereas the contents of
the cecum, as in both normal rabbits, were free from radium.

Elimination of Rowiien from Animals. 375

Experiment 7. Female rabbit; weight, about 2.5 kilos. Double nephrectomy
was performed by the abdominal route. Ligatures were placed at the
cardiac and pyloric ends of the stomach, at the junction of small in-
testines and cecum, and at the junction of large intestines and cecum.
A cannula was introduced into the common bile duct. At 12.35 P.M.
Io mg. of radium bromide were injected. Bile removed at 1.10 P.M.
was not radioactive. Another sample obtained between 1.10 and 2.10 P. M.
was radioactive. At 5.30 P.M. the rabbit was bled to death. The bile
removed shortly before the death of the animal was still radioactive. ‘The
various sections of the gastro-intestinal canal were removed and treated
as in the other experiments. Tests for radium showed that neither
stomach nor contents were radioactive. The small intestine was likewise
inactive, but its contents were radioactive. Examination of the large
intestine as well as the contents gave in both instances negative results.
In this experiment, therefore, the stomach failed to assume any vicarious
function. Moreover, the large intestine, which in all the other rabbits
eliminated radium, has in this case behaved in a rather anomalous manner.
Contrary to expectations, the excretion of radium was entirely inhibited.

The results of our experiments with nephrectomized rabbits do not
answer the question whether the vicarious elimination of radium
might be expected to occur in most rabbits. The recent work of
Meltzer and Lucas! furnishes strong evidence, however, that the
gastro-intestinal canal fails to eliminate more readily, after the re-
moval of the kidney, certain substances introduced into the body.
Thus, when magnesium sulphate was administered to nephrectomized
rabbits, in much smaller doses than those given without special effect
to normal rabbits, deep anesthesia was produced. It is probable,
therefore, that while vicarious elimination may take place through
the walls of the alimentary canal after extirpation, or in diseases, of the
kidneys, in the cases of various organic substances produced within
the body as a result of metabolism, or when introduced from without,
it apparently fails to occur in the cases of certain inorganic sub-
stances, which normally leave by other channels. It is quite possible
that the liver and small intestine, which normally excrete radium,
suffice to meet this need of the body after removal of the kidney.

Finally, we desire to call attention to the study of elimination in the
urinary bladder. Although some experimental evidence on absorption
from the bladder exists, we could not find any reference in the litera-

1 MELTZER and Lucas: Proceedings of the Society for Experimental Biology
and Medicine, 1906, iv, p. Io.

376 Wilham Salant and Gustave M. Meyer.

ture to an investigation bearing on excretion through this organ. It
seemed to us, therefore, that the question ought to be put to an
experimental test. Our results indicate that radium was not ex-
creted through the wall of the urinary bladder.

A summary of our results for radioactivity is appended.

TABLE GIVING THE RESULTS FOR RADIOACTIVITY OF VARIOUS
PARTS IN EXPERIMENTS 2-7.

: Double nephrecto
if T <
Normal dogs. Normal rabbits. mized rabbere!
Experiment No. 2 3 4 5 6 7
|
; ; ; Sli 2
Stomach . .-- Inactive | Inactive | Active Inactive ieee Inactive
: re) ; : Sli :
Stomach contents Siac Inactive Inactive | Inactive Be oe: Inactive
Intestine . . . Inactive
Contents of the
intestine . . Active
Biever as Active Active Active Active Active Active
NeGeS 3) <3, i". Active
BIGOd ee) oo te Active Active Si os Inactive Slightly
active
IEye 5 5 os Active
Wirine? 120% & Active Active Active Active
Small intestine . seer Inactive Active Inactive | Active Inactive
Contents small
intestine . . ey Active Active Active Active Active
Large intestine. Soe Inactive | Inactive | Inactive | Active Inactive
Contents large +e
intestine .. avaie'e Inactive | Active Slightly Active Inactive
Active
Cecum o.) there Be eta Inactive | Inactive Inactive | Inactive
Contents of ce- Slightly
GUAM eer” sane sisiais ae Inactive active Inactive | Inactive

SUMMARY OF CONCLUSIONS.

In dogs and rabbits the kidney, the liver, and the small intestine
eliminate radium. In normal rabbits elimination also takes place
through the large intestine. The passage of radium through the
wall of the large intestine is probably slower than through the wall
of the small intestine. Elimination of radium into the cecum of the
rabbit is slight and in some individuals may fail altogether.

Elimination of Radium from Animals. 377

In a nephrectomized rabbit the elimination of radium takes place
through the small intestine and through the liver. The rate of radium
excretion from the livers of rabbits is not affected by removal of the
kidneys. Ina nephrectomized rabbit elimination of radium through
the large intestine is uncertain; in such a rabbit radium is not elimi-
nated through the cecum, but may pass through the stomach wall.

The channels of elimination for radium appear to vary in different
species and in individuals of the same species.

After removal of both kidneys from a rabbit there is no compen-
satory elimination of radium through those parts of the digestive tract
which are not concerned in its elimination when the kidneys are intact.

We are indebted to Prof. William J. Gies for numerous suggestions.

ON THE CHEMICAL NATURE OF PARANUCLEG-—
PROTAGON, A NEW PRODUCT FROM BRAIN.

By MATTHEW STEEL anp WILLIAM J. GIES.

[From the Laboratory of Biological Chemistry of Columbia University, at the College
of Physicians and Surgeons, New York.)

CONTENTS.
Page
Introduction . . . i 5 &, Ren oS
Review of the pened aaa ane cane aE iar ane Lelli on protesan ane para-
nucleoprotagons. 3. s° ie 0 6 Se se ema)
On protagon .. . rere mere mre ees ey SG Ge Sell
Preparation of Eemtucleoatecen anieare POLS ate ies 37

Cleavage process for the separation of paranuclein and srotagen ‘fon para-
nucleoprotagon, and a new method for the isolation of protagon . . . . . 383

What is the percentage proportion of phosphorus in protagon? . . . . . . 384
Paranuclein from paranucleoprotagon- . . . - . « © « + © ie eRueRaemeemees
General.deductions <9. 20.03 3 (sa «ss nh so) ne

Experimental :
Preparation of paranucleoprotagon .. . é MP i a
Preparation of protagon and paranuclein fed paraawclen protean MO tera el
Cleavage process inaleohol 2) 5) ts) te ere
Isolation of protagon by the classical process . . ... ... +. - 393
Protagon filtrate . . . PPPs oc Gc St
Paranuclein and its Palorofora washings: ele ® » © lo. nes
Summary of quantitative'data . 9... = 3 =. 5s) = 2 -) )o) ene
Summary of general conclusions . -. . «=... s « © & =» » = 0/=sny ese

INTRODUCTION,

WO years ago Posner and Gies published numerous results of

an extended study of brain protagon.!. They concluded that
protagon, whether prepared by the classical methods or by Cramer’s
coagulation process, was without physical or chemical definiteness, .
7. e., that it was merely a mzxture of substances. They made it evi-
dent, also, that, even with the use of exactly the same method of
preparation, products of particularly divergent composition were

1 POSNER and Gigs: Journal of biological chemistry, 1905, i, p. 59-
378

ee

Ox the Chemical Nature of Paranucleoprotagon. 379

obtained, when the physical conditions attending the brain ex-
tractions were slightly modified.

Among the deductions drawn by Posner and Gies from their proof
of the heterogeneity of protagon was one pertaining to Ulpiani and
Lelli’s paranucleoprotagon.2 They said regarding the latter: 3

1 Shortly after the conclusion of the experiments described in this paper,
LOCHHEAD and CRAMER published a few results of a study ‘‘On the phosphorus
percentage of various samples of protagon,”’ which led them to conclude, contrary
to the recent deductions by PosNER and Gigs (doc. cit.), “that protagon is an
individual substance of a well-defined chemical composition” [Bio-chemical
journal, 1907, ii, p. 350 (June 20) ]. One looks in vain, however, through the
paper by LocHHEAD and CRAMER, for evidence warranting the conclusion that
protagon is zofa mixture. On the contrary, their results seem to support unmis-
takably the position taken by PosNER and Gis. Thus, on recrystallization, all
the products except one that LOCHHEAD and CRAMER referred to, lost phosphorus
in marked degree after each such treatment. When the phosphorus contents of
their protagon products were lowered by recrystallization to the percentage amount
that appeared to them to be about right, however, they arbitrarily discontinued in
each case the recrystallization process, in spite of the fact that repetition of it
promised to decrease further the proportionate contents of phosphorus. Further-
more, they used the process of recrystallization from glacial acetic acid as a method
of purifying one of the protagon products that seemed to be among their best,
apparently without being aware of the very significant fact that Kocu has em-
ployed the same method to remove from crude protagon every trace of phosphorus-
containing substance or substances, prior to the ultimate isolation of one of the
leading compounds contained in the protagon mixture, viz., phrenosin (pseudo-
cerebrin, cerebron, cerebrin). (See Gies: Journal of biological chemistry, 1906,
li, p- 159.)

The foregoing remarks in this footnote are made by us on the assumption
that the products analyzed by LocHHEAD and CRAMER were sufficiently like
typical protagon to warrant the term protagon as the designation of them.
LOCHHEAD and CRAMER appear to be satisfied on this point, but the reader of
their paper will find no evidence in it of any effort on their part to prove the
validity of their new methods for the preparation of protagon. Thus, the only
data that were presented regarding “ protagon F,” for exampie, and which LocH-
HEAD and CRAMER appeared to think showed that it was protagon, were the facts
that it was white, that it contained 1.18 per cent of phosphorus, and that it had
been obtained from brain by extraction with ether, not one of which characters is
a differential quality of protagon. We believe the paper by LOCHHEAD and
CRAMER has confused the situation without adding anything material to the facts
inthe case. The results of a critical examination of LOCHHEAD and CRAMER’S
paper was published by Gres in the last number of the Journal of biological
chemistry (iii, p. 339.)

2 ULPIANI and LELLI: Gazetta chimica italiana, 1902, xxxii, p. 466.

8 PosNER and GiEs: Loc. cit., p. 109.

380 Matthew Steel and William F. Gites.

“ The new proteid, paranucleo-protagon, that was described by Ulpiani and
Lelli (1902) as a definite substance, must be regarded as a mixture of products.
It is our intention to begin at once an inquiry into this matter.”

The intention to make an experimental study of paranucleoprotagon
could not be gratified until quite recently. Meanwhile no one ap-
pears to have confirmed or contradicted the conclusions of Ulpiani
and Lelli, nor have the discoverers of paranucleoprotagon recorded
any further observations regarding it that we know of. The follow-
ing digest of Ulpiani and Lelli’s paper! presents the essentials of the
recorded data pertaining to paranucleoprotagon.

REVIEW OF THE REMARKS AND THE WoRK OF ULPIANI AND LELLI
ON PROTAGON AND PARANUCLEOPROTAGON.

Ulpiani and Lelli on protagon.— Ulpiani and Lelli stated that they
made extensive studies of protagon, during the course of which they
came to the conclusion that protagon did not exist in the brain in a
free state, but occurred there united with a protein substance, which
complex compound, after long and arduous endeavor, they finally
isolated and separated by cleavage with alcohol into its (chief?)
components, paranuclein and protagon, and accordingly named para-
nucleoprotagon.? They referred to Walter’s? supposition that ichthulin,
obtained by him from carp eggs, was a protein-protagon compound,
but aiso alluded to Walter’s inability to show that protagon was con-
tained in it.

Ulpiani and Lelli stated that in their studies of protagon they
prepared that material by the Gamgee and Blankenhorn process, and
also remarked that they obtained results in that connection which
were in close agreement with those published by Gamgee and Blan-
kenhorn.! It was their study of protagon obtained by this method and

1 ULprANI and LELLI’s paper was published in Italian (/oc. cz¢:), and no complete
review of it has ever appeared. For this reason, probably, their important paper

has received comparatively little attention. For the same reason our review of it
here will be more detailed than would be desirable otherwise.

2 Their paper was issued in 1902, a long time after the publication of WORNER
and THIERFELDER’S (Zeitschrift fiir physiologische Chemie, 1900, xxx, p. 542),
but shortly before the appearance of the paper by Lesem and Gigs (This journal,
1902, viii, p. 183; also Gres and collaborators: Biochemical researches, 1903,
i, Reprint No. 11).

8’ WALTER: Zeitschrift fiir physiologische Chemie, 1891, xv, p. 477.

* No data were given. They said nothing about phrenosin (pseudocerebrin)
in this connection, and did not appear to be acquainted with GAMGEE’s second

On the Chemical Nature of Paranucleoprotagon. 381

the observation that their protagon products dissolved in chloroform,
and could be precipitated! therefrom with ethyl acetate, acetone, or
alcohol, that led them to use chloroform for the purpose of extracting
protagon from brain, and that resulted finally in their unexpected dis-
covery of paranucleoprotagon, the compound in which, according
to Ulpiani and Lelli, all the protagon that occurs in the brain is
combined.

Ulpiani and Lelli’s remarks on protagon, and on their experiments
with it, lead one to think that they simply skimmed over the surface
of the history and facts relating to protagon? and accepted without
question the accumulated mass of superficial statements regarding
it. As usual, there was the conventional allusion to Thudichum for
his obstinacy in adhering to his well-founded conviction, on the basis
of the results of his well-ordered experiments, that protagon was a
mixture of substances. The important work of Worner and Thier-
felder on phrenosin (cerebron), which gave strong evidence of the
heterogenous nature of protagon, was ignored. Ulpiani and Lelli
stated that protagon was without doubt a definite chemical individual,
but they offered no new facts bearing on the question.

contribution on the subject. Evidently, like GAMGEE and BLANKENHORN’S
products, their protagons contained phrenosin (pseudocerebrin). See GIES:
Journal of biological chemistry, 1906, ii, p. 168.

1 Whether completely or not does not appear to have been determined.

2 Thus, they laid special stress on their observation that protagon is soluble in
chloroform, something they declare had not been noted before. Evidently they
were not familiar with the well-known prior work of WORNER and THIERFELDER
(Zeitschrift fiir physiologische Chemie, 1900, xxx, p. 542), in which chloroform
was employed with alcohol to dissolve from protagon the constituents associated in
it with phrenosin (cerebron) (GriEs: Journal of biological chemistry, 1906, ii, p.
159). Phrenosin itself was found by WORNER and THIERFELDER to be soluble
in hot chloroform, and it seems very probable that phrenosin is appreciably soluble,

also, in co/d chloroform solutions of the substances which, with phrenosin, compose
_ protagon. ULPIANI and LELLI stated that protagon is precipitated from its
chloroform solution by alcohol or acetone. W0ORNER and THIERFELDER used
mixtures of alcohol-chloroform (50 per cent) to separate phrenosin from protagon
by fractional precipitation from cooled extracts (made at 45-50° C., but in which not
all the protagon dissolved). Chloroform-acetone solution was used for purification
by recrystallization. That the significance of these and further observations by
WORNER and THIERFELDER and others was not appreciated by ULPIANI and
LELLI is apparent not only from the failure on their part to note the above-mentioned
use to which the former investigators put chloroform, alcohol, and acetone in
perfecting the mechanical separation of phrenosin from protagon, but also from
various other oversights that become evident as one reads their paper.

382 Matthew Steel and William F. Gres.

Ulpiani and Lelli’s preparation of paranucleoprotagon. — Paranucleo-
protagon was prepared by Ulpiani and Lelli as follows: Horse
brains,! mechanically freed from extraneous matter, were extracted?
with chloroform.? Extraction at 45° C. favored stratification of the
chloroform, without detrimental effect on the paranucleoprotagon ; *
at room temperature, however, the brain-chloroform mixture was too
intimate to permit of ready mechanical separation.® The filtered
extract® was treated with an equal volume of acetic ether, which
caused the separation of an abundant precipitate (paranucleoprotagon,
with impurities). This precipitate’ was filtered off, extracted with
ether,® filtered off again, and then extracted in a Soxhlet apparatus,
first with ether and finally with chloroform.* The remaining product
was dried zz vacuo over sulphuric acid, pulverized and analyzed.’®
Single determinations of the percentage proportions of four elements

1 Whether macerated or not was not made clear.

2 Length of time was not indicated.

Proportion was not mentioned.

How this was ascertained the reader is obliged to guess.

In the unstated proportions employed.

Whether freed from the water layer or not was not specified.
Its qualities were not referred to.

8 Conditions were not stated.

° No explicit statement was made as to the exact purpose of this treatment,
how long it was maintained, nor how thoroughly it was effected. Presumably it
was intended to remove adherent cholesterin, etc., and was continued until no more
substance could be dissolved out, but the reader finds nothing to reassure him upon
either of these points.

The insolubility of paranucleoprotagon in chloroform at this point in the process
is in marked contrast to the solubility of the product in chloroform at the
beginning. ULprani and LELLt explain this by the assumption that the product
was soluble in the original chloroform extract because of the presence there of
lecithin, cholesterin, etc., —in short, that paranucleoprotagon is soluble in chloro-
form solutions of lecithin, cholesterin, and similar bodies extractable from brain
by chloroform, but is insoluble in pure chloroform. No experimental data were
offered to substantiate this very plausible explanation, however.

10 The physical properties of the substance, moist or dry, were passed over in
complete silence by the authors. Whether it was white or black, pasty or fibrous,
for example, appeared to be of no significance to ULPIANI and LELLI. That it
could be pulverized when dry was the only hint as to its consistency. No state-
ment was made by them regarding either the absolute or the proportionate quantity
of paranucleoprotagon obtained from brain, —a particularly striking omission, in
view of the assumption by Utprani and LELtt that all the protagon of the brain
occurs there in the form of paranucleoprotagon.

a an » ©

-

On the Chemical Nature of Paranucleoprotagon. 383

were made in but one preparation of paranucleoprotagon,! the results
of which are appended:

C H N P
60.79 8.74 6.20 1.62

The product prepared in the manner indicated was called para-
nucleoprotagon, because Ulpiani and Lelli believed that it was a
compound that yielded paranuclein and protagon, by cleavage with
alcohol. Their method of preparing protagon from their paranucleo-
protagon product is given briefly below.

Ulpiani and Lelli’s cleavage process for the separation of paranuclein
and protagon from paranucleoprotagon, and their new method for the
isolation of protagon. — The dry paranucleoprotagon was treated with
85 per cent alcohol at 45° C., for the purpose of effecting cleavage
of the product.2. After cooling to 0° C., the mixture was filtered.’
“ After this treatment,” said Ulpiani and Lelli, “the substance which
at first was insoluble in chloroform became in large part soluble in
that medium.” Next they treated in a Soxhlet apparatus with chlo-
roform, the matter that was insoluble in the cold (0° C.) alcohol, and
obtained, in the cumulative chloroform extract,* the major part of the
substance thus handled. On treating the chloroform extract with
acetic ether, a white substance was precipitated ® which physically
and chemically was regarded as being identical with protagon.’
Analysis of the only product mentioned ® by Ulpiani and Lelli gave
the following percentage results :

Cc H N P

66.67 10.45 2.64 1.38

66.67 10.54 2.51 1.24
Average . . 66.67 10.50 257 E. 3

1 There is no recorded evidence that ULPIANI and LELLI made more than one
preparation of paranucleoprotagon.

2 Nothing was said of the proportions of alcohol used, of the length of the period
of treatment, or of the solvent effect.

8 No statement was made as to the length of time the refrigeration process was
maintained. The filtrate received no attention, and was apparently discarded; at
least no information was given regarding anything contained in it.

4 The portion in the receiver was kept hot, of course. The extraction was
continued apparently as long as anything dissolved.

5 The insoluble portion, “ nuclein,” is referred to farther on.

6 Precipitates were also obtained with alcohol and acetone.

7 What the filtrate contained was not suggested. There was no intimation that
the filtrate received any attention.

8 Whether dried to constant weight and how were not indicated.

384 Matthew Steel and William F. Gies.

The only additional chemical inquiry into the protagon-like qualities
of this product was the subjection of the latter to hydrolysis for
twenty hours in 7.5 per cent hydrochloric acid, with a determination
of the presence of reducing material! among the cleavage products,
—a reaction given by phrenosin or any cerebroside or cerebroside
mixture, and not at all characteristic of protagon among brain educts.

Ulpiani and Lelli appear to have done nothing to assure themselves
and the reader that protagon can be prepared by the method employed
in isolating this product from what they call paranucleoprotagon.?

What is the percentage proportion of phosphorus in protagon? —
The “ protagon” obtained by Ulpiani and Lelli contained a compara-
tively large proportion of phosphorus, as may be seen from the figures
in the summary on pages 386 and 387.

The general average of the figures on pages 386 and 387, excluding
Ulpiani and Lelli’s, is 1.07 per cent. It might be inferred from this
that Ulpiani and Lelli’s substance with 1.31 per cent of phosphorus
could not have been protagon. A glance at the data given in the
footnotes pertaining to the summary referred to makes it evident, how-
ever, that our average value for phosphorus content (1.07 per cent),
although it agrees perfectly with the classical Gamgee and Blankenhorn
figure, was derived arbitrarily by excluding all results for phosphorus
content in protagon that, for one reason or another, departed fairly
far from 1 per cent.? If, as Kossel and Freytag suggested, and as
Cramer also thought, there were many protagons (homoprotagons, as
Cramer proposed that they be designated) differing widely in compo-
sition, the average figure referred to, 2. ¢., 1.07 per cent, would be of
little or no differential value; protagons might then contain much more
or much less phosphorus than that (per cent) according to circum-

1 Tested with Fehling’s solution.

* The method employed by ULPIANI and LELLI for the isolation of protagon at
this point is analogous to LOCHHEAD and CRAMER’S process for the preparation of
their “ protagon F,” which was extracted from brain with doz/7zg chloroform, and,
after cooling the extract, was precipitated from the latter by the addition of three
volumes of ether: LOoCHHEAD and CRAMER found, however, that their product
contained 1.18 per cent of phosphorus, but learned nothing else chemically about
it. See footnote, p. 379.

* We have done this, against our sense of the fitness of things, merely to repeat
the method of CRAMER (and other supporters of the hypothesis that protagon is
something definite), and then to show how misleading is such a one-sided presen-
tation of the case, and how meaningless is the conclusion, based on it, that Locu-
HEAD and CRAMER have recently drawn regarding the chemical individuality of
their products. See footnote, p. 379 of this paper.

On the Chemical Nature of Paranucleoprotagon. 385

stances, and Ulpiani and Lelli’s product might be (a) protagon, despite
the decided disagreement of their figure for its phosphorus content
with the average for the same in most protagons as already given
(1.07 per cent). If, however, there is only one protagon, then it is
obvious that we cannot tell who ever had a pure sample of it or what
its percentage content of phosphorus is, for the data in the footnotes
below the summary on the next two pages are quite as significant
as the purely arbitrarily selected data to which preference has been
shown by their placement in the table as particularly representative
figures (following the fashion of most undiscriminating writers on
subjects pertaining to protagon), in which event Ulpiani and Lelli
have no evidence whatever, from their data for percentage phosphorus
content, upon which to base their claim that their product with 1.31
per cent of phosphorus was protagon. But if protagon is always
merely a mechanical mixture of substances, as we think must be con-
ceded, the remarkable discordance among the figures given in the
summary and corresponding footnotes (pp. 386-387), is an inevitable
result of that state of invariable heterogeneity, and, under such con-
ditions, Ulpiani and Lelli’s figure for phosphorus content (1.31 per
cent) implies as much or as little as any one else’s, so far as protagon
characters are concerned. It seems to us that, in the present state
of our knowledge in this relation, 7. ¢., that mechanical fractions
of protagons may contain from 0.I per cent to about 2 per cent of
phosphorus, an analytic result, within that range, for percentage
content of phosphorus in a product is altogether too uncertain in
significance to base upon it with confidence a claim that the product
in question is or is not protagon.

It is significant that Kossel and Freytag obtained from an ether
extract (containing considerably more lecithin and similar material
than an alcoholic extract) a protagon product having a phosphorus
content equal to 1.35 per cent — one apparently having a greater
admixture of lecithins than that commonly present in the protagon
obtained from an alcoholic extract. In their use of chloroform and
ethyl acetate, for the isolation of protagon from paranucleoprotagon,
Ulpiani and Lelli probably brought about a similar result.

Ulpiani and Lelli’s paranuclein obtained from paranucleoprotagon. —
Having separated completely what they thought was protagon from
what they termed paranucleoprotagon, Ulpiani and Lelli concluded
that that portion of their product which had been filtered from cold
(0° C.) alcohol, and which likewise remained insoluble in the chloro-
form that dissolved away the liberated protagon, was a paranuclein.

386 Matthew Steel and Wiliam F. Gies.

PERCENTAGE AMOUNTS OF PHOSPHORUS IN TYPICAL PROTAGON PRODUCTS.

PHCDEEICNS BU 9 2.0? gy set as) 4 Re LO Rappel so ce my ooo ctl oe
Gamgee and Blankenhorn? 1.07 Chittenden and Frissell’ . 1.12
Dhurcienutan see)? ss PCT.e0 Zuelzer*’. ts, sn,
Baumstagk®” 210) ae “Tos Noll? 34 ane
Kossel and Freytag’ . . 0.97 Gulewitsch® -.) )... SMe

1 LIEBREICH: Annalen der Chemie und Pharmacie, 1865, cxxiv, p. 29. One
product obtained from human brain contained 1.5 per cent of phosphorus. Com-
parable results obtained by FrEmy and others before LIEBREICH’S publication are
purposely omitted from the above table.

2 GAMGEE and BLANKENHORN: Zeitschrift fiir physiologische Chemie, 1879,
iii, p. 279. The authors obtained by recrystallization, after continuous warming in
ether, one product containing only 0.72 per cent of phosphorus. This fall in per-
centage content was ascribed erroneously to decomposition, but was certainly due
merely to physical fractionation of the mixture of products of which protagon is
composed. Those who favor the hypothesis that protagon is a chemical individual
never mention these points.

8 THUDICHUM: Annals of chemical medicine, 1879, i, p. 258. THUDICHUM
obtained by simple fractional recrystallization from 85 per cent alcohol and wash-
ing with ether, protagon products having as much as 2 per cent of phosphorus,
and as little as o.1 per cent of that element. The range of fluctuation in the pro-
portions of phosphorus in the products obtained from protagon by THUDICHUM
in his fractional recrystallization process, including those in mother liquors, was
0.1 per cent to 2.9 per cent. GAMGEE obtained from “impure” protagon, by
repeated recrystallization from 80 per cent alcohol, a fraction containing only 0.08
per cent of phosphorus. This product was called pseudocerebrin (Gamgee: A
textbook of the physiological chemistry of the animal body, 1880, i, p. 441).

* BAUMSTARK: Zeitschrift fiir physiologische Chemie, 1885, ix, p. 145.

5 KossEL and FREYTAG: Jdzd., 1893, xvii, p. 431. The figure given above for
phosphorus content is that commonly quoted for their “dest” product, which also
contained 0.51 per cent of sulphur. Another product had 1.06 per cent of phos-
phorus. A third contained 1.35 per cent of phosphorus and 0.88 per cent of sulphur.
Those who support the idea that protagon is something definite chemically always
ignore this evidence against that conception.

6 RUPPEL: Zeitschrift fiir Biologie, 1895, xxxi, p. 86.

7 CHITTENDEN and FRISSELL: Science, 1897, v (N. S.), p. gor (Proceedings of
the American Physiological Society, May, 1897). Frissell obtained results on
subjecting his protagons to fractional recrystallization from 85 per cent alcohol at
45° C. that were similar to THUDICHUM’S. CHITTENDEN assumed, however, that
FRISSELL’s results were due to ‘“‘ decomposition ” of the protagon instead of to its
mechanical partition, regardless of GAMGEE’s previous statement that “ pure
protagon is remarkably rebellious to the action of even dozing alcohol, though
that action be continued for hours.” (GAMGEE: A textbook of the physiological
chemistry of the animal body, 1880, i, p. 429.) ;

8 ZUELZER: Zeitschrift fur physiologische Chemie, 1899, xxvii, p. 255. One of
the products had a phosphorus content of 0.72 per cent. The figure above was the
only one for the product of the two referred to. ZUELZER regarded the product
containing only 0.72 per cent of phosphorus as that of a “decomposed” protagon
—merely a guess. Cramer might have mentioned this as one of the hypothetical
‘‘homoprotagons.” See footnote 13, next page.

® NoLL: Zeitschrift fiir physiologische Chemie, 1899, xxvii, p. 376. The carbon
content of NoLv’s product was unusually high — 67.15 per cent. See footnote 13,
next page.

10 GULEWITSCH, cited by NoLi: Jééd.

On the Chemical Nature of Paranucleoprotagon. 387

PERCENTAGE AMOUNTS (continued).

Uipiani and Lelli*® . . 4.31 Posner ant Gies* .. .- - 10.93
Eesem and. Gies’** .. . -: - 1.22 Lochhead and Cramer® . 1.01
ROME os. Se ha; Oe EOS

0, ULPIANI and LELLI: Gazetta chimica italiana, 1902, xxxii, p. 466.

12 LrsEM and GiEs: This journal, 1902, viii, p. 183. Fractional recrystalliza-
tion from 85 per cent alcohol at 45° C. yielded protagon products with phosphorus
contents ranging from o.12 per cent to 1.23 percent. Fractional products in the
mother liquors contained proportions of phosphorus ranging between 0.85 per cent
and 2.59 per cent. The authors agreed with THuUDICHUM and disagreed with
CHITTENDEN in their interpretation of these results. (Gries and collaborators:
Biochemical researches, 1903, i, Reprint No. 11.)

18 CRAMER: Journal of physiology (English), 1904, xxxi, p. 31. CRAMER
stated (p. 34) that in one specimen, which was prepared zz the same way as the
others, he got “different analytic results with regard to the percentage of carbon.”
Its carbon content was 64.7 percent, that of the normal product was 66.3 per cent;
its phosphorus content was 1.16 per cent, that of the normal product is given above.
CRAMER added: “It is probable that this substance is another representative of
the group of protagons (!) assumed to exist by KosseEt (see footnote 5, p. 386),
which, in order to distinguish them from the protagon usually described and
analyzed, might well be called homoprotagons. . . . A substance belonging to this
group has only once been isolated and described, by NOLL (Loc. cét.). It contained
67.15 per cent of carbon.” Inshort, CRAMER would have it that, whether the prod-
uct under examination has much more or much less carbon (per cent) than that in
protagon, it is a “homoprotagon.” On the other hand, the greater probability that
protagon is nothing definite chemically, and that his “ homoprotagon ” was evidence

‘in that direction, were ignored by CRAMER. If CRAMER’S suggestion were ac-
cepted, the protagon products prepared by WORNER and THIERFELDER (Zeitschrift
fiir physiologische Chemie, 1900, xxx, p. 543) and having carbon contents ranging
from 62.37 per cent to 64.62 per cent, were “‘homoprotagons.” But each of WOr-
NER and THIERFELDER’s products yielded much phrenosin (cerebron) by simple
fractional recrystallization. The proportion of phosphorus in their protagons was
not determined. However, a crystalline product was obtained from the mother
liquor that remained after the isolation of some of the phosphorus-free phrenosin
(cerebron), which contained an average of about o.g per cent of phosphorus, but
the preparations of which showed wide fluctuations in composition. The melting-
point was about 190° C. CRAMER’S normal protagon melted at 192.5° C.

14 PosNER and Gies: Journal of biological chemistry, 1905, ii, p. 59. The
figure quoted above pertains to the phosphorus content of a product that had been
‘recrystallized ten times from 85 per cent alcohol at 45° C. Two protagon products
were obtained having contents of phosphorus equal to 1.7 per cent. On recrystal-
lization, products were isolated from each having contents of phosphorus as lowas
0.15 per cent, whereas products from the mother liquors ranged in phosphorus
content between o.8 per cent and 2.44 per cent. These results confirmed the
conclusions of THUDICHUM and of LESEM and Gigs, that protagon is a mechanical
mixture.

16 LOCHHEAD and CRAMER: Bio-chemical, journal, 1907, ii, p. 350. The fig-
ure given above for phosphorus content represents the average of LOCHHEAD and
CRAMEr’s four products of “highest purity.”? Recrystallization fram the second
to the third time resulted in reductions of phosphorus contents equal, for three
products, to 0.17 per cent, 0.2 per cent, and 0.29 per cent, respectively. Fractional
recrystallization of a protagon containing 1.22 per cent of phosphorus, yielded two
products with 1.14 per cent and 1.05 per cent of phosphorus respectively. See
footnote, page 379 of this paper.

388 Matthew Steel and William F. Gees.

This residue from which the supposed protagon had been separated
by continuous extraction with chloroform was found to be insoluble
in alcohol, ether, chloroform, and similar reagents, but was soluble in
dilute alkali, from which it could be precipitated by acid added to
excess. Dried over sulphuric acid zz vacuo and analyzed, the recorded
data for percentage elementary composition were the following:

C H N Le

54-19 Vere UES 7, 1.86

54-66 7-72 ip ie 1.90

Average . . 54.43 7.71 11.41 1.88

The physical properties noted above for this residue, vzz., insolu-
bility in chloroform, alcohol, etc., solubility in dilute alkali and pre-
cipitability from the latter by addition of excess of acid, together with
the proportions of phosphorus and nitrogen revealed by analysis, led
Ulpiani and Lelli to believe that the material in question was a nuclein
or aparanuclein. In following up this idea they found that 1.6256
gm. of the product, kept for forty-eight hours in artificial gastric
juice,’ remained “ unaltered.”* Tested for purin bases by Kossel’s
method after cleavage with 10 per cent sulphuric acid, the results were
negative, and Ulpiani and Lelli concluded that the material was para-
nuclein. It gave the biuret and Millon tests for protein, and yielded
phosphate on cleavage with alkali.

General deductions pertaining to Ulpiani and Lelli’s paper.— Ulpiani
and Lelli concluded their paper with the general deduction that
protagon is not found in the brain in a free state, but is combined there
with a paranuclein as paranucleoprotagon. They dwelt on the impor-
tant point that, whereas their paranucleoprotagon failed to yield pro-
tagon to chloroform during the first washing with chloroform,? it did
yield it to chloroform, they thought, after the treatment with alcohol.
They alluded to the lecitho-proteins in egg yolk, discovered by Hoppe-
Seyler* and studied in some detail by Osborne and Campbell,> and
concluded that the cleavage of paranucleoprotagon into paranuclein
and protagon by alcohol was analogous to the similar severing influ-

1 Temperature and strength were not indicated.

2 Presumably it was undiminished in amount.

3 No observations were recorded in the paper in this connection. Whether this
was merely assumed or not is not clear.

* HopPE-SEYLER: Medicinisch-chemische Untersuchungen, 1865, ii, p. 215.

5 OSBORNE and CAMPBELL: Journal of the American Chemical Society, 1900,
xxii, p. 416.

On the Chemical Nature of Paranucleoprotagon. 389

ence of alcohol on lecitho-nucleovitellins, with the consequent parti-
tion of the latter into lecithin and nucleovitellin. Ulpiani and Lelli
stated that they repeated Hoppe-Seyler’s experiments in this connec-
tion with egg yolk and obtained confirmatory results. They also
tested the effect of preliminary extraction with chloroform in place of °
ether. They thoroughly extracted egg yolk in a Soxhlet apparatus
two days with chloroform, in which lecithin dissolved readily. The
residue was then treated with alcohol, which immediately removed
much lecithin, 7. ¢., split off lecithin, and dissolved it?

Before proceeding to the description of our own work in this con-
nection the reader’s attention should be concentrated upon the fol-
lowing general references to the statements and data in Ulpiani and
Lelli’s paper.

Their description of the method by which paranucleoprotagon was
prepared is lacking in important details of procedure, so that it may
be impossible, by following them as closely as their general direc-
tions permit, to obtain their quantitative yield of paranucleoprotagon
(whatever it was), although gualitatzvely the product would probably
be approximately the same, if paranucleoprotagon was a definite
chemical individual and not a mechanical mixture.

The ordinary physical qualities of paranucleoprotagon were not
mentioned. In repeating the work, as well as the meagre descrip-
tion allows, there is little or nothing in the remarks by Ulpiani and
Lelli that would enable one to be sure that a perfect duplicate of their
product had been obtained, or that would lead to its recognition from
any special physical characters attributed to it.

They regarded protagon as being without doubt a single definite
substance instead of the mixture it is now considered to be.?

In their effort to isolate the protagon-like cleavage product from
paranucleoprotagon they used a new method, — extraction with chlor-
oform and precipitation with acetic ether, after preliminary treatment
with alcohol, — which they failed to show could be used satisfactorily
to prepare protagon.

1 The assumptions the reader is apparently expected by ULPIANI and LELLI to
make here are that all free lecithin was removed before the treatment with alcohol
and that the chloroform did not effect cleavage. These points were not established
by evidence given in their paper.

2 HALLIBURTON: British medical journal, May 4 and 11, 1907 (Oliver-Sharpey
Lectures, April 29 and 30, 1907); also Gries: Journal of biological chemistry,
1907, iii, p. 339.

390 Matthew Steel and Witham F. Gres.

The protagon-like product that was isolated by them held more than
the usual proportion of material containing much phosphorus, as the
high figure for percentage amount of phosphorus in their “ protagon”
indicated. They did not show definitely that the product they called
protagon was sufficiently like the classical protagons to deserve the
name.

The statement by Ulpiani and Lelli that protagon does not occur
free in the brain, but exists there combined with paranuclein, is not
supported by any experimental evidence whatever. If correct, the
statement implies that all the constituents of protagon are thus con-
tained in the brain, in which event the yield of paranucleoprotagon
must be relatively very large. Ulpianiand Lelli failed to mention the
quantity obtained by them. Noll stated that about 20 per cent of
the solid matter of the brain consists of protagon.!

Our own study of paranucleoprotagon was conducted in the way
described below, where, it is hoped, sufficient details of procedure
are given, without redundancy, to enable future students of this sub-
ject to repeat our work easily and exactly.

&
EXPERIMENTAL. PREPARATION OF PARANUCLEOPROTAGON.

Method of extraction. — Fifty sheep brains were finely minced in a hashing-
machine, and the resultant hash treated in a large jar with about ro litres
of chloroform, which was added gradually. Intimate mixture was effected
by thorough stirring frequently. The material was kept covered with a
glass lid so as to prevent undue evaporation of chloroform. For thirty
hours the mixture stood at room temperature, during which period the
chloroform solution showed relatively little tendency to stratify. At the
end of the time stated, the mass was thoroughly stirred once more, trans-
ferred to glass-stoppered bottles and kept in the latter at 45° C. for about
twenty-four hours. This treatment caused perceptible flocculation of the
previously viscid mixture and favored more decided separation of the
chloroform solution. The mixture was filtered under cover over night in
the ordinary way, but into closed vessels, and the residual fluid was
expressed, filtered also, and added to the original filtrate. The liquid
consisting of the combined filtrates is designated farther on as ‘‘the first
chloroform extract.”

The brain residue was returned to the bottles, and again treated with
about 5 litres of chloroform, at 45° C., for sixty hours, Filtration, expres-
sion of the residual fluid, and filtration of the latter were effected as for

* NOLL: Zeitschrift fiir physiologische Chemie, 1899, xxvii, p. 384.

On the Chemical Nature of Paranucleoprotagon. 391

the first chloroform extract. The liquids comprising the combined filtrates
at this point are designated below as “the second chloroform extract.”

Separation of crude paranucleoprotagon from the first chloroform ex-
tract. — The first chloroform extract filtered readily, was clear, and, as it
filtered, it stratified promptly into two layers. The chloroform portion,
which was very slightly yellowish, was isolated with a separatory funnel,
and treated gradually with a little more than an equal volume of acetic
ether, whereupon bulky flakes of a spongy, sticky, yellowish mass were
immediately precipitated, which speedily sought the surface of the solu-
tion and collected there in a slightly tenacious layer. After about twenty-
four hours the precipitate was skimmed from the solution, placed on a
hard-paper filter, washed with chloroform-acetic ether (50 per cent each),
transferred to a stoppered bottle, covered there with ether and thoroughly
washed in fresh portions of the latter occasionally for several days. The
substance did not harden particularly as a result of this treatment, and was
apparently a mixture of grayish ‘and yellowish materials. This crude
product is referred to below as paranucleoprotagon A.

The watery portion of the first chloroform extract, from which the
chloroform layer has been separated, did not yield a precipitate on treat-
ment with acetic ether.

Separation of crude paranucleoprotagon from the second chloroform
extract. — The second chloroform extract was like the first in general
appearance. ‘The watery proportion was less. Separation and treatment
of this chloroform portion were conducted as they were for the first
extract. Only an insignificant amount of substance was precipitated from
the chloroform portion by addition of acetic ether even in large excess.”

The watery layer, as in the first instance, failed to yield a precipitate
when treated with acetic ether.

The precipitate (paranucleoprotagon B) obtained from the chloroform
portion of the second extract, while comparatively slight in amount, was
apparently like that from the first chloroform extract. It was subjected
to the same treatment as that accorded paranucleoprotagon A and was
added to the latter in ether. -

Purification of the crude paranucleoprotagon products. —-- Mixed para-
nucleoprotagons A and B, after the preliminary washing with ether, were
subjected to continuous extraction with ether in a Soxhlet apparatus for
two days. At the end of that time the last portion of ether that bathed
the material yielded a very slight residue on evaporation, but all of it was
readily soluble in chloroform. Continuous extraction with chloroform

1 A larger proportion of the ether caused no further precipitation.

? If the brain mass had been merely washed with chloroform after the removal
of the first extract, the second extract would have contained still less, probably
none.

392 Matthew Steel and Witham F. Gees.

was next effected in the Soxhlet apparatus, for twenty-four hours, at the
end of which time the last portion of distilled chloroform in contact with
the substance left no appreciable residue on evaporation. ‘The material
was then washed free from chloroform with ether on a hard-paper filter
and dried in a desiccator over sulphuric acid.

Qualities of purified paranucleoprotagon.— At the conclusion of this
treatment the product consisted of fairly soft, somewhat tenacious,
grayish and yellowish particles. In spite of the fact that no appreci
able mechanical losses occurred, only about 10 gm. of this para-
nucleoprotagon were separated from the fifty sheep brains taken,
obviously not enough to warrant Ulpiani and Lelli’s guess that all
the protagon of the brain occurs in that organ as paranucleoprotagon
nor to support their assumption that protagon may be completely
separated from brain with chloroform in the form of paranucleo-
protagon. Fifty sheep brains will readily yield much more than 10
grams of protagon.}

That the product possessed protein qualities was shown by its
response to various protein color tests. It was found to contain only
0.75 per cent of phosphorus.? Ulpiani and Lelli’s product contained
1.62 per cent of phosphorus.

In view of the fact that our product was small in amount, we made
no other qualitative tests with it, but used practically all of the rest
for Ulpiani and Lelli’s cleavage operation in alcohol (p. 383), as
follows:

PREPARATION OF PROTAGON AND PARANUCLEIN FROM PARANUCLEOPROTAGON.

Cleavage in alcohol. The remaining quantity of paranucleoprotagon (ap-
proximately 5 gm.*) was treated for about twenty hours with a litre of
85 per cent alcohol at 45° C. Little change was noticed, although the
bulk of the solid mass appeared gradually to diminish somewhat.

Ulpiani and Lelli’s purpose, in carrying out this treatment, was to
effect a division of paranucleoprotagon into paranuclein and protagon,

1 POSNER and GIkEs: Loc. ctf. See also NOLL: Loc. cit.

2 Our washing operations were probably more thorough than theirs. We have
already called attention (p. 382) to the inadequacy of ULPIANI and LELLI’s
description of their procedures, and the consequent impossibility of following them
closely.

8 The material was inadvertently used for this purpose before its exact weight
had been taken.

On the Chemical Nature of Paranucleoprotagon. 393

after the manner of cleavage, by alcohol, of lecitho-nucleovitellin
into lecithin and nucleovitellin.! After the treatment with alcohol,
they cooled the wzfi/tered mixture to 0° C., apparently with the special
purpose of precipitating whatever protagon had been broken off by
the alcohol, and of keeping it associated with the non-protagon part.
They filtered the cold mixture (0° C.), and apparently made no exam-
ination of the filtrate. From the solid matter filtered off at the low
temperature, however, they extracted, in a Soxhlet apparatus, with
chloroform,? what they thought was protagon, although they presented
no evidence that protagon could be prepared from such a chloroform
extract in the manner described by them.?

It seemed to us that a better way of determining the presence of
protagon in, or its absence from, the alcoholic liquid would have been
the employment of the classical method for its isolation. Accordingly
we proceeded from this point as follows:

Special treatment of the warm alcoholic filtrate containing the protagon.
After the paranucleoprotagon had been in the 85 per cent alcohol at 45° C.
for about twenty hours, and the product had perceptibly diminished
in bulk, as indicated on p. 392, we quickly //tered the warm extract and
washed the residue with 85 per cent alcohol at 45° C. The washings,
which were small in volume, were added to the alcoholic filtrate. (The
insoluble part is referred to on p. 395 as “ residue in warm alcohol.’’)

The alcoholic filtrate was then quickly cooled to o° C. and kept at that
temperature about five hours. As the temperature fell, the liquid soon
became opalescent, then milky, and finally white flakes began to form.
They consisted of amorphous matter. Separation of the white precipi-
tate was effected very rapidly by pressure filtration. The resultant “ prota-
gon” was washed with a small quantity of cold alcohol (o° C.). The
washings were added to what is called, on p. 394, the “ protagon filtrate.” #

The protagon. — The precipitate (protagon) showed no crystalline
characters. On drying it formed waxy lumps, which could easily be
powdered. The general resemblance to protagon was quite marked,
but its phosphorus content was only 0.73 per cent. That of
Ulpiani and Lelli’s protagon, obtained by a different process, how-

1 OSBORNE and CAMPBELL: Loc. cit.

2 The extract in the receiver was kept boiling, of course.

8 They ignored RUPPEL’s statement that protagon is decomposed by boiling
chloroform. RupPeEL: Zeitschrift fiir Biologie, 1895, xxxi, p. 86.

4 Under the general circumstances attending preliminary preparation we thought
it well not to wash with ether, but to get the effects of treatment with alcohol alone.

304 Matthew Steel and William ¥. Gites.

ever, was practically twice as much, 2. @, 1.31 per cent. Only
0.095 gm. of protagon-like product was obtained, and it was neces-
sary to use all of it for the phosphorus determination, so that no
other tests could be applied to it. That the material resembled
protagon was very obvious. That it was chemically very unlike
Ulpiani and Lelli’s product, however, was made unmistakable by our
result for phosphorus content.

That more than one substance was present in the original warm
alcoholic filtrate was made evident by the following results of our
examination of the alcoholic filtrate obtained at 0° C., which may be
conveniently called the “ protagon filtrate.”

The protagon filtrate. — The alcoholic filtrate, from the protagon-like precipi-
tate that was obtained at o° C., was evaporated gradually to dryness on a
water bath. As the bulk of the solution diminished and the percentage
of water in it increased, the liquid became yellowish and very viscid. On
drying it was yellowish brown and had an odor similar to that of lecithin
preparations. In these respects it resembled, under similar conditions,
the products of high phosphorus contents obtained by Thudichum, by
Lesem and Gies, and by Posner and Gies, in the filtrates from their
fractionally recrystallized protagons.

The amount of substance obtained was 0.955 gm., practically ten
times the quantity of the protagon-like product. Its phosphorus
content amounted to 2.25 per cent. It appeared to be a mixture
containing lecithin-like material. It was readily soluble in alcohol,
in ether, and in chloroform, and was partly, perhaps wholly, precipi-
tated from its chloroform solution by ethyl acetate or acetone.!_ The
available quantity of the product was insufficient for further study of its
qualities.

Material like this must have been present in the corresponding
filtrate obtained by Ulpiani and Lelli, but they failed to look for it.
Possibly they precipitated some of it by their method of isolating
their “protagon.” That the latter was a mixture seems equally
probable. By reason of its greater proportion, the unprecipitated
product obtained by evaporation of our alcoholic filtrate must be
regarded as of more significance than the corresponding protagon-
like material. At least two substances were present in the original
warm alcoholic filtrate. Probably there were more of them. It is
not very likely that both or all of these were liberated by alcoholic

1 See p. 383 for ULPIANI and LELLI’s method of precipitating their protagon.

On the Chemical Nature of Paranucleoprotagon. 395

cleavage of only one primary substance. We have already referred
to the heterogeneous appearance of our paranucleoprotagon.

Residue in warm alcohol, extracted with chloroform. — The residual mat-
ter, which was filtered from the warm alcohol at the conclusion of the
treatment that had been intended to separate protagon, was subjected in
a Soxhlet apparatus to continuous extraction with chloroform for about
twenty-four hours. The dried residue, corresponding to Ulpiani and
Lelli’s paranuclein, will be referred to below.

The chloroform extract at this point, which, in Ulpiani and Lelli’s work,
doubtless contained material similar to our protagon-like product (and
possibly also some of the substance or substances of high phosphorus
contents present in our “ protagon filtrate’’), was treated by the Ulpiani
and Lelli method for the precipitation of protagon, z. ¢., by the addition
of ethyl acetate, but no precipitate could be obtained in any proportions
of mixed chloroform extract and ethyl acetate. Acetone likewise failed
to produce a precipitate. It was apparent that everything obtained by
Ulpiani and Lelli at this point had already been separated by us by the
method already described. We finally evaporated the unprecipitable
chloroform-ethyl acetate solution to dryness.

The residue amounted to 0.3 gm., and contained only 0.03 per cent
of phosphorus. Our reagents were phosphorus free, and yielded only
exceedingly slight residues on evaporation.

Ulpiani and Lelli paid no more attention to the chloroform-acetic
ether filtrate than they did to the cold alcoholic filtrate. They were
obviously very arbitrary, therefore, in calling their product paranu-
cleo-protagon, even if protagon had certainly been an individual sub-
stance with the well-defined physical and chemical qualities they
attributed to it.

Paranuclein. — The portion of the paranucleoprotagon that resisted
solution, first in the warm alcohol and then during the continuous
treatment with chloroform, corresponded with Ulpiani and Lelli’s
paranuclein fraction of the paranucleoprotagon. It was hard and
brittle, — quite different in these respects from the original paranu-
cleoprotagon. The particles continued to show grayish or yellowish
coloration, as they did from the beginning. A little more than 3 gm. |
was recovered (3.3). The phosphorus content was only 0.38 per —
cent. Ulpianiand Lelli’s product contained 1.88 per cent. Although
their product and ours were the outcome of essentially the same
treatment,! they differed to this surprising extent in phosphorus con-

1 So far as their meagre description permits of near approach to duplication of
procedure.

396 Matthew Sieel and William F. Gees.

tent. This outcome agrees, however, with our other observations in
pointing to heterogeneity as the chief characteristic of both Ulpiani
and Lelli’s paranucleoprotagon and our own.

This ‘‘ paranuclein ” gave sharp responses to the protein color re-
actions. Strongly peptolytic artificial gastric juice appeared to have
no effect on it. After standing for three days in 0.5 per cent potas-
sium hydroxid at room temperature, the appearance of the substance
seemed to be unchanged and its amount undiminished. The filtrate,
on neutralization and cautious acidification, became only faintly tur-
bid. It required considerable heating to bring about appreciable solu-
tion promptly in 0.5 per cent potassium hydroxid.

Summary of quantitative data. — The foregoing quantitative data
pertaining to our paranucleoprotagon and its cleavage products, with
some corresponding data from Ulpiani and Lelli’s paper, are sum-

marized below:
Percentage of phosphorus.

: Weight : See
Fractional product. a Senae: te eee
Ty el. Brotaconi.® 2). cee 0.095 0.73 1.31
2. Solid matter in the alcoholic filtrate
fromthe protagon’)s 9.1). £iae FOOS5 2525 a0. ate
Ils 12 Parantcleinycys acige i -iier a) mes oe 0.38 1.88
2. Solid matter in the chloroform
washings of the paranuclein . 0.314? 0.03 2
I” 41. Substancesecovercd: | 2) i een ara eeu
2. Paranucleoprotagon taken. . 3 0.75 1.62

It will require extended study to determine more definitely and

finally the chemical and physical qualities of paranucleoprotagon.
The above data show clearly, however, that the term paranucleo-
protagon is a misnomer. The material designated by that name is,
nevertheless, an important product, and although it is apparently a
mixture of substances, it seems to contain at least one combination
which, like lecitho-nucleovitellin, may be severed by cleavage with alco-
hol. It is our purpose to investigate further in this laboratory the qual-
ULPIANI and LELLI failed to examine the corresponding material.
A very small proportion was derived from the reagents used. See p. 395.
This does not exactly represent the total, because small portions of some of
the fractional products were used for qualitative tests before the materials were
dried and the weights taken.

4 Approximately 5 gm. of anhydrous material. By an oversight the exact

weight was not taken (see the footnote on p. 392). The figures for phosphorus
contents are none the less significant, however.

1
2
3

On the Chemical Nature of Paranucleoprotagon. 397

ities of paranucleoprotagon, and to study also the substance or sub-
stances in it that resemble the lecithoproteins in the feature
indicated.

SUMMARY OF GENERAL CONCLUSIONS.

The brain product called paranucleoprotagon, as Ulpiani and Lelli
stated, resembles lecithoproteins in undergoing a certain degree of
cleavage when treated with warm alcohol, although resisting such
cleavage by ether and by chloroform. The extent of this chemical
cleavage as distinguished from mechanical partition has not been
determined.

The substances produced from paranucleoprotagon by 85 per cent
alcohol at 45° C are not simply paranuclein and protagon, as stated
by Ulpiani and Lelli, but one or more additional products are sepa-
rated by such treatment. The term paranucleoprotagon, as applied
by Ulpiani and Lelli, is a misnomer therefore.

Ulpiani and Lelli failed to show that paranucleoprotagon is a defi-
nite, individual substance. The difficulty of getting a product exactly
like theirs by careful repetition of their own method (so far as that
is possible from their inadequate description of it), together with the
observation that several products are obtainable from paranucleo-
protagon by their prescribed treatment with alcohol, indicate that
paranucleoprotagon, like protagon itself, is a mixture of substances.
Further, like protagon, paranucleoprotagon yields different products
when acted on by 85 per cent alcohol at 45°C. It is not very prob-
able that all these products are derived by cleavage from a single,
very complex molecule. If more than one of them is a cleavage
product, it is also quite probable that some of them result simply
from mechanical fractionation.

Ulpiani and Lelli have not proved that their paranuclein is a typi-
cal nuclein. They failed to show that it is not a mixture of residual
- substances.

The facts pertaining to paranucleoprotagon lend no support what-
ever to the assumption that protagon is anything definite, physically
or chemically. On the contrary, the observations recorded in this
paper agree with those previously reported from this laboratory, in
showing that protagon has always been a mechanical mixture of
substances. We have indicated reasons for concluding that the data
published recently in dissent by Lochhead and Cramer actually sup-
port this view of the nature of protagon.

398 Matthew Steel and William F. Gies.

The amount of paranucleoprotagon that is obtainable from brain
by the Ulpiani and Lelli process is much less than the quantity of
protagon that may be extracted from an equal amount of brain by
the classical method for the isolation of protagon. Consequently,
Ulpiani and Lelli’s conclusion that a// the protagon that is separable
from brain occurs there combined with paranuclein in the form of
paranucleoprotagon is incorrect. Ulpiani and Lelli have indeed not
even shown definitely that azy protagon (2. e., the constituents of
protagon) exists in the brain in the form of the product they named
paranucleoprotagon.

THE EFFECT OF UNIFORM AFFERENT IMPULSES
UPON THE BLOOD .PRESSURE; AT. DIFFERENT
EV ELS,

' By W. T. PORTER.

[From the Laboratory of Comparative Physiology in the Harvard Medical School.|

E.

ile 1903 it was shown? that the stimulation of the depressor nerve

in rabbits in which the blood pressure had been lowered by
injuries to the abdominal viscera produced a percentage change
in the blood pressure as great as that produced by a stimulus of
equal strength applied to the nerve when the blood pressure was
at its normal level. In 1907 similar results were published? re-
garding the sciatic and the brachial nerves in rabbits, cats, and
dogs, in which uniform stimuli were applied to these nerves when
the blood pressure was normal, and when it had been lowered by
injuries to the brain.

The present investigation is an inquiry into the effect of the
uniform stimulation of the depressor, brachial, and sciatic nerves
at various levels from the normal down to about 9 mm.

Ji

The animals employed were rabbits, cats, and dogs. They were
invariably etherized during the operations performed upon them.
Before etherization a small quantity of morphia was injected be-
neath the skin of some of the dogs. Tracheotomy was done
and cannulas were placed in the crural or carotid artery and, when
necessary, in the crural or jugular vein. Afferent impulses were
obtained by the electrical stimulation of the depressor, sciatic, and

1 W. T. PorTER and W. C. Quinpy: Boston medical and surgical journal,
1903, cxlix, pp. 455, 456; see also This journal, 1903-04, x, pp. Xil, xiii.
2 W. T. PorTER and T. A, STOREY: This journal, 1907, xviii, pp. 181-199.
399

400 W. T. Porter.

branches of the brachial plexus in or near the axilla. For con-
venience, these branches will be called in the text the brachial
nerve. Precautions were taken to prepare the nerves with the least
possible injury and to protect them against drying. The nerves
were always severed and the proximal end lifted into the air when
the, electrodes were applied. A constant current from Daniell or
gravity’ cells supplied the inductorium, and the secondary coil
was kept at a uniform distance from the primary coil.

Before stimulating the brachial or sciatic nerve curare mixed with
from 15 to 25 c.c. of warm normal saline solution was very slowly
injected into the crural or jugular vein. From time to time the
nerves were stimulated to determine whether just enough curare
had been given. By these precautions it was possible to operate
with the least injurious dose. Efforts were made to have the
curarization uniform throughout each experiment. Care was also
used to equalize as far as possible the stimulating action of the
ether. It is important in such experiments to prevent the magne-
sium sulphate, used to prevent clotting, from being drawn through
the arterial cannula into the circulation as the blood pressure falls.
To this end the membrane manometer was employed in almost
all the measurements, as its liquid displacement is so small that
the movement of about I c.mm. of magnesium sulphate solution out
of the manometer chamber, and thus out of the cannula, would
record a fall of almost 100 mm. in blood pressure. In some
experiments the mercury manometer was used as a control, but
the stopcock between the cannula and the manometer was closed
curing every considerable fall in blood pressure, and the connec-
tion re-established only after the pressure in the manometer had
been reduced by opening the proximal limb to the air. This pre-
caution ought on no account to be neglected, though, unless care be
taken, not enough magnesium sulphate may be left in the cannula
to prevent clotting.

In all the graphic records he blood pressure was measured
from the atmospheric pressure line to the lowest point in the blood
pressure curve. This point can of course be easily determined
and is little exposed to inertia errors, whereas it is much more
difficult to determine the mean or the systolic pressure. The use
of the lowest point in the curve makes all readings lower than
if the mean or the systolic pressure had been chosen, and this
is particularly to be noted with regard to the very numerous cases

“ce

—" a

The Effect of Uniform Afferent Lmpulses. 401

in which the usual difference between the systolic and the diastolic
pressure was increased by a lessening of arterial tension not com-
pensated by an equivalent lessening in the force or frequency of
the ventricular stroke.1

In recording the effect of stimulating afferent nerves on the
blood pressure at various levels, advantage was taken of the spon-
taneous changes in blood pressure seen in many experiments, but
for the lower levels it was necessary to resort to agencies leading
to what surgeons would call “shock.” These agencies were sec-
tion of the spinal cord, exposure and mechanical injury of the
abdominal viscera, application of nitric acid or zinc sulphate to
the peritoneum, cauterization of the skin of the limbs, hemorrhage,
injuries to the brain, and section of the splanchnic nerves.

Wali

Graphic records were secured from thirty-eight rabbits, twenty-
seven cats, and four dogs. The total number of measurements
was 765, of which the brachial nerves supplied 196, the sciatic
248, and the depressor 321.

Measurements of this order group themselves about a central
value or type. Thus the central end of the sciatic nerve of the
rabbit was stimulated in seven individuals, while the blood pressure
was at 60 mm. Hg. The blood pressure rose to 81, 88, 92, 92,
92, 110, and I10 mm., respectively. If the arithmetical mean
be taken, 95 mm. Hg will be the typical value which the blood’
pressure may be expected to attain when the afferent fibres of
the sciatic nerve are stimulated in the rabbit with an initial blood
pressure of 60 mm.

The individual observations in the above series are distributed
about the central value through the operation of constant or
accidental errors.

Excessive curarization would be a constant error, tending to
make all the readings in the over-curarized rabbit too high. The
precautions against over-curarization have already been mentioned.
So far as can be judged, there were no constant errors in the
observations.

1 Some of these points have been mentioned in the paper of PORTER and
STOREY (/oc. cit.), but it was necessary to repeat them here, partly for the con-

venience of the reader and partly because the observations collected by PORTER
and STorey are included in the material analyzed in the present investigation.

402 W. 7. Porter.

An accidental error is one that is just as likely to fall on one
side of the central value as on the other. It follows that if the
number of observations be sufficiently large, those that fall at
a certain distance on one side of the central value will be balanced
by those that fall at the same distance on the other side. There
should therefore be no choice between the individual observations
that are thus by mutual compensation to reveal the hidden type.
Every observation made has accordingly been included in the
material here presented.!

It is evident that by the operation of accidental errors the individ-
ual observations will be distributed symmetrically about the central
value, which can then be calculated with an exactness proportionate
to the number of observations. Where the number of observations

at any one unit of measurement is small, the accidental errors will .

correspondingly be few in number and compensation will be very
imperfect. It will be, then, of advantage to treat the observations
in larger groups. For example, instead of analyzing the results of
stimulation at each millimetre of blood pressure between 60 and
70, all the observations between these levels may be treated as one
group. This has been done in the present investigation. Table I
presents the arithmetical mean of all observations recorded when
the blood pressure at the beginning of stimulation was between
II and 20, 21 and 30, 31 and 40 mm. Hg, and thus by increments
of 10 up to 180 mm. Hg. The difference between the mean of
the blood pressures at the beginning of stimulation and the mean
of the highest or, in the case of the depressor nerve, the lowest
values reached in consequence of stimulation was recorded as
the typical change produced by stimulation. For example, the
central end of the sciatic nerve of the rabbit was stimulated seven-
teen times while the blood pressure was between 41 and 50 mm.
The average height to which the blood pressure rose on stimu-
lation was 82 mm. The difference between 82 and 45, the mean
level before stimulation, is 37 mm. Hg, which is therefore the
typical value obtained by stimulating the sciatic nerve of the rabbit
at the mean level of 45 mm. Hg.

The figure just obtained is the absolute change in blood pressure
upon stimulation of an afferent nerve. But for purposes of com-
parison it is clear that the percentile value should be employed.

1 This material was collected during four years. None of the experiments was
made for the present investigation.

a

Ee

a

The Effect of Uniform Afferent Impulses. 403

In one series the stimulation of the sciatic nerve in the rabbit
while the blood pressure was 100 mm. Hg caused a rise of 35 mm.,
and when the blood pressure was 50 mm. a stimulus of equal in-
tensity still caused a rise of 35 mm. The absolute change was the

TABLE I.

THE ABSOLUTE AND PERCENTILE CHANGE IN BLOOD PRESSURE UPON STIMULATION
OF THE CENTRAL END OF THE SCIATIC, BRACHIAL, AND DEPRESSOR NERVES.

Absolute change on Percentile change on

:
Number of observations. stimulation. stimulation.

Blood pres-

SEIRE GL ee ae
stimulation. F S
Sciatic.| Brach-| De

ial. |pressor.

}} ES oe ee es ee ee —-
mm.Hg =a a faite p. c. rise. | p.c. rige.| p.c. fall.
151 to 160 sc ae 22

141 to 150 23 18
131 to 140 26 24
121 to 130 30 23
111 to 120 33 28
101 to 110 37 37
91 to 100 46
81 to 49
71 to 59
61 to 54
51 to 58
41 to
31 to
21 to
1l to

same in both, but in the first instance this change was 35 per cent,
while in the second it was 70 per cent. The effect of the same
stimulus was only half as great in the first as in the second
instance.

The values gained by the above methods show the irregularities
inseparable from the statistically small number of observations,
but they suffice for some instructive conclusions.

404 W. 7. Porter.

IV.

In Table I are shown the absolute and the percentile change in
blood pressure upon stimulation of the central ends of the sciatic,
brachial, and depressor nerves, together with the number of observa-

mm.Hg

{dae E RSE S RES RRS SSE eB EERE!
TT LLL ea SSIs ieee
PALLIAT Ra A ele eo Sei

ime BanEnES

te ol See i Ce
a Cha aL Set oD eter een
| SASSER ERE ERERSS TSEC
PEEEEEEE EEE BLEEP
HEEEEE ae EeeaNKGnE
SEE ar a I] Boge PNNCE
10 Bee C BSRESES CS!
PENA de ested oa ry] en ke N 5
Epleteabial aie ata eae Bae:
%, aan es - Ne
aa Geese Res Bee a D
Sridastenifostesstecetorttc=
HES HESSEEPennes |
LC aan eZ
cE EECEEEEEE HELE EER CBA HEREE
*, z SS =
PERE ERE EEE
fe sar gaeeasuaeeuaueceace Zc
40 sercs TL ELC E CE egal
Pre EEEPEEPRRECEEEEEEE
et pone as
20 a
ea
BOEu anal oA»
ped Ae Pe |e een aa

0 EERE ER EEE EEE EEE EE EEE EE EEE a
155 135 115 95 75 55 35 15

FIGURE 1.—The absolute and percentile change in blood pressure upon stimulation of

. the central ends of the sciatic (unbroken line), brachial (line of dashes), and depressor
nerves (dotted line). The abscissz give blood pressure in millimetres of mercury.
The ordinates for the absolute curves give blood pressure in millimetres of mercury,
ordinates for the percentile curves give per cent.

tions upon which each result is based. The alterations in blood
pressure are shown graphically in Fig. 1.

Table I and Fig. 1 justify the following statements:

(1) The absolute change in blood pressure upon stimulating these
afferent nerves remains almost unchanged until the blood pressure
has fallen to about one-third its normal height.

(2) The relative or percentile change in blood pressure, which is
the true index of the condition of the vasomotor cells, increases as

EE

The Effect of Uniform Afferent Impulses. 405

the blood pressure falls. In the case of the sciatic nerve, this in-
crease persists until the blood pressure is 30 mm., which is about
the level reached when the spinal cord and bulbar vasomotor centre
are destroyed. In the case of the brachial nerve the increase lessens
when the blood pressure has fallen to 65 mm. Hg, but with both
the brachial and the sciatic nerves the percentile rise is greater even
when the blood pressure has fallen below 30 mm. Hg than at the
normal level of about 150 mm. Hg. The percentile change on
stimulation of the depressor nerve increases until the blood pressure
falls to 85 mm. Hg, whence it diminishes slowly to the 35 mm.
level, and then more rapidly to the 15 mm. level. But it should be
particularly noted that the depressor reflex is as great with the
blood pressure at 30 mm. Hg as with the blood pressure at 145 mm.

(3) The data presented are wholly opposed to the hypothesis that
would explain surgical shock by the exhaustion of the vasomotor
centres. The effect of the afferent impulses upon the vasomotor
cells is as great at 35 mm. Hg as when the blood pressure preceding
stimulation is at the normal level. Exhaustion is always preceded
by fatigue, and fatigue is a gradual process. These measurements
give no evidence of a gradual fall from normal power. The reflex
fails only when the blood pressure sinks to a level at which anemia
of the vasomotor cells is certain. Indeed, all we know regarding
these and similar cells strengthens the belief that their endurance
under stimulation is very great. On the other hand, they are ex-
traordinarily sensitive to variations in their blood supply.

(4) The course of the absolute and percentile curves in Fig. I
is not the same for the bulbar and spinal vasomotor reflexes. The
curves suggest a specific difference between the bulbar and the
spinal vasomotor cells. The probability of this difference is in-
creased by the effect of certain drugs.’

(5) It will be noted in Fig. 1 that as the blood pressure falls, the
power of the brachial and sciatic fibres increases. The brachial and
sciatic nerves here display a protective action. The same stimulus
produces relatively a larger increase in the blood pressure as the
danger of bulbar and spinal anzemia increases. The greater the
danger, the greater the reflex.

1 The action of these drugs upon the vasomotor reflexes is now being studied
in this laboratory.

é ya AE er in

atv ys Oe)’

f ' itive re hoes Here ph. OR a
; & WD @ 1)

=
vf
a

SOME OBSERVATIONS ON THE BEHAVIOR OF THE
BULOMATIC RESPIRATORY AND CARDIAC MECH=
ANISMS AFTER COMPLETE AND PARTIAL ISOLA-—
TION FROM EXTRINSIC NERVE IMPULSES.

BY GaN.) st EWART,

[From the Physiological Laboratories of Western Reserve University and the
University of Chicago.)

CONTENTS.

Page-
PGC GUM Med Ivete cons, o! 6! le! votibeil (oN as, ety telcos, ieM.<, 3. vgicdgret «>, (eMhOU
Ordinary results of double vagotomy, especially on the pulse and respiration . . . 408.
Mheniialsrateiof respiration in'resuscitation -. 9. . 4... 2 eee ew we os 4
The rate of the heart when isolated from its extrinsic nerves . . . . .... . 422
Exceptional recoveries after double vagotomy indogs . ...... ... . 425
Partial section of one vagus with total section ofthe other. . . . .... . . 430
PARSE OME ALIONAOUM AE VAGUS te aks eee ss wl Sk el a ee ABs
SUPLEARLt ARAVA mMSTaE SENS as Coos itp ty (is) hia, ahah at lo oe RH le a ewe ole Gaede 4g

ORE than ten years ago I began a series of observations on.

the consequences of simultaneous and successive division of
the vagi. Although the theme is a somewhat hackneyed one, the
contradictory statements of fact, and especially the conflicting inter-
pretations of the observed facts in the literature, seemed to justify
a renewed investigation. Only very brief references to some of
the results have hitherto been published.t. In the meantime experi--
' ments made in conjunction with some of my pupils? on the partial
or complete isolation of the respiratory and other bulbar centres by
anzmia, sometimes combined with section of both vagi and of the
“upper paths,” have given these observations a new interest. There-
fore I desire to put them on record in greater detail and to discuss
their bearing on our more recent work, some further points in
which may advantageously be considered in this connection.

1 STEWART, G. N.: American year book of medicine, 1901, p. 548; Science,
1905, xxi, p. 889.
2 STEWART, GUTHRIE, BURNS and PIKE:, Journal of experimental medicine,
1906, viii, p. 289. STEWART and PIKE: This journal, 1907, xix, p: 328; xx, p. 61.
407

408 | G. N. Stewart.

The extensive literature of bilateral vagotomy has been well
summarized by Pierre Herzen * and Boruttau,* and only such papers
as bear upon the observations now communicated will be referred
to here. The classical clinical picture presented by such animals
as the dog, cat, and rabbit, when they have suffered division of
both vagi in the neck, is too well known to need description. The
slow, deep respiration, in which the duration of inspiration greatly
exceeds that of expiration, the rapid heart beat, the frequent vomit-
ing or regurgitation, the progressive emaciation in animals which
survive for more than a day or two, as dogs often do, and the ter-
minal double pneumonia which so frequently cuts off the animal,
are very familiar phenomena. It is the way in which these symp-
toms are produced, their permanence, their relative importance in
bringing about the fatal result, and the constancy of the latter, par-
ticularly when an interval has been allowed to elapse between the
division of the two vagi, which have chiefly been subjects of dis-
cussion and dispute. Practically everybody is agreed that in ani-
mals like the dog which not infrequently survive for a considerable
time (several weeks, or occasionally months) the pulse rate tends
to return towards the normal. But there is by no means the same
unanimity as regards the respiration, some writers asserting that
it gradually tends to increase in rate, while others maintain that
the slow, deep respiration established immediately on the elimination
of the pulmonary vagus fibres is not essentially modified as time
goes on. It may at once be stated that in a series of more than
fifty dogs, except in those rare instances where the animals survive
for an indefinite period the section of both vagi, I have seen no
such tendency to increase in the respiratory rate as has been de-
scribed, for instance, by Nikolaides,> in dogs when an interval was
allowed to elapse between the division of the first and second vagi,
even when the animals eventually died from the consequences of the
vagotomy. For instance, in the
Experiment of Fanuary 9, 1897, a fox terrier, whose vago-sympathetics were

simultaneously divided in the neck, lived 31 days, being fed at first by
rectal enemata of eggs beaten up in milk. It received food by the mouth
for the first time 8 days after the operation. As shown in Table I, the
pulse rate, which before the section, with the animal under morphia, was 65

8 HERZEN, PIERRE: Thése, Lausanne, 1897.
4 Borutrau: Archiv fiir die gesammte Physiologie, 1895, Ixi, p. 39.
5 NIKOLAIDES: Archiv fiir Physiologie, 1905, p. 465.

|
:
|

Automatic Respiratory and Cardiac Mechanisms. 409

a minute and after the section 154, had diminished by the fourteenth day
to11o. The rate of respiration, which was 15 a minute before section
(under morphia) and 11 immediately after section, never on succeeding
days rose higher than ro, and was usually lower, except on one occasion
when the animal was excited by seeing a mouse, when it reached 12}.

TABLES I:

Days |
since Respira- Respira-
opera- F tion. tion.
tion.

Ratio.

10

18.3

1 In excitement caused by sight of mouse. Some of the respirations are shallow.
Counting only the deep respirations, the respiratory rate is not increased.
2 The dog is very weak.

The ratio of the heart rate to the respiratory rate, which was 4.3 to 1
before section was never less than 11 to 1 afterwards and nearly always
considerably more. In another fox terrier operated on at the same time
and kept under the same conditions, except that it ate for the first time on
the third day, the pulse rate soon after the operation was 160 and the rate
of respiration 1o (ratio, 16: 1). The pulse rate on the ninth day had
sunk to 118, but the respiration was only 8 (ratio, 16 : 2.1). On the
twelfth day the numbers were 120 and 7 (ratio, 17.1: 1). It died on the
thirteenth day.

410 G. N. Stewart.

In a collie on the evening before operation the pulse and respiratory
rates were 78 and 22 respectively (3.5 :1); on the morning before
operation 82 and 28 (2.9 :1); under morphia and ether 96 and 16
(6:1); after section of the first vagus 126 and 20 (6.3:1); after
section of the second vagus 170 and 18 (soon 14) ; on the day after the
operation 158 and 11.6 (13.6: 1) ; on the second day 160 and 16 (10:1),
on the fifth day 143 and 9.3 (15.4: 1), on the seventh day 140 and 11.4
(12.2: 1), on the twelfth day 128 and 16 (8:1), on the thirteenth day
117 and 12 (g.7:1). The animal died on the fourteenth day.

In a spaniel the pulse and respiratory rates on the evening before
operation were 94 and 16 respectively (5.9: 1); on the morning before
operation 84 and 18 (4.6: 1); under morphia and ether 68 and 18
(3.8: 1); after section of the first vagus 140 and 10 (14:1); immedi-
ately after section of the second vagus 160 and 15 (10.6: 1); on the first
day after the operation 188 and 6.7 (28:1) ; on the second day 170 and
9 (19:1); on the fifth day 152 and 6.8 (22.3:1); on the sixth day 144,
and 6.6 (21.8:1) ; on the seventh day 146 and 7.8 (18.7:1); on the
ninth day 154 andg (17.1:1); on the day of its death (the eleventh day)
Rgr ands; (21.55 = Fe

Not only does the rate of respiration show a relatively high
degree of stability in one and the same dog after bilateral vagot-
omy, but there is a far smaller range than normal in the frequency
in different dogs in the absence of the pulmonary regulating im-
pulses. This is well illustrated in Table II, which gives the pulse
rates, the respiratory rates, and the ratios between them before
section, after section of one vagus, and after section of both in
twenty-three animals, all of which died. In only one case was the
rate of respiration more than 12, after section of both vagi, when
the immediate excitation caused by the section had passed off. In
eighteen cases it was g or less; in four cases 4 or 5. In 50 per cent
of all the cases the rate was 6 to 8. The ratio of pulse to respira-
tory rate was invariably much greater than the normal, although,
owing to the tendency of the pulse rate to diminish in animals which
survived more than a few days, the ratio tended towards the normal
as time went on, but without ever reaching it, except in those rare
cases of complete recovery already mentioned.

It is an interesting question how this relative stability in the res-
piratory frequency, which is notoriously susceptible to so many
influences in the normal animal, is maintained after elimination of
the vagi. Some have laid stress on the vicarious control of the

Automatic Respiratory and Cardiac Mechanisms. 411

TABLE II.

SECTION OF BOTH VAGI IN Docs.

Ratio of pulse to
respiration.

:
Pulse rate. Respiratory rate.

After After After After

Before
sec-
tion.

After | After
section | section section | section section | section
sec- sec-
of one jof second oe of one of second | Gion of one | of second
vagus. | vagus. | * | vagus. | vagus. | * | vagus. vagus.

| Before Before |

‘79 | 140 160 10 | so.) 14 10.6
1881 Y 25.01
% | 126 170 | 20 6.8 63 94
1581 13.61
154 43 14.0
16+ | 5.9 41.0
| 365 26.0
59 20.5
32 25.1
1.6 32
3.0 15.8
41.0
6.5 191
6.2 20.6
4.4 18.1
6.8 245
4.4 115
33 8.6
5.9 : 21.4
9.2
2231
17.5
105
19.25
30.0
20
406
178

1 After one day. 2 Vagi crushed thoroughly without being divided.

3 Pup three months old.

4 Second vagus divided fifty-nine days after the first. The counts were made dur-
ing anesthesia with morphia and ACE mixture.

> Seven hours after operation. 6 After five or six minutes.

412 G. N. Stewart.

“higher paths.” We have seen good reason to believe not only that
these paths from parts of the brain lying higher than the bulbar
respiratory centre do exert an influence upon that centre after
section of the vagi (in cats), but that they are normally active since
section of them, the vagi being intact, produces some diminution in
the frequency of the respiratory movements, contrary to Marck-
wald’s observations (on rabbits). Yet the stability in the rate of
the respiratory discharge is mainly due, we believe, not to the
development of an effective extrinsic regulation which takes the
place of the regulation normally exercised by the Hering-Breuer
fibres, but to the relative isolation of the bulbar centre from ex-
trinsic impulses, which permits its “native”? automatic or autoch-
thonous rhythmical discharge to dominate the situation. For
nothing is more striking than the diminished susceptibility of the
respiratory rhythm of the vagotomized dog to such influences, in-
cluding psychical disturbances, as powerfully affect the respiration
of the normal animal, although experimental stimulation of the
central end of the vagus or of the sciatic still produces a great
effect. In the experiment of January 9, 1897, for instance, already
quoted, the intense psychical excitation produced by the sight of a
mouse, which obviously aroused the keenest interest on the part of
the dog, only increased the respirations to 12% a minute, and this
at a time when the animal was in fairly good condition, on the
twenty-fifth day after the operation. It would almost seem as if
the bulbar centre in the absence of vagus control developed its own
rhythm to such a degree that it remained stable in the presence of
extrinsic excitations which in its normal conditiom would have
caused a marked alteration in it.

In conjunction with Dr. F. H. Pike™ I have recently brought for-
ward fresh evidence that the bulbar respiratory centre is not only
capable of such automatic rhythmical discharge at a point in the
resuscitation of the brain and cervical cord after a period of
anzemia when as yet all the afferent paths leading to it, including
the vagi, are still incapable of conduction, but that the initial rate
of discharge under these circumstances is remarkably constant. The
character of the respiration at this time irresistibly reminds one of
the respiration after double vagotomy. Although these resuscita-
tion experiments were made mainly on cats, with a few observations

‘

6 MARCKWALD: Zeitschrift fiir Biologie, 1887, xxiii, p. 149; 1890, xxvi, p. 260.
7 STEWART and PIKE: This journal (oc. czt.).

Oe a,

Automatic Respiratory and Cardiac Mechanisms. 413

on rabbits and very few on dogs, which are not suitable on account
of the free collateral circulation to the brain through the intra-
vertebral vessels, it cannot be doubted that in some points they
throw light on each other. I have the less hesitation in so apply-
ing them that the practically constant initial rate of discharge
(about 4 a minute) of the resuscitated respiratory centre of the cat
agrees extremely well with the rate observed by Loewy § in rabbits,
after section of the vagi and the brain stem above the bulb, and
with the rate observed by Katschkowsky ® and by myself (Table II)
in a certain number of vagotomized dogs. The constancy of this
rhythm points to its dependence upon a fundamental property of
the nervous elements in which it originates. A qualitative similar-
ity between the respiratory centres of animals even of widely
distant groups is universally admitted. There is nothing improb-
able in the existence of a quantitative similarity between not very
widely separated mammals. It is a very natural supposition that
in a certain number of animals, especially when fully anzsthetized,
the higher parts of the brain should be exerting so little influence
that section of the vagi alcne is sufficient to bring about that degree
of isolation of the respiratory centre from afferent impulses which
is associated with the development of its fundamental rate of dis-
charge, although in the majority the influence of the higher paths
must still be removed before the necessary isolation is attained.

The relative constancy of the discharge of the respiratory centre
(in cats) so long as it is isolated from the afferent impulses is
well illustrated in the experiments cited in Table III, where the
isolation was accomplished by subjecting the brain and cervical cord
to a period of anzemia, and in those of which Table IV gives some
Specimens, where, in addition to the anemia, the brain stem was
divided at the posterior boundary of the posterior corpora quad-
rigemina or farther forward. In some cases both vagi were also
cut, and occasionally the spinal cord as well in the lower cervical
region. In determining the initial rate it must be remembered that
when the return of function of the respiratory centre is very rapid,
as after relatively short occlusions, the rate obtained by counting
several respirations or even by measuring the interval between the
first two may be greater than the true initial rate, since the afferent
paths may begin to be opened up even in this short interval.

8 Loewy: Archiv fiir die gesammte Physiologie, 1888, xlii, p. 245.
® KATSCHKOWSKY: Archiv fiir die gesammte Physiologie, 1go1, lxxxiv, p. 6.

414 G. N. Stewart.

TABLE III.

OCCLUSION OF HEAD ARTERIES IN CATS.

Length | Time | Rate of Length | Time | Rate of
Date of | of occlu- after re- | respira- Date of | of occlu-| after re- | respira-
experiment. | sion in lease in | tion per || experiment. | sionin | lease in | tion per

| minutes. minutes.| minute. minutes. | minutes.| minute.

1905. | |
April 19

April 29

| 2d occlusion

Feb. 27

March 12
March 13

(2d occlusion

March 18
March 19

March 20
April 24

March 30

2d occlusion

1 First respiration. 2 After stopping the artificial respiration.

3 Imperfect occlusion. * Both vagi had been cut.

5 Exactly 6 large respirations a minute with apparently an additional small move-
ment between each pair of large ones.

® This is not the initial rate, as the respiration had returned earlier, at a time not
definitely noted.

Automatic Respiratory and Cardiac Mechanisms. 415

The constancy of the initial rate in resuscitation does not depend
on a constant stimulation of the centre connected with such physical
factors of the circulation as the blood pressure, since it is equally
observed in animals during resuscitation after a period of cerebral
anemia when the general arterial pressure is much lower than

TABLE IV.

SECTION OF THE BRAIN ABOVE THE BULB AND OCCLUSION OF HEAD
ARTERIES IN CATs.

Length of | Time after
occlusion in| release in Remarks.
minutes. minutes. respiration.

June 26 13 91 ac One minute later stimulation of
the brachial causes distinct
acceleration of respiration.

July 3 5c Exactly the same rate during
stimulation of the brachial
or the vagus.

After stopping the artificial res-
piration. Spinal cord divided
and aorta clamped below the
diaphragm.

1 First respiration.
2 Section of both vagi produces no effect on the respiration.

normal (Table III), in animals after section of the vagi and the
brain stem above the bulb without cerebral anzemia when the blood
pressure may be normal, and in animals after cerebral anemia
‘when the blood pressure has been kept abnormally high by liga-
tion of the aorta just distal to the origin of the head arteries or at
any point between this and the lower surface of the diaphragm
(Table V). Nor is it necessarily dependent on any particular
chemical condition of the blood, its content of carbon dioxide or
oxygen for example, since in resuscitation after cerebral anaemia,
with artificial respiration of constant rate and depth, the spontane-
ous respiratory movements return in different experiments at very
different intervals after the restoration of the circulation, and the

416 G. N. Stewart.

time of their return and their initial frequency are quite inde-
pendent of the rate of the artificial respiration or of the total ven-
tilation produced by it. It is easy to see that if the respiratory dis-
charge depends on stimulation by the carbon dioxide of the blood
the “ excitability ”’ of the centre must be restored before discharge
can take place. Perhaps the tension of the carbon dioxide in the
centre must be reduced to a certain point before this excitability is
restored. In any case the initial rate must, as has already been
remarked, be connected with some internal property of the respira-
tory centre, and the fact that whether the respiration returns early
or late in resuscitation, after a long or short period of anzemia, its
rate is at first the same, suggests that the property on which this
rhythm depends is a very fundamental one. For it is impossible to
suppose that in every particular the chemical or physical condition
of the respiratory centre is the same at the moment when it first
resumes its discharge, whether the occlusion has been long or short,
single or repeated, the return of respiration tardy or prompt, the
animal deeply or lightly narcotized before the occlusion, the blood
at or near the normal temperature or decidedly below it, the animal
young or old, a rodent or a carnivore.

An interesting observation illustrating the relation of the circula-
tion to the respiratory centre during resuscitation may generally be
made in experiments where the aorta is clamped below the origin
of the head arteries, so that a relatively high pressure is restored
in the head immediately on release of the head arteries. Respira-
tion, having returned, ceases after a time, as is very often the case
also where the aorta has not been clamped, a state of apncea being
induced. When the aorta is now released so as to lower the blood
pressure in the head, the respiratory movements soon-return, as in
the experiment of March 30, 1905 (Table V). It would be of
interest to determine whether heat dyspnoea can be obtained imme-
diately after the return of respiration in resuscitation, and whether -
the rate of discharge of the centre is affected by increase of tem-
perature before the afferent paths are opened up; also whether the
time of appearance of the first respiration is influenced by the tem-
perature of the blood. But hitherto I have not been able to test
these questions. As the afferent paths, especially the pulmonary
vagus paths, begin to open up in resuscitation, there is an increase,
often abrupt, in the mean rate of discharge (Table VI in addition
to Tables III and V). There is an equally significant diminution

——

Automatic Respiratory and Cardiac Mechanisms. 417

in its stability, stimulation of the afferent fibres not only in the vagi

but in the ordinary peripheral nerves affecting it more and more.
At a certain stage in resuscitation the influence of the incoming

TABLE: Ve

OCCLUSION OF THE HEAD ARTERIES IN CATS AFTER LIGATION OR CLAMPING OF
AORTA DISTAL ‘TO ORIGIN OF LEFT SUBCLAVIAN.

Length of

occlusion
in minutes.

Time after
release in
minutes.

Rate of
respira-
tion.

Remarks.

March 29
(11)

March 30 |

April 3

1907
March 7

Second oc-
clusion. .!

March 29
April 2

Second oc-
clusion ..

10

5.101
11.30

5 (exactly)
11.2
12

18

about 3
5
26
3 30
6

Half-grown cat. Internal mammaries tied.
Jaws, tongue, forelimbs, and ribs move
strongly in resp., but not the diaphragm

Stim. of left phrenic causes no cont’n of
the diaphragm, nor does direct stim.

Resp. now mainly head and jaw move'’ts.

Released aorta.

Half-grown cat. Internal mammaries
tied. 20sec. between first and second
respiration, then respiration goes on at
5 aminute. Includes head and chest.

Including some small resp’y mov’ts, one of
which sometimes precedes a deep one.

Natural respiration exactly synchronous
with artificial.

Diaphragm contracts on direct stimulation
but not on stimulation of phrenic.

The diaphragm does not participate in the
movements, though direct or indirect
stimulation of it causes contraction.

Respiration has ceased.

Diaphragm still contracts on stimulation
of phrenic.

Released aorta. Resp. soon started.

Prepared cat.

After 15 minutes asphyxia produced by
clamping the trachea, respiration is 5 a
minute. After 25 minutes asphyxia,
65 a minute.

After § minute asphyxia respiration is 2 a
min., after 25 minutes asphyxia 6a min.

1 First respiration.

418 G. N. Stewart.

impulses from the vagi comes so to dominate the automatic rhythm
that in some experiments the natural respiratory movements have
been seen to acquire precisely the same frequency as the artificial ©

TABLE VI.

EXPERIMENT May 13,1905. Car. TIME oF OCCLUSION, 60} MINUTES.

eee
| Rate of respiration

per minute.

Rate of respiration

i r rel a i
Time after release per minute.

Time after release.

2to4 . 47m. 16

92 . 19m. 16
$2 . 15m. 19
4.33 . 21m. 22
10.7 . 26m. 18
14 : : 30
14 | : 30
14 19
14 ; ‘ 23
13

6.5

1 First respiration. The second respiration followed 30 seconds after the first, the
next five at intervals of 15, 30, 15, 10, and 12 seconds respectively. Then the rate
became nine a minute, half of the respirations being weak and the rest very strong.
Each weak respiration began in 7 seconds, and each strong one in 12 seconds, from
the beginning of the last respiration.

2 Strong and weak alternately as before.

3 Very strong, without any weak respirations interposed.

At 3h. 123 m. artificial respiration was stopped. From 9h. 15 m. till 10 h. 31 m.
ether was administered to control the spasms. At 11 h.12 m. ether was again begun
and continued till 15 h. 22 m. after release, when the animal died of ether poisoning.

respiration and to persist at exactly this rate for some time after

the stoppage of the artificial respiration, an artificial rhythm having

been, so to speak, impressed upon the respiratory centre by the

rhythmical excitation of the pulmonary vagus fibres. For example,

in the

Experiment of March 24, 1905, after an (imperfect) occlusion of 30 minutes
in a young female cat, respiration had returned for several minutes,

Automatic Respiratory and Cardiac Mechanisms. 419

when it was noticed that the natural respiratory movements were ex-
actly synchronous with the artificial respiration, which was at the rate of
42 a minute. Artificial respiration was now stopped, and the animal
went on breathing at precisely the same rate for about 25 seconds. This
was 14 minutes after the release of the head arteries. Seven minutes
later the spontaneous respiration was going on at the rate of 32.4 a min-
ute ; 52 minutes after release it was 75 ; and 80 minutes after release, 171
a minute. In another cat (a large adult male), after a perfect occlusion of
5 minutes, respiration returned 35 seconds after release. On stopping
the artificial respiration, 8} minutes after release, the cat went on breath-
ing for 3} minutes at exactly the same rate (42 a minute). The respira-
tions then became quicker, increasing to 54 a minute 14 minutes after
release, and to 128 a minute 38} minutes later. In the experiment of
March 22, 1905, in a very large adult male cat, after an imperfect occlu-
sion of 10% minutes, spontaneous respirations did not disappear, and soon
after release it was observed that they were proceeding at exactly the
same rate as the artificial respiration (48 a minute). In the first minute
or two after release the spontaneous respiration ceased if the artificial was
stopped. This was also the case for the last minute or two of occlusion,
as if the excitation of the respiratory centre depended upon the afferent
impulses set up in the vagi by the movements of the Jungs. When the
artificial respiration was stopped, 52 minutes after release, the natural
respiration went on with exactly the same rhythm for a little time, and
then, in less than a minute, began to get quicker.

Much evidence has been obtained in the course of this work
that, great as the influence of afferent impulses on the rate and
character of the bulbar respiratory discharges is, it is, as a rule,
only capable of being exerted when the respiratory centre is already
discharging itself automatically. There are, however, conditions in
which a perfectly quiescent respiratory centre can be roused to dis-
charge by the excitation of afferent nerves, as in the following
experiment.

Experiment of November 5, 1890.— Dog weighing 13 kilos received 0.16
gm. morphia hydrochlorate subcutaneously. Five hours later, stimulation
of the central end of the anterior crural nerve caused violent respira-
tory movements. Chloroform was now administered, and the respiration
stopped. Stimulation of the anterior crural always causes respiration to
begin and to go on again violently for a little time. The animal struggles
also, but respiration soon ceases. The pulse is much accelerated. The
corneal reflex is well marked, and there are occasional swallowing move-
ments. Obviously the bulbar centres are not all paralyzed. Yet, if left

420 G. N. Stewart.

alone, respiration soon ceases. Air enters the lungs freely, and there is
no obstruction in the trachea. If the condition is one of apneea, it has
not been induced by artificial respiration, since none has been employed.
The animal was kept alive for more than an hour by occasional stimulation
of the anterior crural, a burst of respirations following each stimulation,
but spontaneous breathing never returned, although no chloroform was
given from the time the respiration first stopped. At last the animal was
killed by bleeding. After blood had ceased to flow, respiratory move-
ments began spontaneously and continued for a minute or two.

This seems to show that the respiratory centre, injured too much
by the chloroform and morphia to act spontaneously and only dis-
charged itself when stirred up by the arrival of afferent impulses,
recovered its natural automatic power when the blood was drawn
off, either because some of the still circulating poison was with-
drawn, or because the anzemia increased temporarily the excitability
of the centre or its power of developing autochthonous stimuli, or
the intensity of the “ blood stimulus.”

S. J. Meltzer and J. Auer?® have found that magnesium salts
produce such an effect on the respiratory centre that the respiration
stops, although if the dose is not too great stimulation of the sciatic
may cause respirations to be discharged. The fact that in such
conditions the respiratory centre can be roused to activity by affer-
ent impulses is, of course, no more an argument against its normal
automaticity than the existence of conditions in which a perfectly
quiescent heart can be caused to beat by stimulating its augmentor
nerves is an argument against the automaticity of the heart’s beat.
I showed long ago?! that the frog’s heart when reduced to com-
plete standstill by cautiously heating it in salt solution, can be
caused to beat by stimulating the cervical sympathetic, even while
the temperature necessary for standstill continues to be maintained.
A long series of beats ensues which much outlasts the stimulation.
This, as well as the distance of the sympathetic from the heart,
eliminates the possibility of escape of current on to the heart being
the cause of the contractions, —a criticism to which Schelske’s ob-
servations '? on the effect of stimulation of the frog’s vagus during

10 MELTZER, S. J., and AUER, J.: This journal, 1905, xiv, p. 366; 1906, xv, p.
387.

1) STEWART, G. N.: Journal of physiology, 1892, xiii, p. 93.

12 SCHELSKE: Uber die Veranderungen der Erregbarkeit durch die Warme,
Heidelberg, 1860.

Automatic Respiratory and Cardiac Mechanisms. 421

heat standstill of the heart are justly exposed, since he states that
the stimulation causes a tetanic condition and that single induction
shocks are followed by single contractions. My observations con-
stitute, I believe, the first satisfactory demonstration that the aug-
mentor nerves of the heart have the power of causing beats when
the heart is entirely at rest, although they seem to be unknown to
H. E. Hering, who long afterwards 1* demonstrated a similar action
of the accelerantes in mammals on hearts reduced to standstill in
various ways. He incidentally observed a similar rousing of auricles
and ventricles to activity in a dog, some years earlier,’* but spe-
cifically states that he could not tell whether the veins might not still
have been pulsating at the time of the stimulation. Hering rightly
considers that Schelske’s vague statements do not constitute a dem-
onstration of this action for the frog’s vagus.

S. A. Matthews 1° has shown that, in the standstill of the dog’s
heart produced by magnesium salts, stimulation of the accelerantes
causes it temporarily to resume beating.

Another analogy, and an interesting one, between the automatic
respiratory and the automatic cardiac mechanism is the sudden
and strong return of respiration in resuscitation after cerebral
anemia and the sudden and strong return of the ventricular beat
under the influence of direct massage in resuscitation after cardiac
failure induced by asphyxia, ether poisoning, and in other ways.
For example, in the experiment of May 16, 1905, after an occlu-
sion of 81 minutes the heart was found to have stopped about 16
minutes after release. The chest was opened, the aorta clamped
below the origin of the head arteries, direct massage of the heart
started, and the animal put into a hot-water box, artificial respira-
tion, of course, being kept up all the time. The auricles began to
beat very soon after massage was begun, as usually happens, but
the ventricles beat very poorly and intermittently. About 20 minutes
after the stoppage of the heart and 15 minutes after the starting
of massage, the ventricles quite suddenly began to beat well. No
further massage was necessary throughout the experiment. More
than five hours later the heart was beating excellently, when the
experiment,had to be stopped. Here it would seem that the nervous

18 HERING, H. E.: Zentralblatt fiir Physiologie, 1905, xix, p. 129; Archiv fiir
die gesammte Physiologie, 1906, cxv, p. 354.

4 Tbid., tgo1, lxxxvi, p. 578; Zentralblatt fiir Physiologie, 1gor, xv, p. 683.

15 MATTHEWS, S. A.: This journal, 1907, xx, p. 323-

422 G. N. Stewart.

mechanism, if it is nervous, which conducts the impulses from the
auricles to the ventricles when its resuscitation reaches a certain
point suddenly begins to conduct, just as the respiratory centre
when its resuscitation reaches a certain point suddenly begins to
discharge. Under the conditions of our experiments the “all or
nothing ”’ law appears to apply to the respiratory centre when iso-
lated from afferent impulses as it normally does to the heart.

We might also consider the fact that the respiratory centre,
isolated as described, discharges itself not continuously but with
a constant rhythm, —a token that, as regards such excitation as
is exerted on it by the blood, it possesses a refractory period.

The relative constancy of the heart rate in the first hours or
days after, double vagotomy in different individuals of the same
species and even in the different mammalian species studied, what-
ever the rate may have been before section, may also be considered
analogous to the relative but more marked constancy of the respi-
ratory frequency after isolation of the respiratory centre either
from all afferent impulses or from that most influential contingent
which reaches it through the vagi. Of course, greater variations
are to be expected in the rate of the heart, which is still in con-
nection with the greater part of its accelerator supply and probably
even with a certain portion of its inhibitory supply after section
of both vagi, than in the rate of respiration after complete elimi-
nation of the afferent paths to the bulbar centre. But still the
range is singularly narrow after the immediate effects of the section
and of the anzesthetic have passed off. What the respective shares
of the accelerator mechanism, whose tone has been investigated
| by Tschirjew,!® Friedenthal,1’ Hunt,’® and others, and of the
local regulating mechanism in the heart which many writers as-
sume, may be in bringing about the gradual return of the heart
rate towards the normal in dogs which survive for some time,
cannot be settled at present. There is some evidence that inhibitory
fibres may reach the heart by other paths than the vagi, perhaps
issuing from the upper thoracic cord with the accelerator fibres,’®
and these, although perhaps normally of comparatively slight im-
portance, may come to play a more important role when the main
cardio-inhibitory path has been severed.

16 TscHIRJEW: Archiv fiir Physiologie, 1877, p. 116.
FRIEDENTHAL: Archiv fiir Physiologie, 1902, p. 135.
18 HuNT: This journal, 1899, ii, p. 395.

19 Cf. FRIEDENTHAL: Lace. cit.

~
-1

———————— ee

>

=

Automatic Respiratory and Cardiac Mechanisms. 423

The rate of the heart isolated from all extrinsic innervation
and under given conditions of temperature, coronary pressure, and
nutrition, appears to be still more constant in individuals of the
same species and in different mammalian groups than the rate after
elimination of the vagi alone. Taking the point in resuscitation in

TABLE VII.

mm. of Hg.
Time after

release in

minutes.
mm. of Hg.

occlusion in
Length of
occlusion in
pressure in
Pulse rate.

Length of
minutes.
Time after
release in
minutes
pressure in
Pulse rate.
minutes.
Blood

1907.
April 5
(rabbit)

ioe)

S

“ID WG

£ON
ie)
rs)

wm

445(1)| 14.151
19.15!

Testis (a) 7.45
18.45 5 April 20
PNAS (rabbit) | 19.30 (1) |
24.45 : (im perf.)
38.45 6 (2)

— a)
ADNAO
ep ep

Onur

1 No reflex inhibition of the heart can yet be obtained.
2 Reflex inhibition first obtained.

3 Before occlusion.
A number in parentheses in the occlusion column indicates whether the occlusion
was the first or second in experiments in which there was more than one occlusion.

about forty cats (specimens of the data are given in Table VII) at
which both the inhibitory and the accelerator mechanisms have not
yet recovered their tone while the heart muscle is still in a tolerably
normal condition and the pressure and temperature of the blood
are not far from normal, the average pulse rate is between 155
and 160. In the few rabbits which yielded results suitable for this
determination the average was between 160 and 165. In an experi-
ment on a dog by Tschirjew after elimination of the vagi and the
accelerators with the knife I find the average of all the observations
is about. 151. But if one selects observations where the blood

424 G. N. Stewart.

pressure was within the normal range, the average is higher (155
to 160). The average rate for the twenty-two dogs referred to in
Table II after section of the vagi alone is 163, when only observa-
tions made after the primary excitation effect of dividing the vagi
has passed off are included. The rate for the dog’s heart isolated
completely from extrinsic impulses must be less than this on account
of the accelerator tone. The average of all the observations on a
rabbit quoted by Tschirjew I find to be a little over 165. The sug-
gestion is that, under precisely similar conditions, the hearts of the
rabbit, cat, and dog when isolated from the central nervous system
beat very much at the same rate, just as the respiratory centres of
these animals when isolated from all afferent impulses discharge
themselves very much at the same rate. Of course such calculations
are open to a certain amount of error, and it is possible that careful
experiments specially directed to this point, in which the temperature
and blood pressure were exactly controlled might show constant
differences even in individuals of the same species. But it can be
safely predicted that the range would not be found a wide one even
between not too distant groups. And without pressing the analogy
too much, it may be considered, perhaps, an additional argument in
favor of the idea that the mechanism which originates the cardiac
rhythm is of the same nature as the mechanism which originates the
respiratory rhythm, namely, a nervous one. That is to say, if the
relative constancy of the fundamental respiratory rhythm, that of
the isolated respiratory centre, 1s a deep-seated attribute of nervous
structures, the existence of a similar constancy in the fundamental
cardiac rhythm, that of the heart isolated from extrinsic impulses,
is an indication that the seat of that rhythm may be of nervous
nature too.

It has been mentioned that as a rare phenomenon a dog both of
whose vagi are cut in the neck at the same time may recover and live
for an indefinitely long period, comporting himself to all intents
and purposes as a normal dog. I have had one, perhaps two, such
cases among more than fifty dogs subjected to simultaneous bilat-
eral vagotomy. One of these unfortunately escaped, owing to the
carelessness of an attendant, when all danger seemed to be passed
and it was gaining weight rapidly, 24 days after the excision of
about an inch of each vago-sympathetic. The other was kept for
300 days and then sacrificed for autopsy. The ends of the left vago-
sympathetic were found separated by about % inch. In the right

————————————

Automatic Respiratory and Cardiac Mechanisms. 425

vago-sympathetic the seat of the section was indicated by a fine
transverse scar visible to the eye, and seen on cutting longitudinal
sections after hardening in Muller’s fluid embedding in celloidin,
and staining by Berkley’s “ rapid” modification of the Weigert-Pal
method #° to be occupied by scar tissue. Scar tissue free from fat
cells occupied the part of the sheath opposite the transverse lesion.
The section was complete except for a strand of sheath at the outer
side of the nerve, which perhaps was responsible for keeping the two
portions of the severed nerve so well in line. Some regeneration
appeared to have taken place, since stimulation of the nerve with
induction shocks, before death, caused slowing and weakening of
the heart. This, however, does not account for the recovery of the
animal, since all the serious symptoms had disappeared long before
any regeneration could have occurred. For the same reason it can-
not explain the rapid return of the pulse rate, the respiratory rate
and the ratio between them to the normal (Table VIIJ). When the
animal vomited, as it did repeatedly from the eleventh day after the
operation till the end of the third month, it always made the sound
associated with that act in the normal dog. The other vagotomized
dogs made no sound while vomiting, indicating that apparently in
this exceptional dog a portion of the innervation of the glottis had
escaped. Slapping the dog’s sides elicited the sound of vomiting
although he did not actually vomit, and he not infrequently made
this sound spontaneously without vomiting. When he did vomit, it
was usually only frothy liquid and not more or less digested solid
food, as in the case of ordinary dogs. These cases of complete
recovery were mentioned in an article written in November, 1900.7?
At the Madrid International Medical Congress of Medicine, in April,
1903, Ocafia showed a dog which, ten weeks after simultaneous
section of the vagi in the neck, was in perfect health and remained
so for more than six months. After death, he states, the complete
division of both vago-sympathetics was verified.

Several observers have supposed that when a considerable interval
is allowed to elapse between the section of the two vagi the fatal
result is postponed. Nikolaides ?? states that two dogs out of three
operated recovered completely when the second vagus was cut 47
days and 57 days respectively after the first. He seems to assert

2 LEE, BOLLES: Vade-Mecum, 5th edition, p. 403.
21 STEWART, G. N.: American year book of medicine, /oc. cit.
22 NIKOLAIDES: Zentralblatt fiir Physiologie, xiv, p. 197; xv, p. 482.

426

G. NV. Stewart.

that as a general rule large strong adult dogs will survive the opera-
tion and recover if it be performed in two stages with such an

interval between.

February 5, 1897.
(ratio 3.1); morning before operation, pulse 90, respiration 38 (ratio 2.3)
phia and ether, pulse 63, respiration 23 (ratio 2.7); after exposure of the vagi, pulse 85, _

Skye terrier.

TABEE Vill:

Whether the comparatively genial climate of

Pulse in the evening before operation 120, respiration 38

; after mor-

respiration 24 (ratio 3.5); after division of first vagus, pulse 86, respiration 19 (ratio
4.5); after division of second vagus, pulse 86, respiration 19.

Days
since op-| Pulse.
eration.
( 153
1 157
) 165
2 130
140
3 126
128
4 135
§ 160
5 1a 144
119
6 130
146 +
ae
(esa
1
10 115
150
am 140
12, 140
13 128
14 100
17 114
18 85
19 112
21 110

Respira-
tion.

——

tor wo
SNS FOLD

SOS ONWA

is)

300

Ratio.

126
10.9
16.2
540

5.8

4.5

33
ifs)

6.0
5.2
SS.
2.5
4.7
4.1
4.6
3.8
2.3
2.8
3.0
3.6

Days

since op-
eration.

300

Pulse.

1307

Respira-
tion.

|

Ratio.

* A large abscess due to infection by the syringe used to inject morphia was
opened in the morning. The counts indicated by the asterisk were taken in the
Before the abscess was opened the pulse, and especially the respiration,

afternoon.

were accelerated.

‘celeration of the respiration was ever seen.
¢ Animal excited so that the respiration could not be counted.

In ordinary dogs after double vagotomy no such extraordinary ac-

Automatic Respiratory and Cardiac Mechanisms. 427

Athens may have had something to do with this result I forbear to
speculate. Undoubtedly in my experience, as in that of others, a
sudden drop in the external temperature or other unfavorable cli-
matic change does sometimes appear to determine the onset of the
fatal terminal pneumonia, for example in the experiment of Janu-
ary 9, 1897 (Table I). But I have seen no reason to believe that the

_ result is in any essential respect modified by the lapse of even a long

time between the division of the two nerves, provided, of course,
it is not long enough for effective regeneration to have occurred. I
agree with Friedenthal 7° that such a substitution of the influence
of one vagus by the other paths as will render the elimination of the
second vagus compatible with long life is not realized, at least under
our conditions. For instance, in Experiments 14 and 23 (Table I),
59 days elapsed between the first and the second operation, yet the
animals died on the second and fourth days respectively after divi-
sion of the second vagus. In a bitch whose right vago-sympathetic
was cut on January 23, 1901, and the left vago-sympathetic on June
Sth of the same year (interval of 136 days), the animal was found
dead and in rigor on June roth. The pulse rate before the anzsthetic
was given for the second operation was 88, and the rate of respira-
tion 20 (ratio, 4.4:1). Under the anesthetic before section of the
left vago-sympathetic, the rates were 60 and Io respectively; after
section of the nerve 180 and 7 (ratio, 25.7: 1).

I do not propose to discuss here the somewhat discordant views
which have been taken by different observers of the cause or causes
of death after double vagotomy. Two main groups may be distin-
guished, — those who lay stress on the pulmonary changes (so-called
“section pneumonia”), so frequently seen at autopsy, and those
who emphasize the digestive troubles. There can be no doubt that
the immediate cause of death is often the pulmonary lesion, and
Pawlow 24 antl others have shown that when infection of the lungs
by the entrance of swallowed or vomited matter is avoided, as by
leaving one recurrent laryngeal intact to innervate the muscles that
provide for closure of the glottis, or by first making a double cesoph-
ageal and gastric fistula, dogs may survive indefinitely. It may be
quite clearly demonstrated, however, in animals which live for some
weeks after complete section of the vagi in the neck, that notwith-

28 FRIEDENTHAL: Archiv fiir Physiologie, 1902, p. 137-
2 Forthe literature see KATSCHKOWSKY : Archiv fiir die gesammte Physiologie,
loc. cit.

428 G. N. Stewart.

standing the paralysis of the glottis associated with aphonia, no pul-
monary symptoms may be present till a day or two before death.
The picture presented in these cases is that of an animal suffering
above all from alimentary disturbances, difficult deglutition, fre-
quent and persistent vomiting, and progressive emaciation in spite
of an appetite which is sometimes ravenous. The animal appears to
be slowly starving to death because of the difficulty of swallowing,
retaining, and digesting its food. The respiration is, to be sure,
very different from the normal in frequency, depth, and type, but
there is nothing to suggest that the lungs are the seat of any patho-
logical process. Suddenly the picture changes. Pulmonary symp-
toms obtrude themselves. The physical signs of consolidation may
be detected, and in a short time the animal is inevitably dead. Some-
times, as has been mentioned, the determining cause of the pul-
monary lesion seems to be some external circumstance, as a sudden
fall of temperature. The idea is exceedingly apt to present itself
to the observer that the pneumonia is an accident, an acute intercur-
rent affection breaking the course of a chronic malnutrition, which
in any case must have ended in death, unless the disturbances in
the functions of the alimentary canal could have been met, as in
Pawlow’s experiments, for instance, by direct gastric feeding
through a fistula. Of course the vagotomized animal is predisposed
to this accident, but there is no definite time after section of the
nerves at which it must take place.

The cause of the regurgitation of partially swallowed food which
is present in addition to true vomiting has generally been supposed
to be the accumulation of the food in the lower part of the paralyzed
cesophagus. Cannon?° states that in the cat elimination of the
vagus fibres for the cesophagus only causes a temporary paralysis
of the peristaltic movements of the lower part of the tube which is
provided with unstriped muscular tissue, although the upper part
remains permanently paralyzed. There is no doubt that in the dog
swallowing remains difficult, although food can be forced into the
stomach after the first few days. The muscles of the mouth and
pharynx concerned in deglutition can be observed to contract more
persistently than normal. as if struggling with the morsel. There
is a good deal of difference in different vagotomized dogs in the
ease with which food is swallowed. In the exceptional “ immune ”
cases the power of satisfactory deglutition returns early.

25 CANNON: This journal, 1907, xix, p. 436.

Automatic Respiratory and Cardiac Mechanisms. 429

As to the explanation of the existence of such cases, little that is
satisfactory can be said. Their rarity is great, and the absence of
any sign by which they can be distinguished beforehand are obstacles
to their study which one has been unable to overcome. The possi-
bility, a slender one, that exceptionally early regeneration might
take place where the severed ends of the nerves or of one of them
were specially well situated with regard to each other, led to a series
of experiments on twelve dogs, in which one vagus or vago-sympa-
thetic was divided wholly or partially with as little disturbance of
the relations as possible and with the sharpest possible edge. In
some of the experiments where the nerve was totally divided, the
head was kept fixed during and for some time after the operation by
mechanical arrangements, so as to prevent retraction of the frag-
ments. The following extracts from the protocols will illustrate
the results:

February 25, 1901. — Total section of left vago-sympathetic with small, narrow
knife, without disturbing relations.

March 2. — Killed dog. The ends of the nerve were separated by not more
than 6 mm., and lay well in line with each other and in contact with caro-
tid. Stimulation of upper end before death caused no effect on the pupil,
and stimulation of the lower end no effect on the heart.

February 25, 1901. — Another dog. Divided left vago-sympathetic completely
without separating it from the carotid. Neck bent while cutting the
nerve, and for some time after.

March 5. — Occasionally makes sound of vomiting, although never actually
vomits.

March 6. — Killed. The ends of the nerve are not separated at all. They
lie quite in line and are united by grayish material. A casual examination
might not show that the nerve had been divided, although when it is fully
exposed the scar is seen. Results of stimulation the same as in the other
dog operated on on the same day.

March 8, 1901.— Divided left vago-sympathetic with a razor, inserting an
aneurism needle between the nerve and the carotid and cutting on the
needle. A thin strand of nerve sheath was left uncut.

March 16.— The nerve was found united, with only a narrow scar and slight
thickening central to the scar. The cardio-inhibitory and pupillo-dilator
fibres were quite inexcitable.

In a second experiment on March 8, 1901, the left vago-sympathetic
was completely divided in the same way. On March 16 the ends were
seen to lie close together, and united by a narrow band of whitish mate-
rial. The nerve looked as if only a nick had been made in it, although

430 G. N. Stewart

there is no doubt that it was completely divided. Stimulation had the
same effect as in the first dog operated on on March 8.

In a third dog operated on on March 8, 1go1, in the same way a gap
of about 12 mm. was found on March rz between the ends. Stimulation
of the lower end causes stoppage of the heart, stimulation of the upper
end marked dilation of the pupil.

In another dog 6 days after section of the vago-sympathetic stimulation
of the lower end caused acceleration of the heart, the inhibitory fibres
having degenerated sooner than the accelerators which run in the vagus.”*

These results confirm the foregone conclusion that preternaturally
rapid regeneration of the divided vagus is not the explanation of
the exceptional cases of survival. Another suggestion is that a suffi-
cient number of fibres necessary for life may in these dogs, owing
to some anatomical vagary, escape section when the vagi are divided.
It is known that some of the fibres usually carried in the vagus may
run an aberrant course, efferent cardio-inhibitory fibres, for exam-
ple, being occasionally present in the rabbit’s depressor. May it
not be that a part of the motor fibres of the larynx and cesophagus
usually running in the recurrent laryngeals may leave the main
trunk by the superior laryngeals? This would explain the absence
of paralysis of the glottis, and the greater efficiency of the deglu-
tition mechanism after vagotomy, which would be ‘sufficient to
account for the more favorable position of the animal, particularly
if associated with such an abnormal course of, say, a portion of the
Hering-Breuer and the cardio-inhibitory fibres as permitted a suffi-
cient proportion of these to be spared.

In connection with this question and also on account of the inter-
est of the problem in other relations, a series of experiments was
made in which one vagus was completely divided, while varying
tractions of the total cross section of the other were severed, in order
to determine how small a proportion must remain uncut to prevent
death. J. Steiner?’ stated that in the rabbit the efferent fibres go-
ing to muscle (cardio-inhibitory, cesophageal, and laryngeal fibres)
eccupy a position on the median side of the vagus, while the afferent
(pulmonary) fibres which affect the respiration are contained in a
lateral bundle, which it is possible to separate with a thin instrument

26 Cf. ScuIFF: Archiv fiir die gesammte Physiologie, 1878, xviii, p. 172;
ARLOING: Archives de physiologie normale et pathologigue, 1896, viii, 5e serie,
P- 75-

27 STEINER, J.: Archiv fiir Physiologie, 1878, p. 218.

Automatic Respiratory and Cardiac Mechanisms. 431

from the median bundle. At the level of the medulla oblongata the
cardio-inhibitory fibres run in the middle and lowermost rootlets
of the 9th, roth, and 11th cranial nerves, and the afferent respira-
tory fibres in the uppermost rootlets. Cadman ?* has shown that in
the dog the afferent respiratory fibres and the afferent cardio-
inhibitory fibres enter the bulb in the upper rootlet of the gth, roth,
and 11th cranial nerves. The-efferent cardio-inhibitory fibres leave
the bulb in the lowest rootlet. In the cat, although the rootlets are
more numerous, the various groups of fibres occupy similar posi-
tions. Friedenthal ?° has taken advantage of these facts to eliminate
all the extrinsic nerves of the heart without causing death.

So far as I am aware, the precise distribution in the vagus trunk
of the dog of the various functional groups of fibres is unknown.
Judging from the similarity of the paths in the rootlets of the rabbit
and dog, it is not very hazardous to assume that the efferent cardio-
inhibitory fibres are probably mainly at the median side, and the
afferent respiratory fibres mainly at the external side of the nerve.
A larger or smaller part of the cross section was therefore divided,
beginning either at the-median or at the lateral side of the nerve,
the amount actually cut being controlled by micrometric measure-
ments on the hardened material obtained post .mortem. Several
dogs were operated on with aseptic technique. In one of these (a
young male fox terrier) the vagi were slightly separated with a
needle from the sympathetics in the neck. Rather more than half
the cross section of the right vagus proper, beginning at the median
side, was divided. The partial section of the right vagus caused no
change in the pulse rate (138); the respiratory frequency slightly
diminished (from 41 to 36). Then the left vagus was completely
divided. The effect on the pulse rate (increase to 177) on the res-
piration (diminution to twelve a minute with deep, prolonged inspt-
ration), and on the voice indicated that all the cardio-inhibitory
fibres, many of the Hering-Breuer fibres, and a part of the motor
supply of the glottis muscles in the right vagus had been cut. This
would suggest that the cardio-inhibitory fibres lie at the median side
of the nerve and the Hering-Breuer fibres at the lateral side, as
Steiner says is the case in the rabbit. The wound healed by first
intention. The animal recovered entirely and was killed for autopsy
after 55 days. Among the symptoms in regard to which this animal

28 CADMAN: Journal of physiology, 1900, xxvi, p. 42.
29 FRIEDENTHAL: Archiv fiir Physiologie, 1902, p. 135.

432 G. N. Stewart.

in the first days after the operation conformed to the classical clinical
picture of double vagotomy were the greatly accelerated pulse and
the diminished respiratory frequency. A striking difference was
the rapid restoration of a normal rate of respiration and a normal
pulse rate. (Respiration on the following day 20, pulse rate 126;
on the second day 20 and 139; on the fifth day 22% and 104 respec-
tively.) As regards the digestive tract, the animal swallowed with-
out difficulty from the first, indicating the escape of a portion at any
rate of the esophageal motor fibres, which therefore do not, it is to
be supposed, all lie at the median side of the dog’s vagus, as Steiner
states is the case in the rabbit. There is some possibility, perhaps,
that a larger portion of gullet is provided with smooth muscle
which sooner recovers its power of peristalsis in some dogs than in
others. Nevertheless there was some gastro-cesophageal disturbance
indicated by numerous abortive attempts at vomiting, accompanied
by the characteristic sound, repeated at generally increasing inter-
vals during the first four weeks. Actual vomiting or regurgitation,
which forms so constant a symptom after total section of both vagi,
was never seen. Partial aphonia was present for a while, the at-
tempted whine or bark being shriller than usual. Frequent hawking
or coughing was observed from the end of the second till the end of
the fourth week, possibly excited by some irritation arising in the
partially severed nerve. On the second, third, and fourth days there
were several peculiar convulsive attacks, never seen, in my experi-
ence, in dogs after total vagotomy. The animal would fall down on
the floor on its back, the limbs extended in spasm, with moderate
opisthotonus and twitching of the eyelids. Consciousness was not
lost. The anal sphincter was relaxed. The heart was decidedly
slowed (to 92 a minute immediately after one attack and 96 a min-
ute after another). The animal soon got up, but was weak on its
legs for some time. The normal acceleration of the heart towards
the end of inspiration was already present on the day after the
operation. After total section it is not seen. Apparently a small
number of vagal cardio-inhibitory fibres had been spared. The
number could not have been large, since the heart after the operation
was accelerated to the rate ordinarily seen after total vagotomy. It
is interesting that the removal or diminution at the time of the inspi-
ratory acceleration of the cardio-inhibitory tone maintained by this
relatively small number of fibres should produce a very distinct
effect. However, it has never been demonstrated that the acceler-
ator fibres in the vagus may not be concerned in this acceleration.

EE ————— ere” ~”

Automatic Respiratory and Cardiac Mechanisms. 433

Ancther young dog died in less than three days with the typical
symptoms seen after complete section of both vagi, although some-
what less than one third of the total diameter of one vagus proper
remained uncut on the external side (total diameter, 71; uncut por-
tion, 23). On the day after the operation the pulse was 164 in
the morning, 1&0 in the evening, and the respiration 8. Vomiting
was frequent, and deglutition difficult. Aphonia was present.

In other experiments the vagi were divided piecemeal and the
effects of successive partial seetions on the pulse and respiration
observed. Thus in the experiment of January 25, 1905, in a dog
under morphia and ether, the following observations were made:

Pulse Respiration
Both vago-sympathetics exposed . . . . . . 102 25.3
94 22.6
Cut § of left v.s. fromouterside . ... . . 96 21.4
or of leit v.'s. from outerside . . . . . + Tio 23.3
: 114 21.3
Mrmmeemomicitws Ss: 55 3 er ee ke EEA Oy Be
116 18.6
Cut more than } of right v. s. from median side . 116 19
116 19
Cut 2 of right v. s. from median side . . . . . 184” 14.6
oa 13.6
Memincouenmiphtv:S ° - : - 2. . 2 s + 6184, i
200 12.8

Another set of experiments was made in which the vagus was
stimulated above or below the level at which successive partial
sections had been made, in order to study the quantitative differ-
ences in the effects produced on the pulse rate, the respiration and
blood pressure by stimulation of given strength when the propor-
tion of the total cross section of the nerve stimulated was varied.

Experiment of February 13, 1905, isan example. A large dog was an-
zesthetized with ACE mixture. The left vagus was isolated from the
sympathetic. The sympathetic was cut and the lower end stimulated.
No effect on the heart was caused by the stimulation. The peripheral
end of the isolated vegus was now stimulated (by shielded electrodes fixed

80 These pulse counts are somewhat too low. as a weak beat was missed occa-
sionally in the femoral artery, where the pulse was counted. In the last observa-
tion it was counted over the thorax.

434 G. N. Stewart.

once for all in position on the nerve) with a strength of current sufficient
to stop the heart completely. Made a partial section of the vagus below
the electrodes extending a little more than one-third across, beginning at
the median side. Stimulated vagus with same current as before; the
heart is now only slowed, not stopped. Increased the strength of the
stimulus, the heart is stopped. Deepened the cut (Post mortem it was
shown by hardening the nerve and measuring the total section *! and the
uncut portion of the section that the ratio between them was 11.8: 4.8,
so that about three-fifths of the fibres were severed). Stimulation with
the weaker current previously used caused no effect whatever on the
heart. The stronger stimulation caused standstill. Now made a second
cut a little lower down on the nerve than the first, but beginning at the
outer side, and dividing about one-third of the total section of the nerve.
No effect on the heart is caused by stimulation with the weaker stimulus
used before. Now made a third cut one-third of an inch below the
second, beginning at the median side of the nerve and involving between
one-fourth and one-third of the total section. Stimulation at the same
position as before with the stronger stimulus caused stoppage of the
heart. Deepened this cut so as to include rather more than one-half the
section of the nerve. Stimulation with the stronger current causes stop-
page of the heart. Deepened the cut somewhat. The stronger stimulus
has now no effect on the heart. Since the various portions of the nerve
are still in physical contact with each other, none of the previous inhibi-
tory effects are to be attributed to escape of current below the incisions
or to electrotonic currents. A still stronger current causes stoppage of
the heart. Deepened the third cut so as to leave only between one-eighth
and one-ninth of the total cross section undivided (as shown by fost
mortem measurement). Stimulation with the strongest current has no
effect on the heart.

In some experiments one vagus or vago-sympathetic was stimu-
lated without dividing it, and then partial sections were made above
or below the point of stimulation. The other vagus was either
intact or previously divided. The conflicting or conspiring effects
of simultaneous excitation of efferent and afferent fibres in the
nerve were sometimes analyzed in an interesting way by means of
the partial sections. For instance, in the

Experiment of Fanuary 10, 1905, the left vago-sympathetic of a dog
anesthetized by chloretone and ether was stimulated without division, a

51 By making outlines of the sections with the camera lucida, cutting out pat-
terns from thin sheet copper and weighing them, and also by measuring various
diameters with the eyepiece micrometer.

Automatic Respiratory and Cardiac Mechanisms. 435

blood-pressure tracing being taken from the carotid. ‘The relations of
the nerve were not disturbed, and the other vago-sympathetic was left
intact. Both with weak and strong stimulation slowing of the heart and a
slight fall of blood pressure were obtained. ‘The nerve was then partially
divided below the level of the electrodes, the cut beginning at the median
(vagal) side and extending fully half-way across the vago-sympathetic.
Stimulation with the strong stimulus used before caused slowing of the
heart and only a slight fall of pressure The cause of the small effect on
the pressure was revealed on deepening the cut, for stimulation was now
without effect on the heart rate but caused a marked increase of pressure,
the afferent pressor fibres being now permitted to prevail over the efferent
inhibitory fibres. The left vago-sympathetic was then completely divided.
The pulse was now 146, the respiration 22. Partial section of the vago-
sympathetic beginning at the median side reduced these rates to 140 and
20. A further cut left the pulse rate unchanged, but reduced the respir-
atory rate to 18. On increasing the depth of the cut still further, the
pulse rate rose to 150, the maximum seen in the experiment, even after
total section of the right vago-sympathetic somewhat later, and the respir-
atory frequency sank to 14, the minimum rate. Although the increase of
the pulse rate to the maximum might seem to indicate that all the cardio-
inhibitory fibres had been divided, this was not the case, since stimulation
of the nerve above the cut still caused some slowing of the heart, with a
slight fall of blood pressure and stoppage of respiration. When the cut
was extended still more, stimulation no longer produced any effect upon
the heart, but caused a slight fall of blood pressure, after a longer latent
period than the fail previously obtained, and probably due to the
excitation of depressor fibres.

Good evidence was obtained in several of the experiments that
while inhibitory effects can be produced on the heart by excitation
of a relatively small proportion of the total number of cardio-
inhibitory fibres, the threshold is raised when the proportion is
reduced beyond a certain limit, and further that the effect produced
is less profound and more transient. The period of recovery of
the heart when it has begun to beat again is especially shortened
when the number of inhibitory fibres is reduced, even when the
stoppage of the heart is as prompt and its complete quiescence
as prolonged as before. For this reason the fall of blood pressure
caused by cardiac inhibition produced by excitation of a limited
number of inhibitory fibres, even where the inhibition is complete,
is less persistent than where a greater number of fibres have been
excited. Thus, in the experiment of February 7, 1905, the right

436 G. N. Stewart.

vago-sympathetic of a dog anesthetized with morphia and ACE
was exposed and the strength of stimulation determined which
just caused stoppage of the heart. A partial section was now
made below the electrodes, beginning at the median (vagal) side
and extending between one-third and one-half way across the
nerve. The pulse rate was 86 and the respiratory rate 32.8 before
the section, 82 and 27.4 respectively after it. Stimulation of the
same strength stops the heart only for an instant. It requires
a stronger stimulus to cause stoppage as complete and lasting as long
as that caused before section. When the peripheral end of the left
vago-sympathetic was stimulated with such a strength of current
as caused complete inhibition of the heart, no inhibitory effect was
obtained after a section of the nerve below the position of the elec-
trodes, beginning at the lateral (sympathetic) side and dividing
everything except about one-tenth of the total cross section of the
vagus proper at the median side. The existence of accelerator
fibres in the vagus of the dog was clearly demonstrated in many
of the experiments, and indications were observed that an effective
proportion of them may escape section when all or nearly all the
inhibitory fibres have been cut. For example, in the experiment
just referred to, when the peripheral end of the vago-sympathetic
was stimulated the heart was stopped, and when it resumed beating
after stimulation was ended there was distinct acceleration, as
is very commonly observed in dogs. A partial section, beginning
at the median side and extending nearly one-third across the vagus
proper, was now made below the level of the electrodes. Stimula-
tion of the same strength caused the same effects as before. But
when the cut was extended so as to sever about one-half of the
vagus proper, the same stimulation caused only slight slowing of
the heart, although quite as great an after-acceleration as before
the section.

One or two similar experiments on partial stimulation of the
afferent fibres in the sciatic which influence the blood pressure
and the respiration were performed, although not enough to permit
any very definite conclusions to be drawn. In a dog anesthetized
with chloretone and ether the central end of the sciatic was stimu-
lated (after section of both vago-sympathetics) with strong induc-
tion shocks. The respiration was markedly quickened, then slowed,
and there was a slight fall of blood pressure. The nerve was now
partially divided centrally to the electrodes. Stimulation of the

spine dele

Automatic Respiratory and Cardiac Mechanisms. 437

same strength as before caused no change in the blood pressure,
but the same respiratory effects as previous to the section. On
deepening the cut and repeating the stimulation, slowing of the
respiration without previous acceleration was obtained. This
apparent reversal of the original effect may be due to the escape
of a group of inhibitory respiratory fibres, while the fibres which
accelerate the respiration have been cut. It is also conceivable,
however, that the excitation of a relatively small number of afferent
fibres might so affect the respiratory centre as to reduce the rate
of its discharge, while stimulation of a larger number of similar
fibres might increase that rate, just as stimulation of one and the
same group of fibres may have different effects according to the
strength of the stimulus.

SUMMARY.

1. After double vagotomy in dogs the ratio of the pulse rate to

the respiratory rate, which is at first much increased through the
quickening of the heart and slowing of respiration, tends to diminish,
as time goes on, in animals which survive more than two or three
days. This change is due to a fall in the pulse rate, the rate of
respiration in the great majority of cases showing no tendency to
increase.
- 2. The relative constancy of the respiratory rate after double
vagotomy depends largely on the fact that in the absence of the
Hering-Breuer fibres the fundamental rhythm of the respfratory
centre becomes predominant.

3. This fundamental rhythm is revealed in the initial rate of
respiration in resuscitation after anemia of the brain and cervical
cord before the afferent paths to the respiratory centre have been
opened up. The initial rate (about four a minute) has been studied
mainly in cats; but it appears to be remarkably constant not only
in different individuals of the same species, but in the different
mammalian species investigated (dog, cat, rabbit).

4. At a certain stage in resuscitation, after respiratory move-
ments have returned, an artificial rhythm may be impressed upon
the respiratory centre through the pulmonary vagus fibres, by in-
flating and deflating the lungs at a given rate. The breathing
becomes exactly synchronous with the artificial respiration, and
persists at the acquired rate for some time after the artificial
respiration is stopped. |

438 G. N. Stewart.

5. The rate of the heart when isolated from its extrinsic nerves
by cerebral anzemia is relatively constant when the external condi-
tions (for instance, the temperature and pressure of the blood)
are kept constant.

6. The fatal result of double vagotomy in dogs is not obviated
when a considerable interval is allowed to elapse between division
of the two vagi.

7. In very rare cases dogs may recover completely after simul-
taneous bilateral vagotomy above the origin of the recurrent laryn-
geals. This result is not connected with any specially favorable
situation of the ends of one or both of the divided nerves which
ensures prompt regeneration. It may be due to some anatomical
peculiarity.

8. When one vagus was divided completely and somewhat more
than one-half of the other (beginning at the median side, and thus
sparing to a great extent the Hering-Breuer fibres), a dog recov-
ered completely. When somewhat less than one-third of the total
diameter of the second vagus remained uncut (at the external
side), another dog died with all the usual symptoms of complete
vagotomy.

g. When one vagus is stimulated and partial sections of the nerve
increasing successively in depth are made between the electrodes
and the heart, evidence is obtained that up to a certain point exci-
tation with a given strength of stimulus produces the same effect
upon the heart whether all the cardio-inhibitory fibres in the nerve
are stimulated, or only a certain number. Beyond this point the
inhibitory effect is diminished, but can be made as great as before
by increasing the strength of the stimulus. Beyond this, again, no
increase in strength of stimulation suffices to cause the same inhibi-
tory effect. When a relatively small number of cardio-inhibitory
fibres is excited, the heart may stop as promptly as when all are
excited, but it recovers more rapidly.

I have to thank my friend Dr. F. H. Pike for correcting the
proof in my absence.

——

PANT HiN: AS A CAUSE. OF FEVER. AND ITS
NEWT eAIZATION. BY SALICYLATES:

By ARTHUR R. MANDEL.

[From the Physiological Laboratory of the University and Bellevue Hospital Medical
College. ]

1S a former paper the writer! showed that a constant relationship

exists between the height of fever and the quantity of purin
bases eliminated in the urine. The experiments were performed
upon patients suffering from so-called surgical or aseptic fevers.
In all the cases examined the patients were placed upon a milk diet
and comparison was made between the purin elimination before and
after the operation.

In one case of hernia, with primary union, no infection, being a
true aseptic fever, the purin bases rose from 38 to 60 mgm. as the
temperature rose from 98.9° F. to 1o1.0° F. Morgenbesser,? in a
research from the chemical laboratory of this institution, has shown
that the normal elimination of purin bases in man on a purin-free
diet is between 20 and 30 mgm. daily.

In further confirmation of the discovery that fever and high purin
base elimination were coincident, the writer showed that xanthin
administered subcutaneously to a monkey caused a rise in tempera-
ture, and that a strong decoction of coffee caused a fibrile temperature
in himself. He called attention to the fact that von Jacksch * had
found an increase of purin bases in the urine of tuberculous patients.
Also Benjamin ‘4 reports a case of typhoid whose urine contained the
large quantity of 0.1 gm. of purin bases. ;

The writer can confirm Benjamin’s work, for he has found as
high as 64 mgm. of purin bases in the case of a typhoid patient on
a milk diet.

1 MANDEL: This journal, 1904, x, p. 452.
2 MORGENBESSER: New York medical journal, April 14, 1906.
8 VON JACKSCH: Zeitschrift fiir klinische Medizin, 1902, xlvii, p. I.
4 BENJAMIN: Salkowski’s Festschrift, 1904, p. 61.
439

440 Arthur R. Mandel.

Further, an investigation upon the course of purin excretion in
pneumonia, upon a patient on a milk diet, yielded the following
results.

TABLE 1.
CasE A. K., PNEUMONIA.

Date of Average Leucocyte

Purin bases.

disease. | 24 hr. temp. count.

deg.

103.8 23,000
104.6 24,000
103.6 20,000
103.8 19,000
104.4 22,000
103.9 17,000
99.8 15,000
98.8 15,000
99.2 12,000

102°

|
4,0) SIste
FIGURE ].— Case A. K., Purin elimination in pneumonia.

It will be noticed that the purin excretion reaches 67 mgm. and
in general follows the temperature record of the days of the
experiment.

The experiment of injecting a monkey with xanthin, which was
mentioned before, lost some of its significance after the publication
of a description of a wide diurnal variation of temperature in this

ee

Xanthin Cause of Fever, Neutralization by Salicylates. 441

animal. Professor Lusk suggested that another monkey be given
a constant diet, and that on certain days xanthin, and on other days
xanthin and sodium salicylate, be administered.

A monkey was therefore obtained. His weight was 3.7 kgm.
He was kept on a constant diet of one or two eggs, one banana,
and about 200 c.c. of milk daily. When the xanthin or the salicylate
was given, the banana was split and these substances were spread
on the split surface, the two pieces were put together, and the banana
was always greedily eaten without loss.

The following chart is characteristic of the results repeatedly
obtained :

A.M. Se _AM. P.M. _AM. P.M.
Temp ES 1 a sax: Slo es te Ome, § 10 12 2 4 6
103° TY
Etiiitiaccctttt mp
101° achesttetes No
EERE
noae SSSScSSR550058 ass
99°LH EEE FREER EERE EEE

at Nes

ee
ee

1 4
Normal. Xanthin. ca ee and salicylate.

FIGURE 2.— Temperature curves ina monkey. Xanthin, 0.1 gm., was given at 1; sali-
cylic acid, 0.1 gm., at 2; xanthin, 0.4 gm., at 3; and salicylic acid, 0.1 gm., at 4.

The first portion of the chart shows the temperature curve of a
normal day. The following day administration of o.1 gm. of
xanthin caused a rise to a temperature 1.6° F. higher than had been
ebserved on the normal day. On the third day 0.1 gm. of sodium
salicylate was given at 9 A.M. and again the same at noon. At
IO A.M. 0.1 gm. of xanthin was given. It will be noticed that no
rise in temperature followed, such as was the case when xanthin
was given alone.

The salicylate administered by itself had no effect upon the
normal curves.

The results, however, are not based upon a single experiment.
Thus the average of the I Pp. M. readings on twenty-six normal days
showed a temperature of 101.9° F. (38.9° C.); of fifteen readings
on the days when xanthin was given the average was 103.0° F.
(39.4° C.); and of six readings on days when both xanthin and
salicylate were administered the average was 102.1° F. (38.9° C.).

The relationships in the average of all the experiments are exhib-
ited in the curves on the following page.

5 GOLDBRAITH and Simpson: Journal of physiology, 1903, xxx, p. Xx.

442 Arthur R. Mandel.

It is distinctly evident that xanthin administered at IO A.M.
causes a rise in temperature which may be prevented by the simul-
taneous administration of sodium salicylate. It would be an inter-
esting question to determine whether xanthin forms a chemical

A.M. ° P.M. A.M. P.M. A.M. P.M.

oro ——— ——— aS > —_ ——

Tem OF lh: 1 3 5 9 11 1 5 5 9 ll 1 3 5
P- DEERE fe BE
103° LU jealial a a8
YT | BE
= ze
101° ge re
ER SERSCRERRRARAas HA HH

99¢ EGRESS RTORRRURERTeS

Av. noimal temp. Av. xanthin temp. Av. xanthin & sal. acid temp.

FIGURE 3.

combination with salicylic acid similar to diuretin (= caffein +
salicylic acid), thereby rendering it innocuous.

Experiments with antipyrin failed to reduce the temperature rise
caused by administration of xanthin in the monkey.

Neither Professor Lusk nor the writer was able to cause any
rise of temperature in himself after taking 0.4 gm. of xanthin.
Larger doses would probably have been effective, but were not
administered on account of the cost of the substance. It is very
probable that a febrile patient would have been more susceptible
to the ingestion of xanthin.

In reviewing this work it should be recalled that the writer found
a considerable fall in the output of uric acid in aseptic fevers.
This fall in uric acid elimination is probably in consequence of a
vaso-constriction of the blood vessels of the kidney, thereby decreas-
ing the blood supply to that organ, which is a usual accompaniment
of fever.6 The actual concentration of purins in the blood stream
may therefore be far greater than the comparison between the
normal and pathological urinary content betokens.

In surgical fevers the purins must be derived from crushed tissue,
and they must enter the blood stream without previous oxidation to
uric acid. In septic fevers it may well be that toxins lessen the
power of some of the body’s tissues, the muscles for example, to
oxidize xanthin to uric acid, and hence a larger quantity than normal
is thrown into the blood. In both cases these purin bases may be
a considerable factor in that stimulation of cranial vaso-motor
nerve centres, which is part of the syndrome of fever.

6 MENDELSON: VIRCHOW’s Archiv, 1885, c, p. 274.

Xanthin Cause of Lever, Neutralization by Salicylates. 443

It is obvious from this discussion that the use in fever of
purin-free milk, instead of purin-containing meat, has its scientific
justification.
SUMMARY.

1. In fever there is a distinct relationship between the rise in
temperature and thé appearance in the urine of purin bases.

2. The supposition that purin bases are directly concerned in
the production of febrile temperatures, is strengthened by the fact
that the administration of xanthin or caffein will effect a rise of
body temperature.

3. Sodium salicylate may neutralize the temperature rise pro-
duced by xanthin.

CORRECTION.

In Table V, page 90, the word “boiled” should read “ unboiled,”
and ‘“‘unboiled” should read “ boiled,” throughout the table.

THE RELATION OF VAFFERENT IMPULSES
FATIGUE OF THE VASOMOTOR CENTRE.

By W. T. PORTER, H. K. MARKS, Anp J. B.-SWIET, je:
[From the Laboratory of Comparative Physiology in the Harvard Medical School.|

B
INCE the bulbar centre deals for the most part with streams of
afferent impulses, the physiologist is interested in determining
whether at any point fatigue of the nerve cells may prevent these
constant stimuli from calling forth or regulating the discharges on
which many vital functions depend.

To the clinician and especially the surgeon this question is of
supreme importance. The life of the patient hangs on the main-
tenance of balance or equilibrium reactions, such as the pressure of
the blood. These must be held against the violent changes in the
intensity or volume of the afferent stream often witnessed in disease
or injury. Apparently the vasomotor apparatus in every individual
must be forearmed against afferent storms that afflict only the rare
unfortunate, and him as a rule but once. Practice with increasing
loads, the builder up and habitual protector of reflex mechanisms,
is here impossible.

The present research attempts to determine whether this equi-
librium can indeed be seriously affected by the prolonged stimulation
of afferent nerves.

LL

The animals used in this investigation were cats, rabbits, and
dogs. They were anesthetized with ether. The carotid blood pres-
sure was written by a membrane manometer or, in some cases, by
a mercury manometer.

The following stimulations were made: the central ends of the
divided sciatic nerve, the brachial nerves, the posterior spinal roots,

444

Relation of Afferent Impulses to the Vasomotor Centre. 445

the lumbar branches of the spinal nerves; the great splanchnic,
coeliac ganglion, superior mesenteric plexus, gastric branches of the
vagi, various parts of the abdominal sympathetic; the testis.

The stimuli were chiefly induction currents. In some experi-
ments they were provided by the ordinary inductorium; in others
the primary circuit was made and broken by the Ewald interrupter.
The frequency and strength of the induction currents was varied in
the different experiments and at times in the same experiment.
When the Ewald interrupter was used, from four to six Daniell cells
supplied the primary circuit. The duration of stimulation averaged
about three hours. The usual precautions were taken to prevent the
nerves from drying. The electrodes were not allowed to remain
long at the same point upon the nerve.

TE:

The course of the investigation will appear from the following
selected protocols:

Fune 27, 1905. = ie spinal nerves in the lumbar region were stimulated
three hours, fifty-five minutes in a strong female cat, anzesthetized with
ether. Five Daniell cells supplied the primary circuit, which was inter-
rupted by the Ewald device twelve times per ten seconds during the first
hour and subsequently fifteen times per ten seconds. Stimulation began
at 10.50 A.M.; at 2 p.M. the blood pressure had fallen but 20 mm. Hg.

Fuly 10, 1905.— A cannula was placed as usual in the carotid artery of a cat
anesthetized with ether. There was a considerable loss of blood through
a defective seam in the rubber cannula-tube. The posterior root of the
IV lumbar nerve was stimulated three hours. The rate was eleven induc-
tion currents per ten seconds. The currents were recognizable but not
painful when the electrodes were placed on the tongue. In spite of the
prolonged stimulation following the hemorrhage noted above and the
severe operation upon the spinal cord, the blood pressure fell no more
than in a control animal subjected to the same manipulations except
stimulation of the nerves.

The stimulation of the posterior root of a lumbar nerve was
repeated in three other cats, but there was no significant reduction
of the blood pressure.

This closed the first series of experiments, nineteen in all. In
none was there a significant fall of blood pressure following pro-
longed stimulation. As the results were negative, they were not

446 W. T. Porter, H. K. Marks, and fF. B. Swift, Fr.

immediately published, although the number and character of the
experiments hardly admit a doubt as to the conclusion drawn from
them.

Additional evidence follows:

November 5, 1906. —In an etherized dog, in which the carotid blood pressure
was written, repeated efforts were made to lower the blood pressure by
crushing the testes, but wholly without effect.

November 12, 1906.—In an etherized dog the prolonged stimulation of the
testes with strong and with weak tetanizing currents did not lower the
carotid blood pressure, nor did repeated crushing of the testes half an
hour later. The animal was then curarized, and the brachial and sciatic
nerves stimulated at intervals for two hours and forty minutes, but no sig-
nificant fall of blood pressure was observed.

The following year, still other experiments were made. In these,
anesthesia was obtained by dividing the brain anterior to the pons;
the possible influence of ether anesthesia was thus eliminated.

September 6, 1907. — The large hemispheres of an etherized cat were divided
transversely above the pons by a blunt seeker introduced through a small
trephine hole. ‘The ether was then discontinued. At intervals of one or
two minutes, strong currents from the inductorium were passed through
the central end of the sciatic nerve for, thirty seconds. This continued
from 11.30 A.M. to 1.30 P.M. At the beginning the blood pressure was
80 mm. ; on stimulation of the sciatic it rose to 105 mm.; two hours later
the blood pressure was 60 and rose to 113 mm. on stimulation.

On September 9 a similar though shorter stimulation in a rabbit
also failed to produce a significant fall in the blood pressure.

Dy

The numerous stimulations just described uniformly failed to give
a significant fall in blood pressure, nor have we been able to find
in the literature any instance in which the stimulation of a nerve
produced such a fall in the normal animal.?

It is true that cases are of record which seem at first glance to
contradict this statement. The following are examples:

1 It need hardly be said that the temporary fall produced by the stimulation of
depressor fibres is not to be classed with the abnormal sinking of the blood pres-
sure discussed in this paper.

Relation of Afferent Impulses to the Vasomotor Centre. 447

“ After a control had been taken, in which the blood pressure was 146 mm.,
the animal [dog] was subjected to manipulation, operation on the skin,
including extensive removal, irritation of the sciatic nerve and the nerves
of the tracheal plexus, and finally opening of the abdomen and manipula-
tion of the intestines during the period of one hour.””* The blood pres-
sure fell to 20 mm.

In another experiment,’ “ tearing, crushing the brachial plexus, sciatic nerve,
abstracting six ounces of blood, and performing pylorectomy reduced the
dog to profound shock. Respiration was irregular. The blood pressure
fell to 36 mm.’’#

In the large class of experiments which these citations illustrate,
a combination of procedures, including the stimulation of nerves,
is followed by a fall in the general blood pressure. It is impossible
to draw correct conclusions from such experiments because they
confuse hydrostatic, chemical, and nervous phenomena.

Exposure of the intestines inevitably dilates the blood vessels in
the largest vascular area in the body. The general blood pressure
thereupon necessarily falls. Primarily, this is simply an hydrostatic
phenomenon, identical with the fall in arterial pressure produced
in a rubber and glass model of the circulation by lessening the pe-
ripheral resistance. It may indeed be very dangerous—a rabbit
may be bled to death in its own portal system by dividing both
splanchnic nerves —but the cause of death is anzemia of the bulbar
cells; a local anemia. The removal of large portions of the skin
acts also primarily in this hydrostatic way, by dilatation of extensive
vascular areas.

The escape of blood from the arteries and their capillary ter-
minals either outside the body or into the veins produces not merely
an hydrostatic fall in the general blood pressure, but if the disloca-
tion be excessive deprives the bulbar cells of oxygen by removing
large quantities of hemoglobin. It is for this reason that animals
exposed to a very low blood pressure cannot be recovered by the use
of normal saline solutions. The bulbar cells are permanently in-
jured by very brief oxygen starvation. Only when red corpuscles
are present in sufficient numbers can normal saline injections rescue
the asphyxiating cells by driving the haemoglobin carriers to the
bulb.

2 CRILE, G. W.: Blood pressure in surgery, 1903, Experiment 45, p. 62.
Serine, G. W.: Lac. ciz., Experiment 61, p: 67.
Serie, G. Wi: Loc. ciz., pi 67;

448 _W. 7. Porter, H. K. Marks, and F. B. Swift, Tr.

Such experiments as the two just cited afford therefore no evi-
dence that the stimulation of the afferent nerves produced the
observed fall in the blood pressure.

Another class of observations subject to a serious error of inter-
pretation is well illustrated by a case recorded in an interesting
paper by Dr. Harvey Cushing. In operating on a sarcoma, a “ mass
of glands in the neck had been freely exposed by the high incision
and was readily enucleated. Several large branches of the brachial
plexus, however, were spread out over the growth, and a secondary ~
division of this portion consequently was necessitated. When this
was done, the patient’s radial pulse immediately became impalpable.
It continued thready and almost imperceptible during the remainder
of the operation, which was rapidly completed, and for almost
twenty-four hours afterwards.” ®

Such occurrences are often attributed to a reflex lowering of the
blood pressure through the action of afferent impulses on the vaso-
motor centres. Unfortunately it is very difficult to obtain in the
human subject conditions sufficiently uncomplicated to withstand
rigid criticism. There is reason to believe that Dr. Cushing’s patient
suffered primarily from reflex inhibition of the heart. An entirely
parallel instance occurred recently in this laboratory in an experiment
by Mr. Russell Richardson, who is studying with Dr. Porter the
effect of afferent impulses upon the vasomotor centres in different
vertebrates.

LVovember 6, 1907.— A rat was etherized and the carotid pressure written
with a membrane manometer. A small quantity (0.75 c.c.) of very
dilute curare solution was injected slowly into the external jugular vein.
The blood pressure was now 70 mm. ‘The difference between diastolic
and systolic pressure was about 20 mm. On stimulating the brachial
nerves the individual heart beats almost disappeared from the curve,
the blood pressure fell 20 mm., and the writing lever traced an almost
unbroken line. On injecting saline solution, the heart improved, and
the difference between diastolic and systolic pressure rose to about 15 mm.
An effort was now made to stimulate the central end of the already di-
vided sciatic nerve. When the severed nerve was gently raised upon a
thread, the heart again failed and the above phenomena were repeated.
Thirty-six minutes later, a saline injection was given, the heart gradually
recovered, the blood pressure rose to 110 mm., and stimulation of the
brachial and sciatic nerves caused a rise of about 20 mm. Hg.

5 CusuinG, H.: Annals of surgery, 1902, xxxvi, p. 324. The experienced
will not think we are attacking Dr. CrRILE and Dr. CusHiNnG because we find
useful illustrations in their protocols.

EEE eee ee ee ee ee eee ee

Relation of Afferent Impulses to the Vasomotor Centre. 449

Similar occurrences were noted in other animals during curare
poisoning.

Cases like Dr. Cushing’s, in which reflex inhibition of the heart
cannot be excluded, should not be accepted as evidence of vasomotor
fatigue.

We have not been able to find reliable instances in which the
stimulation of afferent nerves produced a significant fall of blood
pressure in the normal animal.

V.

Nor have we been able to find acceptable instances of vasomotor
fatigue in animals not to be classed as normal.

It has sometimes been urged that vasomotor fatigue is indicated
whenever the blood pressure falls upon the stimulation of the sciatic
or the brachial nerves instead of rising, as it normally does. But it
has long been known that certain drugs, such as curare, chloral,
and strychnine, may convert the usual rise into a fall, and that
this abnormal reaction may rapidly give place to the reflex rise
commonly observed. Instances of this have been recorded in this
laboratory in a research in which Dr. Porter and Dr. Clark have suc-
ceeded in demonstrating specific differences between the bulbar and
the spinal vasomotor cells. Yet we have seen no instance in our
own experiments, or in the literature, in which the blood pressure
suffered more than the usual temporary fall in cases uncomplicated
by inhibition of the heart, hydrostatic reduction, or anzemia of the

bulb.

FURTHER OBSERVATIONS ON THE RELATION BE-
TWEEN BLOOD PRESSURE AND RESPIRATORY
MOVEMENTS.

bY €.’C. GUTHRIE AnD F: H. PIKE.
[From the Hull Physiological Laboratory of the University of Chicago. |

BOUT one year ago we! reported some experiments on the
effect of changes in blood pressure upon respiratory move-
ments. Weconcluded that, under the conditions of our experiments,
the usual effect of any considerable increase in blood pressure is
an increase in the respiratory rate with diminution in the amplitude
of the movements during the period of high pressure, while a fall
in pressure causes a slower rate with increased amplitude. The
changes in amplitude are constant, but the changes in the rate are
not always constant. We supposed that, in the production of the
phenomena which we observed, there might be two factors, possibly
acting in opposite directions, — one in the high blood pressure it-
self, tending perhaps to increase the activity of the respiratory
centre, and the other the lower CO, content of the blood, since the
same volume of blood now supplies but about one half of the
tissues of the body, tending to decrease the activity of the centre.
The pressure we believed to be the constant factor, and the CO,
content of the blood to be the variable. In the absence of positive
evidence we suggested that the seat of action might be either periph-
eral or central, but inclined to the view that it was central.

Our conclusions have recently been questioned by Eyster, Aus-
trian, and Kingsley.2, In order to determine whether our results
were due largely to stimulation of afferent nerves in the region of
the thoracic aorta, as these authors allege, or whether the same
results could be obtained under conditions precluding such a stimu-
lation of afferent nerves, we made some additional experiments,
which we present in this paper.

1 GUTHRIE and PIKE: This journal, 1906, xvi, p. 475.
2 EysTER, AUSTRIAN, and KINGSLEY: This journal, 1907, xviii, p. 413.
4st

452 C. C. Guthrie and F. Fi. Prke.

TECHNIQUE OF THE EXPERIMENTS.

The animals were anesthetized with ether, and tracheotomy per-
formed. The respiratory tracings were taken by means of a Marey-
recording tambour connected with the tracheal cannula in the man-
ner described in our previous paper. The right carotid artery was
ligated, closed with a bull-dog clamp some distance below the liga-
ture, and a nick made in one side as for the insertion of a cannula.
A sound similar to that employed by Chauveau and Marey® and
by Fredericq,* fitted at the lower end with a small rubber bulb
which could be expanded or relaxed at will, was now pushed down
the carotid into the descending portion of the thoracic aorta. The
sound was of such a calibre that it could be readily introduced into
a dog’s carotid, and it did not interfere materially with the circula-
tion in the aorta when the bulb was not expanded. The circulatioa
below the end of the sound could be restored after temporary occlu-
sion by relaxing the bulb. All portions of the tubing connecting
the tracheal cannula, tambour, and ether bottle were adjusted so
as to give a good excursion of the writing-point of the tambour,
and left untouched throughout the remainder of the experiment.
The degree of anzsthesia was the same before and after the occlu-
sion of the aorta, and at no time was the corneal reflex present.
A tracing of the normal respiration was first taken, and then the
bulb at the lower end of the sound was expanded so.as to occlude
the thoracic aorta. The position of the bulb was determined post
mortem.

THE EXPERIMENTAL RESULTS.

The results were similar in all respects to those previously re-
ported. In all experiments the depth of the respiration decreased
immediately after expanding the bulb at the lower end of the sound.
The rate suffered either no change or a slight quickening during the
occlusion of the aorta. After reopening the lumen of the aorta,
independently of any change in the rate, there was always a marked
increase in amplitude over that observed before occlusion as well
as over that observed during occlusion. It is worthy of note that
the same variations in the depth of the respiration occurred during

* MAREY: cited by Morat ET Doyon, Traité de physiologie, Paris, 1899,
iL, Pr: oye
4 FREDERICQ: Archives de biologie, 1890, x, p. 131.

Blood Pressure and Respiratory Movements. 453

the occlusion as occurred before it (Fig. 1). The respiratory rate
observed in a typical experiment for twelve-second intervals before
and during occlusion, and after the relaxation of the bulb, are given

So re at

FicureE |]. — Two fifths the original size. Showing the effect upon respiratory move-
ments of occluding the thoracic aorta by means of asound. The depression of the
signal marks the period of occlusion and high blood pressure. Time trace in seconds.
Normal respiration shown in first part of tracing. Vagi intact.

below in Table I. The position of the bulb was slightly below the
level of the diaphragm.

TABLE I
Time of Respiratory Time of Respiratory
observation. rate. observation. rate.
1. Before occlusion . 13 in 12sec. 4. During occlusion . 14 in 12 sec.
2. During “ . 13in1i2sec. 5. After relaxation . 13 in 12 sec.
3. During “ = igo rasec... 6. After ae 4, (gi #2 see.

Essentially the same results were obtained after double vagotomy
(Fig. 2).

DISCUSSION OF THE RESULTS.

Certain errors, in our opinion, crept into the work of Eyster,
Austrian, and Kingsley. It has not been shown that the rate and
depth of respiration are in all respects the same during artificial
respiration with Brauer’s*® apparatus, after opening the thorax, and
during normal respiration with the thorax intact. Again, the
authors do not state the manner of clamping the descending aorta
at the upper part of its course. It should be pointed out that this
is near the place of distribution of the endings of the depressor
fibres, and that compression either by a ligature or by hzemostatic
forceps might well set up a stimulation of afferent nerves which
would complicate or obscure the results due to changes in blood

®§ BRAUER: Mittheilungen aus den Grenzgebieten der Medizin und Chirurgie,
1904, xiii, art. xviii.

454 C.-C. Guthrie and F. Ff. Ptke.

pressure. Indeed, the practice of clamping the aorta in this region
by means of hemostatic forceps causes so much disturbance in the
cardiac rhythm that we abandoned it in our experiments on the
effect of blood pressure on cardiac activity.° Thrusting a pair of
hzemostatic forceps into the thoracic cavity in almost any region
where they will come in contact with the lungs or diaphragm is
almost certain to cause disturbances in the heart rate. We found

INanhhhlhhhhhalhnnhhhannhhshhanbhhanhnnannnhnhnnhhhhhahhhbhahhhnnhhhnhhhghannann nnn ahhh thhhhhhhnhhhhhhhhhhhhhhan

FicurEe 2.— About one-half the original size. Same as Fig. 1, except that both vagi
have been divided.

that a carefully placed ligature drawn up through a glass tube with
smooth ends might be tightened sufficiently to occlude the aorta
well down toward the diaphragm by drawing against the end of
the tube without introducing any cardiac irregularity traceable to
direct stimulation of the nerves. The open thorax and the stimula-
tion of afferent nerve endings must, therefore, be admitted as two
possible sources of error.

Another element to be considered when the aorta is ligated high
in the thorax, as in Eyster, Austrian, and Kingsley’s experiments,
is the great diminution of oxygen-consuming tissue. It is evident
that the consumption of oxygen and the production of CO, would
be much less, and we should expect, as we have pointed out above,
that the respiratory activity would be greatly diminished if this
were the only factor in the case. That this tendency to depression
should be sometimes greater than the opposite tendency towards
stimulation is a reasonable supposition, particularly when the aorta
is ligated high up. A diminution both in rate and amplitude is not,
therefore, cause for great surprise.

It has been objected that our first method of occluding the tho-
racic aorta is a frequent source of error, since traction upon the
ligature causes a stimulation of afferent nerves. That sufficiently

§ GUTHRIE and PIKE: This journal, 1907, xviii, p. 14.

Blood Pressure and Respiratory Movements. 455

strong traction upon a ligature passed about the vertebral column,
but not including the aorta, may cause enough stimulation of affer-
ent nerves to affect the respiration, we do not doubt. But that a
reasonable degree of traction, equal to the lateral pressure of the
blood in the aorta plus a little more than the elastic resistance of
the wall of the vessel, will necessarily do so, we are not so ready to
admit, particularly when the animal is properly anzesthetized, — that
is, insensible to pain. But whatever the effect of such a ligature
upon the afferént nerves, it is most probably avoided by the use of
the aortic sound. The only sources of error due to stimulation of
afferent nerves are (1) excitation of sensory nerves in the wall
of the aorta at the part occluded by the sound, and (2) the stimula-
tion of peripheral endings by the suddenly increased blood pressure
in parts of the vascular system which are still open. As a matter
of fact, we see little reason for distinguishing between the two. If
this should prove to be a source of error in our experiments, it is
obvious that it would apply also to any experiments, such as the
unpublished experiments of Eyster, in which the general systemic
blood pressure is raised by ligation of all or part of the cerebral
arteries, and the results there stated would not, therefore, be due
solely to the circulatory changes, but would be complicated by the
effects of the stimulation of the afferent nerves. It has been shown
also that respiration may persist for a time after the blood pressure
in the circle of Willis, as measured from the peripheral end of one
carotid, has fallen to the base line.

If there is no direct evidence to show that the respiratory centre
is affected by changes of blood pressure per se, there is also no
evidence to show that the increased blood flow per se is the sole
regulating factor in respiratory activity. It is scarcely necessary
to say that a crucial experiment separating definitively the effects
upon the respiratory centre of blood flow from those of tension or
pressure has not yet been performed. Indirect evidence," drawn
from experiments upon the heart ganglion of Limulus, is in favor
of the independent effect of pressure upon the respiratory centre.

We have not made any measurements of the respired air during
the period of high blood pressure, nor did we make any predictions
as to the relative amounts of air required during periods of high
blood pressure and normal pressure. It would be surprising, in-
deed, if an animal whose oxygen-consuming tissue was reduced to

7 CARLSON: This journal, 1907, xviii, p. 149.

456 C. C. Guthrie and F. H. Ptke.

only one half or one third of the total tissues during occlusion of
the aorta should respire as much air at this time as it required under
normal conditions. But in order to show that the increased blood
pressure exerts a depressing action on the respiratory centre, it is
necessary to show that that portion of the tissues supplied with
blood actually receive less oxygen, as measured by the total ventila-
tion of the lungs, during the period of high blood pressure than they
would receive during normal blood pressure. Eyster’s statement
that-a diminution in the rate as well as in the amplitude of respira-
tion occurs during the period of increased blood pressure, or that
the total ventilation of the lungs is less at this time than- when the
pressure is normal, lacks the quantitative exactness necessary to
prove his contention that the increased blood pressure depresses the
activity of the respiratory centre. :

In the light of the present series of experiments we are, for the
present, content to abide by our original conclusions that (1) “ the
amplitude or force of the respiration, as indicated by the excursion
of the writing-point of the tambour, usually is less during high
pressure, but shows a constant augmentation when the pressure
falls’; and (2) “that, under the conditions of our experiments,
the usual effect of any considerable increase in blood pressure is an
increase in the respiratory rate with diminished amplitude, while a
fall in pressure causes a slower rate with greater amplitude.” We
do not believe that these results are due, in all cases, to our method
of occlusion.

Pee R STUDIES ON THE RELATION OF THE
we YGEN SUPPLY OF THE SALIVARY GLANDS
SO THE COMPOSITION OF THE SALIVA,

By A. J. CARLSON anp F. C. McLEAN.
[from the Hull Physiological Laboratory of the University of Chicago.]

N a recent communication to this Journal by Carlson, Greer, and
Becht,’ it was shown that when, during the stimulation of the
dog’s chorda tympani, the blood supply to the submaxillary gland is
artificially diminished to the same degree as during the stimulation
of the cervical sympathetic, the percentage composition of the chorda
saliva becomes identical with that of the sympathetic saliva. By
greatly diminishing the oxygen supply to the gland during activity
following chorda stimulation the rate of secretion is greatly dimin-
ished, and the percentage of organic solids in the saliva increased to
a point equal or exceeding that of ordinary sympathetic saliva.
These results are contrary to Heidenhain’s observations, but in view
of the fact that they were obtained constantly the conclusion seemed
inevitable that the cause of the normal difference between saliva
obtained by stimulation of the cranial secretory nerves and that
. following the stimulation of the cervical sympathetic is the differ-
ence in the vascular condition of the gland rather than a difference
in the character of the cranial and sympathetic secretory nerves.
The existence of true sympathetic secretory fibres to the dog’s sub-
maxillary gland was also demonstrated, and evidence adduced that
on abundant oxygen supply to the salivary gland the sympathetic
saliva becomes just as dilute as that obtained on stimulation of the
chorda or of Jacobson’s nerve.

These facts seem to render the theory of trophic secretory nerves
superfluous. But, as pointed out in that report, one apparent fact
still remains to be brought into line with the above results before
the theory of trophic secretory nerves as stated by Heidenhain can

1 CARLSON, A. J., GREER, J. R.,and Becut, F. C.: This journal, 1907, xx, p. 180.
457

458 A. F. Carlson and F. C. McLean.

be dispensed with entirely. Heidenhain’s experiments seemed to
show, namely, that the stimulation of the cervical sympathetic in-
creases the percentage of organic solids in the subsequent chorda
saliva. The data reported by Carlson, Greer, and Becht are not con-
clusive on this point. They show, to be sure, that a previous stimu-
lation of the sympathetic does not invariably increase the percentage
of organic solids in the subsequent chorda saliva. But some pre-
cautions necessary to render the results conclusive were not taken in
those experiments, as they were carried out with a different purpose.
The main purpose of the experiment now reported was the examina-
tion of Heidenhain’s observations touching this question, and to
determine whether this supposed effect of the sympathetic stimula-
tion on the subsequent chorda saliva can be duplicated by previous
anemia of the gland secured by other means than the stimulation
of the sympathetic. We may say at the outset that we failed to
confirm Heidenhain’s observations. His results were due to experi-
mental errors, as will be made evident in this report.

Further data are also required on the variations of the organic
solids in the chorda saliva obtained during a single period of stimu-
lation, a period of stimulation yielding from ten to fifteen c.c. of
the saliva. It is well known, from the researches of Ludwig? and
Heidenhain, that the chorda saliva grows poorer in organic solids
part passu with the increasing fatigue of the gland. Does this rela-
tion obtain in the case of the first five or ten c.c. of saliva yielded
by the gland on chorda stimulation? It was shown by Carlson,
Greer, and Becht in the previous report that after a period of gland
rest the first c.c. or half c.c. of saliva is much richer in organic
solids than the subsequent two or three c.c. In fact, the concentra-
tion of the organic solids in the first c.c. of chorda saliva following ~
a period of rest is usually as great as that in ordinary sympathetic
saliva. In the case of a single period of stimulation yielding five c.c.
of saliva, does the same relation as regards the organic solids obtain
between the fifth and the fourth c.c. as between the first and the
second c.c.? This point is obviously of importance, because, if we
have such a gradual decline in the percentage of organic solids
during any single period of stimulation, it is necessary to use exactly
the same amount of saliva, secured at exactly corresponding stages
in the gland activity, when comparisons are to be made of the

* BECHER and Lupwic: Zeitschrift fir rationelle Medizin, 1851, N. F. i,
p. 278.

Oxygen Supply of the Salivary Glands. 459

percentage composition of saliva under varying conditions of the
gland.

Unless otherwise stated, the same experimental methods are fol-
lowed as in the work by Carlson, Greer, and Becht.

J. Tue GrapDuAL DIMINUTION IN ORGANIC SOLIDS OF THE
CHORDA SALIVA DURING A SINGLE PERIOD OF STIMULATION.

The data from our several experiments on this point are given in
Table I. In Experiment 5 the organic solids in the second sample
are less than that in the third sample, but in the other experiments
there is a progressive diminution in the organic constituents during’
the period of secretion. In Experiments I and 4 the difference be-

TABLE, I.

Submaxillary chorda saliva. Dog. Variations in the percentage composition of the
saliva yielded in a single period of chorda stimulation with the weak interrupted
current. The three samples represent the total amount of saliva yielded by the
stimulation, except in Experiments 6 and 7.

wat, (eeetor same Onanbiey Solids per 100 c.c. saliva.

ples of saliva of saliva
collected. in c.c.

experiment.

Total. | Inorganic. Organic.

0.31
0.55
0.57

0.25
0.31
0.51

0.23
0.38
0.48

a)
md Ww
Ane enon

i)
ty UN ON
IT G2 G2
RSS Bas

wn
io)

ome Orr ocr

CCE Same guinness
NW WwW
Rese aS
COW
MAD ur

460 A. F. Carlson and F. C. McLean.

tween the second and third sample is very slight, however. In Ex-
periments 2, 3, 6, and 7 the difference is considerable.

This gradual diminution in the organic solids during a single
period of stimulation is, at least in part, independent of the rate of
secretion. According to Heidenhain, the stronger the stimulation
of the chorda, and therefore the greater the rate of secretion, the
greater the percentage of organic solids in the saliva, other things
being equal. Even if this is a fact, it cannot be the cause of the dif-
ference between the first c.c. and the next two or three c.c. of saliva
in our experiments, because by varying the strength of stimulation
the first c.c. may be secured at a slower rate than the following
sample, and yet the first sample is richer in organic constituents.
When the strength of the stimulation is so graduated that the gland
is to yield only fourteen or fifteen c.c. of the saliva in that period,
the last five c.c. will necessarily be secreted at a slower rate than the
first or second samples. But the same difference in the percentage
of organic solids of the second and third samples appears even when
the strength of the stimulation is varied so that the rate of secretion
remains nearly constant. This was done in Experiments 6 and 7,
Table I.

It is not within the scope of this report to consider the cause of
this variation. The rate of diminution of the organic solids appears
in some cases to be too great to be due solely to actual diminution
in preformed secretion granules in the gland cells, although this is
in all probability one of the factors. Now, inasmuch as the com-
position of the chorda saliva varies from one drop to another during
any single period of stimulation, and furthermore, since its com-
position varies with the length of the intervals of rest between any
two periods of stimulation, it is evident that comparisons between
the composition of samples of saliva from the same gland may lead
to faulty conclusions unless these facts are taken into consideration.
It is necessary that the gland have equal periods of rest preceding
the periods of activity, that the oxygen supply remain the same on
stimulation, and that comparisons be made between the same quan-
tities of saliva secured during corresponding stages of the period
of gland activity.

Oxygen Supply of the Salivary Glands. 461

IJ. Tue INCREASE IN THE ORGANIC CONSTITUENTS OF PILo-
CARPIN PAROTID SALIVA ON STIMULATION OF THE CERVICAL
SYMPATHETIC AND ON ARTIFICIALLY DIMINISHING THE Oxy-
GEN SUPPLY.

The work of Carlson, Greer, and Becht on the relation between
the blood supply to the gland and the composition of the saliva was
confined to the submaxillary gland of the dog and the cat. No
work was.done on the parotid gland. There is this difference
between the submaxillary and the parotid gland of the dog, the cat,
and the rabbit, that the cervical sympathetic always causes secretion
from the former, but none at all or at the most a barely perceptible
secretion from the latter. Yet we know, from the work ot Heiden-
hain and others, that by stimulation of the sympathetic in the neck
during the stimulation of Jacobson’s nerve or during the secretion
of the parotid under pilocarpin, the flow of saliva is diminished and
the percentage of organic solids contained in it greatly increased.
Heidenhain considers this fact as a most striking evidence of the
presence of trophic secretory fibres to the parotid gland in the cer-
vical sympathetic; and so it is, to be sure, unless it can be shown
that this effect is not due to the diminished blood supply to the
gland caused by the sympathetic stimulation. Heidenhain found,
apparently, that diminished blood supply does not alter the composi-
tion of the saliva, and most of his experiments touching this point
were made on the parotid gland.

There is little doubt that what is true for the submaxillary gland
is also true for the parotid; and the results of Carlson, Greer, and
Becht on the former show that typical sympathetic saliva is yielded
on chorda stimulation, provided the oxygen supply to the gland 1s
greatly reduced. We decided, however, to repeat the experiments
on the parotid, partly in hope of learning how Heidenhain came to
get only negative results. As Jacobson’s nerve is rather difficult
to isolate for direct stimulation, we made use of intravenous injec-
tion of pilocarpin for producing gland activity.

Care was taken, in collecting the saliva, to secure pure or true
samples. When the change is made from the normal pilocarpin
saliva to the pilocarpin saliva secreted during sympathetic stimula-
tion or during venous or arterial occlusion, it is not enough to
remove the normal saliva from the collecting cannula, but it is also
necessary to discard the first two or three drops yielded under the

462 A. F. Carlson and F.C. McLean.

new conditions, as there is at the start at least that quantity of
normal saliva remaining in the passages of the gland from the pre-
ceding activity.
TABLE SLE:
Parotid saliva following intravenous injection of pilocarpin. ‘The effect on the percentage
composition of stimulation of the cervical sympathetic (Nos. 1-4) and of diminution

of the oxygen supply to the gland by occlusion of the gland veins or the gland
arteries (Nos. 5-10).

Solids per 100 c.c. saliva.

Animal. | Character of the pilocarpin saliva.

Total. |Inorganic.| Organic.

. Normal
. During symp. stim.

0.62 2.61
0.17 4.44

0.08 0.45
0.10 2.56

0.46 0.60
0.32 1.90

0.42 0.80
0.35 1.87

033 2.11
0.17 2.85

0.57 1.91
0.51 2.20

0.26 2.62
0.20 4.88

Cat . Normal | 0.45 Pa
. R. & L. carotids & R. vert. ; : 0.42 4.47

Oto
28

. Normal
. During symp. stim.

wo +O

. Normal

Rabbit ||) Piloc!/sal. during symp. stim.

BR Qe

: : . Normal
Rabbit | 9° During symp. stim.

tN bX
NL

. Normal
Cat . Gl. vein occl.

. Normal
Cat . Gl. vein occl.

ae St

. Normal

Cat . R. & L. carotids & R. vert.

1.
2.
2
3.
2.
2
2
Sy
3H

Oo (ora)
S &&

Cat . Normal A 0.45 2.46
. R.& L. carotids & R. , i 0.52 2.42

Cat . Normal i. 0.33 2.40
. R. & L. carotids & R. ; : 0.25 SWS

Stimulation of the cervical sympathetic during pilocarpin secre-
tion diminishes the rate of secretion and increases the percentage
of organic solids in the saliva. This is true for the cat, the dog,
and the rabbit. The relative increase of the organic solids varies.
Our results are given in Experiments 1 to 4, Table II. Some of
the analyses reported by Heidenhain show a relatively greater
increase.

An examination of Experiments 5 to 10, Table II, shows that this

Oxygen Supply of the Salivary Glands. 463

effect of the sympathetic stimulation is duplicated by diminishing
the oxygen supply to the gland. In Experiments 5 and 6 the oxygen
supply was diminished by occlusion of the gland vein; in the rest
of the experiments the greatly diminished blood supply to the gland
was secured by ligation of the innominate and the left carotid, leav-
ing the left vertebral unobstructed. This ligation usually reduces
the amount of blood passing through the parotid to a degree equal
to that on stimulation of the sympathetic with a moderately strong
interrupted current.

In Experiment 9 there was no appreciable slowing of the secre-
tion during the ligation of the arteries, as compared to the normal
rate previous to the ligation, and in that particular experiment the
two samples of saliva exhibit practically no difference in percentage
composition. It is probable that the degree of diminution of the
blood supply to the gland in this case was not sufficient. In all the
other experiments the saliva secreted during the diminished blood
supply is much richer in the organic and slightly poorer in the in-
organic solids, just as is the case with the sympathetic saliva in
Experiments 1 to 4. The venous occlusion appears to produce less
change than the arterial obstruction.

These results on the parotid and the previous results on the sub-
maxillary gland render it highly probable that the relation between
the oxygen supply to the gland and the percentage composition of
the saliva holds true for all the salivary glands of mammals. The
only explanation that we can offer of Heidenhain’s negative results
is that he did not diminish the oxygen supply to the same degree
that obtains on sympathetic stimulation, or perhaps did not take
sufficient care in removing the normal saliva from collecting can-
nula and gland before securing the sample of saliva secreted during
the period of diminished blood supply.

III. Previous STIMULATION OF CERVICAL SYMPATHETIC DOES NOT
INCREASE THE PERCENTAGE OF ORGANIC SOLIDS IN THE SUB-
SEQUENT CHORDA OR PILOCARPIN SALIVA.

1. Heidenhain found that the saliva secreted on stimulation of
the cranial secretory nerves after a previous stimulation of the cervi-
cal sympathetic is richer in organic solids than the normal saliva.
This he attributes to the actual increase in the organic materials in
the cells destined to pass into the saliva because of the stimulation

464 A. F¥. Carlson and F. C. McLean.

of trophic secretory fibres in the sympathetic. This observation of
Heidenhain does not appear to have been subjected to criticism, and
seems even to-day to be generally accepted as correct.

We know that partial anemia of the gland increases the per-
centage of organic solids in the chorda saliva by diminishing the
secretion of water and inorganic salts. It does not appear from
Heidenhain’s account that he allowed a sufficient period to elapse
between the cessation of the sympathetic stimulation and the begin-
ning of the chorda stimulation to assure the return to the normal
vascular conditions in the gland. The first part of the chorda saliva
may therefore have been secreted under conditions of diminished
oxygen supply. Care must also be taken to discard all of the sym-
pathetic saliva remaining in the gland from the previous stimulation.
These precautions were taken in our experiments. A period of
from five to seven minutes was allowed to elapse between the sym-
pathetic and the chorda stimulation or between the sympathetic
stimulation and the collection of the sample of normal pilocarpin
saliva in the case of the parotid gland. It is hardly necessary to state
that the sympathetic saliva was removed from the collecting can-
nula, and the first three or four drops of chorda or pilocarpin saliva
discarded, as being secreted during the sympathetic stimulation.

Our results are given in Table III. Of the sixteen experiments
thirteen are on the dog’s submaxillary, one on the dog’s parotid,
and two on the rabbit’s parotid. The data in Experiments 5 to 12
are taken from the paper by Carlson, Greer, and Becht already re-
ferred to. Those experiments were made with another end in view;
but as most of the above-stated precautions were taken in collecting
the samples, these analyses are available for this comparison. Ex-
periments 1 to 4 and 13 to 16 were made purposely to test Heiden-
hain’s observation.

An examination of Table III reveals the fact that in three cases
(Nos. 2, 8, and one series of No. 12) the chorda saliva following
the sympathetic stimulation is slightly richer in organic solids than
that preceding it. In four cases (Nos. 3, 4, 9, 14) the percentage
of ‘organic constituents is identical in the two samples, while in the
remaining eight experiments the saliva following the sympathetic
stimulation is poorer in organic solids than that preceding it.

It is evident from these results that stimulation of the sympa-
thetic does not increase the organic solids in the subsequent chorda
saliva, if time is given for the gland to recover from the anemia

Oxygen Supply of the Salivary Glands. 465

TABLE III.

The effect of previous stimulation of the cervical sympathetic on the percentage compo-
sition of chorda submaxillary saliva and pilocarpin parotid saliva. In Experiment 13
corresponding samples of chorda saliva were collected from the right gland, only the
left sympathetic being stimulated. A period of about five minutes was allowed be-
tween the cessation of the sympathetic and the beginning of the chorda stimulation.
The first two or three drops of chorda or pilocarpin saliva following stimulation of
the sympathetic were discarded as being sympathetic saliva.

Duration
of stim. of
the cervi- Chorda saliva.

cal sympa- - - 5
thetic, Total. |Inorganic.} Organic.

Solids per 160 c.c. saliva.
Number of
experiment.

. Normal 32 0.30 1.02
. After symp. 1.01 0.25 0.76
Normal 1.23 0.32 0.91
. After symp. 1.26 0.25 1.01
Normal 0.83 0.17 0.66
. After symp. 0.91 0.29 0.62
Normal 0.97 | 0.26 0.71
. After symp. 1.03 0.31 0.72
Normal 2.12 0.29 1.83
. After symp. Wal 5 O25 0.96
. Normal 1.85 0.43 1.42
. After symp. 1.63 0.40 1.23
Normal 2.21 0.52 1.69
. After symp. 2.01 0.55 1.45
. Normal 1.08 0.24 0.84
. After symp. 1.15 0.26 0.89
Normal 1.16 0.22 0.94
. After symp. 1.17 0.26 0.91
Normal 1.46 0.22 1.24
. After symp. 1.30 0.27 1.03
Normal 4 0.55 1.11
. After symp. ie 0.71
Normal # L 1.00
. After symp. : : 1.24
. After 2d symp. stim. - E 0.59

. Normal. ; es : ; an

. After symp. j R. (0.90)
L. 26 | 0.96

8 min.

7 min.

7 min.
8 min.
About
10 min.
About
10 min.
About
10 min.
About
10 min.
About
10 min.
About
10 min.

About
10 min.

About
10 min.

30 min.

stim. L. side
. After 2d period R.

OS MH WHE DRE NE DERE NEE DE DE DE NE Ne

: 5 0.70)

30 min. of symp. stim. (

L. side 0-76
. Piloc. sal. normal Et u 0.46
. After symp. stim. ! 0.43
. Piloc. sal. normal | : k 0.60
. After symp. stim. L Fe 0.23
- Piloc. sal. normal ; E 0.80
. After symp. stim. bs ; 0.10

14
Dog’s parotid
15
Rabbit’s parotid

45 min.

25 min.

25 min.

16
Rabbit’s parotid

466 A. F. Carlson and F. C. McLean.

prior to stimulation of the chorda or the injection of the pilocarpin.
There is not the slightest basis for the trophic nerve theory in the
after effects of sympathetic stimulation. In the three cases where
we obtained positive results the difference is not great enough to
justify any conclusions. On the other hand, the organic solids may
be considerably scantier in the chorda saliva after a previous stimu-
lation of the sympathetic. The cause of this appears to us quite
obvious. The stimulation renders the gland greatly anemic, and in
consequence interferes, in all probability, with the processes of build-
ing up during gland rest the organic material that passes into the
gland during activity. And, secondly, in case the sympathetic
stimulation actually causes the gland to secrete, the gland is of
course more exhausted after the stimulation than before it; and we
know now that the percentage of organic solids in the saliva dimin-
ishes gradually during a single period as well as during successive
periods of activity.

IV. PartraL ANEMIA OF THE GLAND DOES NOT INCREASE THE
PERCENTAGE OF ORGANIC SOLIDS IN THE SUBSEQUENT
CHORDA SALIVA UNLESS THE CHORDA STIMULATION IS BE-
GUN BEFORE THE GLAND HAS RECOVERED FROM THE ANEMIA.

Further evidence that the increase in saliva solids, in case the
stimulation of the sympathetic is immediately followed by excita-
tion of chorda, is due to the partial anemia of the gland, is secured
by the effect on the chorda saliva obtained immediately after a
partial occlusion of the gland artery or the gland vein. Our ex-
periments on this point are given in Table IV. Partial anemia of
the submaxillary gland was again produced by occlusion of the main
gland vein, by partial occlusion of the gland artery directly, or, in
the dog, by temporary ligation of both carotids and both vertebrals.

In the first six experiments the chorda stimulation was begun
immediately on release of the gland veins or the occluded arteries.
In each case the chorda saliva then secured is more concentrated
than that preceding the interference with the gland circulation. The
analyses in the case of the first two experiments go to show that this
increase is in the organic constituents. Under these experimental
conditions the gland is necessarily anemic during the initial part of
the secretion, and as a consequence we have the typical chorda saliva
of anemia.

Oxygen Supply of the Sahvary Glands. 467

In Experiments 7 and 8 a period of five to six minutes was
allowed to elapse between the restoration of the normal gland cir-
culation and the beginning of the chorda stimulation. Under these

TABLE IV.

The effect on the percentage composition of saliva of previous anemia of the gland by
venous or arterial occlusion. In Experiments 1-6 the chorda stimulation began im-
mediately on restoration of the normal blood supply to the gland. In 7 and 8 the
stimulation was begun five minutes after restoration of the circulation.

Solids per 100 c.c. saliva.

No. of experiment. Chorda saliva.

Total. |Inorganic.| Organic.
|

Cat’s subm.

2
Cat’s subm.

5

Dog’s subm.

4
Dog’s subm

5
Dog’s subm

6
Dog’s subm

7
Dog’s subm

sub. artery

- Normal
. After 20 min. occl. of the |

main subm. vein.

. Normal
. After 10 min. partial occl. |

of gland artery

. Normal
. After 10 min. occl. of main

gl. vein

. Normal
. After 10 min. occl. of both |

carotids and vertebrals

. Normal
. After 10 min. occl. of both |

carotids and vertebrals

. Normal
. After 10 min. occl. of both

. Normal
| ‘ . After 15 min. partial occl. ; ou Od

0.32

0.29
0.25

0.66

0.66
0.74

carotids and vertebrals

. Normal

. After 15 min. occl. of both
carotids and vertebrals

8
Dog’s subm

conditions the saliva following the anemia is either of the same or
of a less concentration in organic solids than the normal.

The conclusion is therefore evident that the effect of sympathetic
stimulation on the percentage composition of the subsequent saliva
secured by stimulation of the cranial secretory nerves can be dupli-
cated by producing partial anemia of the gland prior to such stimu-
lation. This effect is in each case due to the gland anemia at the
beginning of the gland activity, and does not appear when the con-
dition of anemia no longer obtains.

468 A. F. Carlson and F. C. McLean.

SUM MARY.
i

During any single period of activity of the submaxillary gland
produced by direct stimulation of the chorda tympani there is
usually a gradual diminution in the percentage of organic solids
in the saliva. This gradual diminution in the organic solids is, at
least under some conditions, independent of the rate of secretion.

In the case of the parotid gland of the dog, the cat, and the
rabbit, diminution of the oxygen supply diminishes the secretion
rate and increases the percentage of organic constituents of the
pilocarpin saliva. This relation has been previously demonstrated
for the submaxillary chorda saliva of the dog and the cat. It is
probably true for the salivary glands of all mammals.

Stimulation of the cervical sympathetic during pilocarpin activity
of the parotid gland retards the rate of secretion and increases the
percentage of organic constituents in the saliva. Both of these
actions of the sympathetic are duplicated by diminution of the
oxygen supply to the active gland by means of venous occlusion or
obstruction of the arteries. It is therefore probable that the change
in the parotid pilocarpin saliva produced by the sympathetic stimula-
tion is due to the vaso-constriction and not to the stimulation of
trophic secretory nerves.

Stimulation of the cervical sympathetic does not increase the
percentage of organic solids in the subsequent saliva obtained by
stimulation of the cranial secretory nerves or by the injection of
pilocarpin. Such an increase is evident in case the stimulation
of the cranial secretory nerves follows immediately on cessation of
the sympathetic stimulation. This effect is due to the partial anemia
of the gland from the sympathetic vaso-constrictor action, and does
not appear when time is given for the normal vascular condition
to be re-established in the gland. These effects of the sympathetic
stimulation are duplicated by diminishing the oxygen supply to
the gland by means of venous occlusion or obstruction of the
arteries.

These facts, together with the results obtained by Carlson, Greer,
and Becht, render the Heidenhain theory of trophic secretory nerves
to the salivary glands untenable. Still, the salivary glands may be
supplied with “trophic” nerve fibres acting in a way quite dif-
ferent from that conceived by Heidenhain. The atrophy of the

a

eee ee nnn a ee

Oxygen Supply of the Salivary Glands. 469

salivary glands following section of the cranial secretory nerves
indicates that the life of these cells, and therefore the building
up of secretion material, is dependent on impulses reaching them
through these nerves. We probably have “ gland tone,” analogous
to ‘“‘ muscle tone,” sustained by impulses from the ordinary secre-
tory nerves.

HYDROLYSIS OF AMANDIN FROM THE ALMOND.!

By THOMAS B. OSBORNE Anp 5S. o. CLARE

[From the Laboratory of the Connecticut Agricultural Experiment Station.|

HE chief part of the protein substance of the almond (Prunus

amygdalus var. dulcis) consists of a globulin to which the
name amandin has been given. An investigation of the proteins
of the almond made in this laboratory” gave no evidence of the
presence of more than one globulin in this seed. Amandin is char-
acterized by its high content of nitrogen, about Ig per cent, and by
the large proportion of nitrogen which it yields as ammonia on
hydrolysis.

In preparing the amandin which was used for this hydrolysis we
received the assistance of Mr. J. F. Harris, to whom we wish
to express our thanks.

The almonds were freed from their outer coating by immersing
them for an instant in hot water and from the greater part of their
oil by pressure. The remainder of the oil was removed by petro-
leum benzine and the meal ground to a fine powder.

The oil-free meal was then extracted with one tenth saturated
ammonium sulphate solution, the extract filtered clear, and enough
crystals of ammonium sulphate dissolved in it to bring its concen-
tration up to four tenths saturation. The precipitate thus produced
was filtered out, dissolved in dilute sodium chloride brine, and the
solution, after filtering perfectly clear, was dialyzed for several days.
The amandin was thus precipitated for the most part as a somewhat
gummy and coherent mass. This, when washed with water and
with alcohol, dehydrated with absolute alcohol and dried over sul-
phuric acid, formed a snow-white powder.

1 The expenses of this investigation were shared by the Connecticut Agricul-
tural Experiment Station and the Carnegie Institution of Washington, D. C.
? OSBORNE and CAMPBELL: Journal American Chemical Society, 1896, xviii,
p- 609.
470

Hydrolysis of Amandin from the Almond. 471

Five hundred grams of this amandin, equal to 457.2 gm. moisture
and ash-free, were suspended in a mixture of 500 c.c. of water and
500 c.c. of hydrochloric acid of specific gravity 1.19 and warmed
on the water bath until solution was nearly effected. The hydroly-
sis was then made complete by boiling the solution in the oil bath
for fourteen and a half hours.

A preliminary removal of glutaminic acid, performed in the usual
manner, yielded 83.64 gm. of the hydrochloride, or 67 gm. of the
free acid. The filtrate from glutaminic acid hydrochloride was then
concentrated under reduced pressure very sharply, and the residue
esterified with alcohol and dry hydrochloric acid gas according to
the directions of Emil Fischer.

After liberating the free esters from the hydrochlorides and shak-
ing out with ether, the aqueous layer was freed from inorganic salts
and the esterification repeated.

_ The combined esters were then distilled under diminished pres-
sure with the following result:

Temp. of bath

Fraction. up to Pressure. Weight.
I 80° o.8 mm. 14.32 gm.

II g2° 6.5 19.94 “

Ill 100° 6. 701-7 5a.05y5

IV 134° 0.70 “ 50:79 “

Vi 150° O.7he: SO=tqars:

VI 200° 6.70 42-804 -
Matglereh.. ss pe) oes Net. Vevey

The undistilled residue weighed 68 gm.

Fraction I.— This fraction yielded 1.10 gm. of the hydrochloride
of glycocoll ethyl ester. The melting-point was 144°.

Chlorine, 0.2866 gm. subst., gave 0.2925 gm. AgCl.
Calculated for C,H,90,NCl = Cl 25.45 per cent.
Bound a2 .t ite «3. cs == Cea aa es

The filtrate was added to the glycocoll filtrate from Fraction II.

Fraction II. — The esters were saponified with boiling water, the
solutions evaporated to dryness under reduced pressure, and the
residue extracted with boiling alcohol to remove the proline.

The insoluble portion was re-esterified and the glycocoll brought
to separation as the hydrochloride of the ethyl ester. The yield

472 Thomas B. Osborne and S. Hl. Clapp.

was 3.30 gm. By fractional crystallization of the free amino acids
from water and from water and alcohol, there were further obtained
from Fraction II 1.50 gm. of leucine, 6.45 gm. of alanine, 0.75 gm.
of substance having the percentage composition of amino-valerianic
acid and a fraction of perfectly definite appearance, which on an-
alysis gave results agreeing closely for a mixture of equal parts of
leucine and amino-valerianic acid.

Carbon and hydrogen, 0.1221 gm. subst., gave 0.2368 gm. CO, and 0.1087 gm.
H,O.
Calculated for equal molecules of leucine and amino-valerianic acid =
C5312; H 9:66 percent,
Found.- <2... 4 S==1G362.89 #l o:65 percent:

The alanine decomposed at about 290° and gave the following
analysis :

Carbon .and hydrogen, 0.1542 gm. subst., gave 0.2294 gm. CO, and 0.1163 gm.

H,O.
Calculated for C;H,O.N = C 40.45 ; H 7.86 per cent.
Found 3 5 5 A —= Gc 40.573 H Be “ 6c

The amino-valerianic acid was analyzed as follows:

Carbon and hydrogen, 0.1027 gm. subst., gave 0.1931 gm. CO, and o.ogor gm.

H,0O;
Calculated for C;H,,0,N = C 51.28; H 9.40 per cent.
Found, ¢2: 45, 25 0. tee 28 or Oy eee

The preparation of amino-valerianic acid seemed to be homo-
geneous under the microscope, but, owing to the small amount of
substance at our disposal, we are unable to offer any real chemical
evidence of the existence of this substance in the protein.

Fraction I11.— This fraction was saponified and the proline ex-
tracted in the usual way. The part remaining undissolved in alcohol
yielded 19.04 gm. of very nearly pure leucine.

Carbon and hydrogen, 0.4203 gm. subst., gave 0.4840 gm. CO, and: 0.2154 gm.

H,0.
Calculated for C,H;;0,.N — (& 54-96 ; H 9-92 per cent.
Found ~ - ‘, - = G 54-933 H 9-95 “ “

The substance decomposed at about 298°.
In the filtrate from leucine no definite substance could be isolated.
The proline extracts of Fractions II and III were united.

es

Hydrolysis of Amandin from the Almond. 473

The substance was converted to the copper salt, and the levo

‘separated from the racemic with boiling absolute alcohol. The yield

of air-dry racemic proline copper was 2.61 gm., while of the amor-
phous copper salt of levo-proline there were obtained 11.98 gm.
dried at 110°. The racemic proline copper salt was analyzed as
follows:

Water, 0.1163 gm. subst. (air-dried), lost 0.0124 gm. H,O at 110°.
Calculated for C,)H,,O,N,Cu * 2 H,O = H.O 10.99 per cent.
Bounds 2°. c= = H.One.667" 4
Copper, 0.1030 gm. subst; gave 0.0278 gm. CuO.
Calculated for Cj)H,,0,N.Cu = Cu 21.81 per cent.
POUnG eens SE a= OB ee

The levo-proline was identified as the characteristic phenyl-
hydantoine. The melting-point was 142°-143°.

Carbon and hydrogen, 0.1332 gm. subst., gave 0.3244 gm. CO, and 0.0672 gm.

H,O.

2

Calculated for C;.H,;,0.N, = C 66.67 ; H 5.57 per cent.
Peps. 7 9. 0. ha == O6:42 5 HL 5.60-

Fraction Iv.— The ester of phenylalanine was removed in the
usual manner by shaking out with ether. The yield of phenyl-
alanine hydrochloride was 2.94 gm. As from Fractions V and VI
there were further obtained 11.35 gm. of the hydrochloride, the
total yield of free phenylalanine from amandin was II.71 gm., or
2.53 per cent. The substance was identified as the copper salt.

Copper, 0.1133 gm. subst., gave 0.0227 gm. CuO.
Calculated for C,;H0,;N.Cu = Cu 16.24 per cent.
anne ae) tts.) = On r60nr/ Got

From Fraction IV there were further obtained 4.55 gm. of
aspartic acid as the barium salt and 12.74 gm. of air-dry copper
aspartate. From this fraction no glutaminic acid could be isolated
as the hydrochloride.

Fraction v.— There were further obtained from this fraction
6.97 gm. of aspartic acid as the barium salt, 1.06 gm. of glutaminic
acid hydrochloride, and 15.25 gm. of air-dry copper aspartate. The
latter substance, crystallized in the tyrosine-like bundles of needles
and dried in the air, gave the following analysis:

474 Thomas B. Osborne and S. H. Clapp.

Copper, 0.1381 gm. subst., gave 0.0396 gm. CuO.
LVitrogen, 0.2100 gm. subst., required 1.12 c.c. 5/7 N—HCI.
Calculated for C,H;0,NCu 4} H,O = Cu 23.07; N 5.08 per cent.
Found)... .:s. > atv, <i) = 1Gu cor 5) Nive see

The aspartic acid from the barium salt reddened but did not
decompose at 300°.

farbon and hydrogen, 0.2600 gm. subst., gave 0.3458 gm. CO, and 0.1292 gm.

H,0.
Calculated for CsH,O,N = C 36.09 ; H 5.26 per cent.
Found... «de ,2ae eee gable ica
Fraction VI.— The aqueous layer of this fraction, remaining

after the removal of phenylalanine ester with ether, was saponified
with baryta in the usual way. After prolonged standing no barium
aspartate had separated. The barium was accordingly removed and
the concentrated solution separated 14.16 gm. of pure glutaminic
acid. The decomposition point was at about 198°—199°.

Carbon and hydrogen, 0.1840 gm. subst., gave 0.2759 gm. CO, and 0.1064 gm.

lgROR
Calculated for C;H,O,N = C 40.81 ; H 6.12 per cent.
Found». 42° <= "E 40.89 ; H 6.42 é T:

There were further obtained from this fraction 3.4 gm. of
glutaminic acid hydrochloride, while no copper aspartate could be
obtained.

TYROSINE.

Fifty grams of amandin, moisture and ash free, were boiled with
a mixture of 150 gm. sulphuric acid and 300 c.c. of water for ten
hours. After removing the sulphuric acid with an equivalent quan-
tity of baryta, the filtrate and washings were concentrated to small
volume, and the crystalline substance that separated on standing
was filtered out, washed well with water, and recrystallized. The
tyrosine thus obtained in characteristic needles weighed 0.5635 gm.
equal to 1.12 per cent.

The solution from which the tyrosine separated, together with
the mother liquor from the recrystallization, was concentrated and
a second separation obtained. This was dried and treated with

flydrolysts of A mandin Jrom the Almond. 475

glacial acetic acid. All was dissolved, and no more tyrosine could
be obtained from the hydrolysis solution.

WVitrogen, 0.5630 gm. subst., required 4.3 c.c. 5/7 N—HCIl.
Calculated for C,H,;,0;N = N 7.73 per cent.
PGBHGie aes oe aN oa

HISTIDINE.

Fifty grams of moisture and ash free amandin were boiled for
ten hours with a mixture of 150 gm. sulphuric acid and 300 c.c.
of water, and the bases isolated according to the direction of Kossel
and Patten. The solution of the histidine was made up to 300 c.c.
and nitrogen determined in an aliquot part of it.

Vitrogen, to c.c. solution required 0.58 c.c. 5/7 N—HCl= 0.0058 gm. N =
0.1740 gm. N in 300 c.c. = 0.6438 gm. histidine, or 1.29 per cent.

Another determination, made with the same amount of amandin
boiled for fourteen hours with the addition of 15 gm. sodium chlo-
ride,® gave a little higher result, namely, 1.87 per cent.

LVitrogen, 20 c.c. solution, required 1.69 c.c. 5/7 N—HCl = 0.0169 gm. N =
0.2535 gm. N in 300 c.c. = 0.9379 gm. histidine = 1.87 per cent.

The remainder of the histidine was converted into the dichloride,
but, owing to a loss, its identity was not established.

ARGININE.

The solution from this second hydrolysis, containing the arginine,
was made up to 1000 c.c.

LVitrogen, 20 c.c. solution, required 3.68 c.c. 5/7 N—HCl = 0.0368 gm. N =
1.840 gm. N in 1000 c.c. = 5.9189 gm. arginine = 11.85 per cent.

The rest of the arginine was identified as the copper nitrate
double salt.
Water, 0.2223 gm. subst., air dry, lost 0.0217 gm. H,O at 110°.
Calculated for C,,.H:;0,N;Cu(NO3)2 3 HzO = H.O g.16 per cent.
eee te Os se) dy a, OR GEG) : Su

* Cf. Hart, Zeitschrift fiir physiologische Chemie, 1901, xxxiii, p. 348.

476 Thomas B. Osborne and S. FH. Clapp.

Copper, 0.1983 gm. subst. (dried at 110°), gave 0.0292 gm. CuO.
Calculated for Cy.H.»04N,Cu ied = Cui 1.07 per cent.

Found

=u biog a

LYSINE.

The lysine in the two hydrolysis solutions was converted into the
picrate, of which 0.5845 gm. was obtained in the solution to which
the sodium chloride had been added and 0.8833 gm. in the soiution
without this salt. This is equivalent to 0.2277 gm. and 0.3441 gm.

lysine respectively, or 0.46 and 0.70 per cent. of the amandin.

Nitrogen, 0.4260 gm. subst., required 7.9 c.c. 5/7 N—HCl (Kjeldahl-

Jodlbauer).
Calculated for C,H,;,O.N. ° C,H
Found

The results of this hydrolysis were the following:

Per cent.

Glycocoll, os.) jis 92, ghost

AlanineG: = i505 a9 Hot eeeO
Welitie™s;;" "ayia, a fans tec OemtO
Liewcme: 0%. -.)) Many? 2% bales
Proline’. ia SA eed (eee
Phenylalanine — (-° ae 2a5s
Asparticaeid) 21) OS os n eae
Glutaminic:acid, ©\'\5, .y2ganqe

4 OSBORNE and GILBERT: This journal, 1906, xv, p. 350.

30,N, = N 18.67 per cent.

= INGE Oca ie:

Serine
Tyrosine
Arginine
Histidine
Lysine .
Ammonia
Tryptophane

Per cent.
?
I.12

« Tle5

1.58
0.70

ae

. present

59-00

BYDROLYSIS OF THE PROTEINS OF- MAIZE,
ZEN MAYS.

By THOMAS B. OSBORNE anp S. H. CLAPP.

[¥rom the Laboratory of the Connecticut Agricultural Experiment Station.]

HE seeds of maize or Indian corn, Zea Mays, like those of the

other cereals, contain a very small proportion of protein sol-
uble in water or neutral saline solutions, a relatively large amount
of protein soluble in strong alcohol, and a considerable quantity in-
soluble in all neutral solvents but soluble in very dilute alkaline and
acid solutions.

An investigation of the proportions of these several proteins was
made some years ago in this laboratory,” and it was found that the
sample of yellow corn meal then examined contained 5 per cent
of zein soluble in strong alcohol, 3.15 per cent of protein soluble
only in alkaline or acid solutions, and 0.45 per cent of globulins,
albumins, and proteoses. The composition and properties of these
different proteins have been extensively described by Chittenden
and Osborne.*

Owing to the very small amount of globulins and albumins, no
attempt was made to prepare any of these for hydrolysis, but the
zein and alkali soluble proteins were prepared in quantity from
freshly harvested, high-protein, white maize, which was kindly sent
to us by Professor Hopkins of the University of Illinois. We wish
to here express our high appreciation of Professor Hopkins’ kind-
ness in furnishing so large a quantity of these valuable seeds.

After the seeds were dried until they became hard, they were
ground to a fine powder in the laboratory mill and then extracted

1 The expenses of this investigation were shared by the Connecticut Agricul-
tural Experiment Station and the Carnegie Institution of Washington, D. C.
2 OSBORNE, Journal American Chemical Society, 1897, xix, p. 525.
8 CHITTENDEN and OsBORNE, American chemical journal, 1891, xiii, pp. 327
and 385; 1892, xiv, p. 20. i
477

478 Thomas B. Osborne and S. H. Clapp.

with cold 85 per cent (by volume) alcohol as long as anything was
removed. The voluminous extract was filtered perfectly clear and
concentrated under reduced pressure to a small volume. The clear
syrup that remained, after cooling, was poured in a very fine stream
into about eight volumes of distilled ice water, containing a very
little sodium chloride in order to promote the separation of the zein.
The zein, which separated as a flocculent precipitate, soon united to
a coherent plastic mass. This was filtered out on cloth, washed
superficially with water, and again dissolved in 95 per cent alcohol
to a syrup. This solution was then shaken out repeatedly with
petroleum benzine until all the yellow coloring matter was removed,
together with most of the fat.

The perfectly water-clear solution of the zein was then poured
into eight volumes of cold water, containing a very little sodium ,
chloride. The zein that separated was then dried and ground, yield- —
ing a fine powder that was entirely colorless.

The meal that had been extracted with alcohol until nothing more
could be obtained from it was then extracted with a large quantity
of 0.20 per cent sodium hydroxide solution, the extract filtered per-
fectly clear and treated with very dilute hydrochloric acid until the
dissolved protein separated as a flocculent precipitate. This was
then filtered out, again dissolved in the dilute sodium hydroxide
solution, the resulting solution filtered perfectly clear, and the pro-
tein precipitated as before. After washing with water, this precipi-
tate was digested several successive times with fresh quantities of
strong alcohol as long as any zein was removed, and when dehy-
drated with absolute alcohol and dried, formed a dusty powder
entirely free from color.

This preparation represents the residual protein of the seed after
all that is soluble in alcohol has been removed. Of its individuality
as a protein substance we of course know nothing, for it has hereto-
fore been impossible to obtain any considerable quantity of it in a
state of even approximate purity as respects admixture with non-
protein substances. This preparation contains more nitrogen than
the five other preparations that we have previously made, and prob-
ably contains but a small amount of non-protein matter. As to
whether it consists of a single protein substance or is a mixture, we
can say nothing, for we have no evidence whatever on this point.
We have, nevertheless, made this preparation for hydrolysis because
it represents a large part of the protein of the seed and it is im-

Hydrolysis of the Proteins of Maize. 479

portant to know how its products of hydrolysis compare with those
of zein.

In addition to the preparations just described we have made
similar ones from the “ gluten-meal,”’ a by-product of the glucose
works. For this meal we wish to thank Mr. H. C. Humphrey of
the Corn Products Refining Company. The material was sent to
us in cold weather in a moist state, just as it came from the filter
press. It was at once frozen solid and kept in that condition until
we dried it rapidly at a low temperature.

This meal contains about 50 per cent of protein, a large part of
which is zein. The method of preparation of the proteins from this
meal was essentially the same as that described for the corn meal.
The products obtained were used for some of the separate deter-
minations to be described later, as the quantity of the preparation
' from the corn meal was insufficient for all. Where the “ gluten ”
preparations were used special mention will be made of the fact,
although we have no doubt that the two products are entirely alike.

HYDROLYSIS OF ZEIN.

Five hundred and sixty-five grams of zein, equal to 487.2 gm.
moisture, ash, and fat free, were suspended in a mixture of 565 c.c.
of water and 565 c.c. of hydrochloric acid of specific gravity 1.19,
and heated for ten hours at 100°. The hydrolysis was then made
complete by boiling the solution for sixteen hours in a bath of oil.

A preliminary removal of glutaminic acid hydrochloride, per-
formed in the usual manner, yielded 69.80 gm. freed from ammo-.
nium chloride, or 55.92 gm. of free glutaminic acid, which is 10.9
per cent of the protein. The substance melted at about 202°—203°.

Carbon and hydrogen, 0.1607 gm. subst. gave 0.2413 gm. CO, and 0.0953 gm.

H,O.
Calculated for C;H,O,N = C 40.81; H 6.12 per cent.
Eades vos. —C 40.95; H 6.59 %

The filtrate from the glutaminic acid hydrochloride was then con-
centrated very sharply under strongly reduced pressure, and the
residue esterified with alcohol and dry hydrochloric acid gas in the
usual way.

After liberating the free esters with sodium hydroxide and potas-
sium carbonate, as described by Emil Fischer in the original hydro-

480 Thomas B. Osborne and S. H1. Clapp.

lysis of casein, and shaking out with ether, the aqueous layer was
made strongly acid with hydrochloric acid, freed from inorganic
salts, and the esterification repeated. The combined esters were
divided into the following fractions by distillation under diminished

pressure:
Temp. of bath

Fraction. up to Pressure. Weight.
I 70° 12 mm. 11.24 gm.
II 22 12: Ss 25.900"

a 100° pest Az.bas

Lo} “ “c

III; ; ak ' 6c Kee “c
94 I 44.3

d 105° O92 ak SeAyma

[2 145° O-D2y a 42-00

Iv; b By7O° 0:92, 47.00) ="

|< 186° o.92. = 7 Ose

ALOtAl, psc) tugs «| > te St en eee

The undistilled residue weighed 84 gm.

Fraction I — From this fraction no glycocoll could be brought to
separation as the hydrochloride of the ethyl ester. The fraction
consisted essentially of alanine, of which 3.13 gm. were obtained.

Fraction 11. — This fraction was saponified with boiling water,
the solution of the amino acids evaporated to dryness under re-
duced pressure, and the proline extracted with boiling absolute
alcohol. The part remaining undissolved was then examined for
glycocoll, but no crystals of the ethyl ester hydrochloride could be
obtained.*

By systematic fractionation of the free amino acids from water
and from water and alcohol, there were obtained from this frac-
tion 3.36 gm. of leucine and 5.52 gm. of alanine, while the presence
of valine could not be established. The leucine gave the following
analysis :

Carbon and hydrogen, 0.1105 gm. subst. gave 0.2224 gm. CO, and 0.0983 gm.

H,O.
Calculated for C;H,;;0,.N = C 54.96; H 9.92 per cent.
Found .. .0 ge ="@754.00. Moo ee

The alanine decomposed at about 290° and gave the following
analysis:

4 We were, furthermore, unable to detect this substance in the ether distilled
from the esters on the water bath.

Hydrolysis of the Proteins of Maize. 481

Carbon and hydrogen, 0.1496 gm. subst. gave 0.2237 gm. CO, and 0.1066 gm.

H,0O.
Calculated for C;H,O,N = C 40.45 ; H 7.86 per cent.
Bount! 4 _7.aem == 498 3). 79 216.) FE

Fraction III. — The esters of this fraction were boiled with five
volumes of water for six hours, after which time the solution had
ceased to react alkaline to litmus. The leucine that had separated
was then filtered off, and the filtrate rapidly evaporated to dryness
under strongly reduced pressure. The large quantity of proline
present was then dissolved out with boiling absolute alcohol, and
the insoluble portion dissolved in water and systematically frac-
tioned. The yield of leucine was 87.75 gm.

Carbon and hydrogen, 0.1790 gm. subst. gave 0.3610 gm. CO, and 0.1593 gm.

H,O.
Calculated for C,;H:;0.N = C 54.96; H 9.92 per cent.
Beate ee moo, EL gogo 6° >

The substance decomposed at about 298°.

In the filtrate from the leucine there were further isolated 2.04 gm.
of alanine and 1.4 gm. of valine. To isolate the latter substance it
was found necessary to have recourse to racemization. The frac-
tion containing the valine was accordingly heated in the autoclave
with excess of baryta for twenty-four hours at 175°. After remov-
ing the barium exactly with sulphuric acid, a preparation was readily
obtained, which not only gave closely agreeing results on analysis,
but which under the microscope looked perfectly homogeneous.

Carbon and hydrogen, 0.1444 gm. subst. gave 0.2721 gm. CO, and 0.1246 gm.

HO:
Calculated for C;H;,O.N = C 51.28; H 9.40 per cent.
Bounce es) ee es 2s —C 91:38 + Hig.g8

The remainder of the substance was then coupled with phenyl-
isocyanate in alkaline solution and the hydantoic acid recrystallized
from water. The substance separated in the characteristic plates,
which, heated side by side with a similar preparation from phaseo-
lin,> melted simultaneously with the latter at 161° (corr.).

Carbon and hydrogen, 0.1331 gm. subst. gave 0.2974 gm. CO, and 0.0820 gm.

H,O.
Calculated for Cj,.H;,O;N, = C 61.02; H 6.78 per cent.
Found Ce eS a & 60.94 ; H 6.84 73 3

5 OSBORNE and CLAPP: This journal, 1907, xviii, p. 301.

482 Thomas BL. Osborne and S. Hl. Clapp.

The phenyl-isocyanate derivative was then converted to the an-
hydride by dissolving in strong hydrochloric acid and concentrating
on the water bath. The prisms of the hydantoine melted on re-
crystallizing from ether and petroleum ether at 120.5° (corr.), while
the preparation from phaseolin ® melted at 122.5° (corr.), whereas
Slimmer‘ found, in the case of the synthetic a-amino-iso-valerianic
acid, 163.5° (corr.) for the phenyl-isocyanate derivative and 124°-
125° (corr.) for the hydantoine.

The proline extracts of Fractions II and III were united. The
solutions were evaporated to dryness under reduced pressure, and
the residue taken up in boiling absolute alcohol. After prolonged
standing 1.24 gm. of substance, melting at 250°, had separated, the
identity of which was not established.

The filtrate was then slightly concentrated and the proline pre-
cipitated with ether, as a slightly colored crystalline mass, which
dried in vacuum over sulphuric acid to constancy, weighed 31.99 gm.

On redissolving in absolute alcohol, the substance separated in
the characteristic prisms melting at about 203°—205°.

Carbon and hydrogen, 0.2045 gm. subst. gave 0.3889 gm. CO, and 0.1472 gm.

HO.
Calculated for C;H,O,.N = C 52.18; H 7.83 per cent.
FOUNG 50 freyt se cafe = ag -O Os WEI G.OONr esac

For the strict identification the proline was converted to the
copper salt, and the latter substance separated into the levo and
racemic by boiling with absolute alcohol. The undissolved racemic
salt separated from water in the characteristic plates.

Water, 0.1506 gm. subst. (air dry) lost 0.0164 gm. H,O at 110°.
Calculated for C1>H,,O,N2Cu * 2 HzO = H,O 10.99 per cent.
Hound , = © +: yo HSE 0.80 eee mee
Copper, 0.1294 gm. subst. ‘(area at 110°) gave 0.0351 gm. CuO.
Calculated for Ci)H:g0,N. Cu = Cu 21.81 per cent.
Found) 0) (2) Se Caron att

The amophorous copper salt of lzevo proline was converted to
the phenyl-hydantoine. The substance crystallized from water in
the characteristic prisms melting at 143°.

* OsBORNE and CLAPP: This journal, 1907, Xvili, p. 301.
7 SLIMMER, MAx D.: Berichte der deutschen chemischen Gesellschaft, 1902,
XXXV, Pp. 403.

Hydrolysis of the Proteins of Maize. 483

LVitrogen, 0.1897 gm. subst. required 2.5 c.c. 5/7 N—HCI.
Calculated for Cy.H;,0,N. = N 12.96 per cent.
HOUNEE oes eer) cet. Sy oo

Fraction Iv.— From this fraction phenylalanine was removed as
the ester in the usual way. The yield of the hydrochloride of
phenylalanine was 29.11 gm. The free phenylalanine decomposed
at about 270° and gave the following analysis:

Carbon and hydrogen, 0.1150 gm. subst. gave 0.2773 gm. CO, and 0.0699 gm.

H.O.:
Calculated for C,H,,0.N = C 65.45; H 6.66 per cent.
Rte es to hoe FO ae EL Gaga | oe

The aqueous layer was saponified by warming with an excess of
baryta on the water bath. The yield of aspartic acid isolated as the
barium salt was 5 gm.

Carbon and hydrogen, 0.2571 gm. subst. gave 0.3426 gm. CO, and 0.1252 gm.

H,0O.
Calculated for C,H;O,N = C 36.09; H 5.26 per cent.
DEVE LLLECE A ARR ret 36.34; H 5-41 ‘“ “

The filtrate from the barium aspartate was freed quantitatively
from barium, concentrated under reduced pressure, and saturated
with hydrochloric acid gas. The yield of glutaminic acid, which
separated after prolonged standing at 0°, was 20.37 gm. The free
glutaminic acid decomposed at about 201°—202° with effervescence
to a clear oil. From this fraction there were further isolated 3.94
gm. of air-dry copper aspartate, while no serine was obtained.

Copper, 0.1135 gm. subst. (air dry) gave 0.0330 gm. CuO.

Nitrogen, 0.5275 gm. subst. (air dry) required 2.7 c.c. 5/7 N—HCL.
Calculated for C,H;O,NCu 44 H,O = Cu 23.07 ; N 5.08 per cent.

rr. Boned “<\ 400% F VU Nee Ch ogres Ree: FSi

THE RESIDUE AFTER DISTILLATION.

The residue remaining after distillation of the esters weighed
84 gm. It was saponified with baryta and the glutaminic acid
separated as the hydrochloride in the usual manner. The yield
of glutaminic acid hydrochloride was 14.32 gm., which makes the

484 Thomas B. Osborne and S. Hl. Clapp.

total yield of free glutaminic acid obtained on this hydrolysis
83.71 gm., or 16.32 per cent of the protein, while by the direct
method 18.28 per cent was obtained.

GLUTAMINIC ACID.

Osborne and Gilbert * found 16.87 per cent of glutaminic acid in
zein, but as the amount of protein available for that determination
was only 32 gm. we have repeated the determination with the fol-
lowing results: 100 gm. equal to 91.6 gm. moisture, ash, and fat
free zein from the “ gluten-meal”’ was heated on a water bath for
two and a half hours with 200 c.c. of hydrochloric acid sp. gr. 1.19,
and then boiled in an oil bath for twelve hours. After concentrat-
ing the hydrolysis solution to about two thirds its original volume
it was saturated with hydrochloric acid gas and allowed to stand
for some days on ice. There were thus obtained 20.9 gm. of glu-
taminic acid hydrochloride equal to 16.74 gm. glutaminic acid, or
18.28 per cent.

TYROSINE.

Kutscher ® has stated that zein yields 10.06 per cent of tyrosine.
We have made every effort to isolate all the tyrosine possible from
our solutions, but have obtained barely one third as much as
Kutscher. Although we are convinced that tyrosine still remained
in our solutions, we have obtained no evidence that the amount that
defied separation was more than relatively small in proportion to
that which did separate.

As several unusual observations were made in the course of our
efforts to separate the last traces of tyrosine, we will give a brief
description of them.

Three hundred grams of the zein from the “ gluten meal” were
hydrolyzed with 900 gm. of sulphuric acid and 1800 gm. of water
by heating on the water bath for ten hours and then boiling in an
oil bath for ten hours longer. After removing the sulphuric acid
with an equivalent quantity of baryta and thoroughly washing the
barium sulphate, the filtrate and washings were concentrated to
about 2500 c.c. and cooled. The substance that separated, when

c

8 OSBORNE and GILBERT: This journal, 1906, xv, p. 333.
® KUTSCHER, Zeitschrift fiir physiologische Chemie, 1903, xxxviii, p. IIT.

Hydrolysis of the Proteins of Maize. 485

washed with cold water and thoroughly dried at 100°, weighed
22.8 gm., I. The filtrate and washings were concentrated to about
1200 c.c. and similarly yielded 19.88 gm. of II. On further con-
centration 12 gm. of III were obtained and 16.27 gm. of IV. The
filtrate from IV, when concentrated to a syrup, gave 27.75 gm.
of V, from which the thick mother liquor could not be very thor-
oughly sucked out.

The total yield of substances that were thus separated in a solid
state capable of removal by filtration was about one third of the
hydrolyzed zein. Small parts of the five fractions were then treated
with glacial acetic acid in which IV and V were completely soluble
at the room temperature, II and III were mostly soluble, while I
contained much that did not dissolve. The whole of Fractions I,
II, and III were therefore treated with an abundant quantity of
glacial acetic acid, the residue filtered out, washed with absolute
alcohol, and when dried at 100° weighed 8.27 gm. (crude tyro-
sine A). The acetic acid solution filtered from A was freed from
acetic acid as completely as possible by repeatedly evaporating under
diminished pressure with water and with alcohol and then subjected
to extensive fractional crystallization. The less soluble fractions,
which closely resembled nearly pure leucine, when recrystallized
gave no Millon’s reaction, whereas, although the most soluble frac-
tions gave strong Millon’s reactions, no tyrosine could be obtained
from them in characteristic crystals. All the fractions which gave
Millon’s reaction were set aside for further examination.

Fraction IV was dissolved in water, the solution treated with
bone-black and subjected to fractional crystallization. There was
obtained 9.35 gm. of the original 16.27 gm. which did not give
Millon’s reaction; but the second fraction, which was very soluble
in water, gave a Millon’s reaction, while the filtrate from this, on
concentration to a syrup and adding alcohol, gave a gummy deposit
which would not harden under alcohol and gave a strong Maillon’s
reaction.

A similar fractional recrystallization of V yielded no fractions
which did not give a strong Millon’s reaction.

The thick syrup that had been separated from V was then treated
with alcohol, and on long standing a product separated that could
be sucked out but contained so much mother liquor that it could not
be weighed. This was redissolved, its solution treated with bone-
black and concentrated, and 8.8 gm. of substance obtained which
gave a moderately strong Millon’s reaction.

486 Thomas B. Osborne and S. H1. Clapp.

The fractions thus far obtained which gave a Millon’s reaction
were united, dissolved in a considerable quantity of 5 per cent sul-
phuric acid, and treated, as long as a precipitate formed, with a
20 per cent solution of phosphotungstic acid dissolved in 5 per cent
sulphuric acid. After washing the precipitate with a dilute solution
of phosphotungsic acid in 5 per cent sulphuric acid, these two acids
were removed from the filtrate and washings with an excess of
baryta and the excess of baryta with an equivalent quantity of sul-
phuric acid.

On concentrating and cooling the solution and subjecting it to
fractional crystallization, 3.78 gm. of tyrosine were obtained in
characteristic crystals suitable for weighing. Although long-con-
tinued and persistent effort was made to bring more tyrosine to
separation, no more than traces could be obtained. Occasionally
small fractions crystallized out in which a few needles that looked
like tyrosine could be seen under the microscope, but no weighable
quantity could be separated. As before, the tyrosine accumulated,
so far as could be judged from the Millon’s reaction, in the most
soluble fractions. So extensive and thorough was this fractionation
that two fractions were obtained, weighing 0.50 and 1.10 gm.,
equal to 0.57 per cent of the zein, which, after a single recrystalliza-
tion, were found to be pure serine.

The crystals, which had the characteristic form and sweet taste
of serine, browned at about 215° and melted with effervescence to
a brownish mass at about 240°.

Carbon and hydrogen, 0.2644 gm. subst. gave 0.3317 gm. CO, and 0.1623 gm.

H,O.
Calculated for C;H;,O,;N = C 34.29 ; H 6.67 per cent.
Found),< 402.4: We2d@iga 2o% Gide, hea

Further in the filtrate from the serine 8.5 gm. substance was
obtained which crystallized in needles, had the composition of
alanine, and decomposed at about 285°.

The preparation gave a somewhat too high Eien determination.

Carbon and hydrogen, 0.2226 gm. subst. gave 0.3338 gm. CO, and 0.1614 gm.

HO:
Calculated for C,H;0,N = C 40.45; H 7.86 per cent.
Found 4....!) ea 40.89 ; H 8.05 6 6c

The crude tyrosine A, when recrystallized from water, yielded
5.47 gm. of nearly pure tyrosine. The substance in the filtrate,

flydrolysis of the Proteins of Maize. 487

which still contained some tyrosine, was added to the solution that
was treated with phosphotungstic acid, as already described. The
total quantity of tyrosine thus obtained in weighable form was
9.25 gm. As this was obtained from 274.77 gm. of moisture, ash,
and fat free zein, it is equal to 3.37 per cent.

Another determination of tyrosine was made from the same prep-
aration of zein with a similar result. A quantity weighing I00 gm.,
equal to 91.6 gm. moisture, ash, and fat free zein, from the
“ gluten-meal,’’ were heated with a mixture of three times its
weight of sulphuric acid and six times its weight of water for two
hours in a water bath and boiled for twelve hours in an oil bath.
The hydrolysis solution was then diluted with water, the sulphuric
acid removed with an equivalent quantity of baryta, and the barium
sulphate boiled out four times with water and washed very thor-
oughly. The filtrate and washings were then concentrated to about
500 c.c. under diminished pressure, with the addition of barium
carbonate, filtered from the barium carbonate, and the latter boiled
out with water and thoroughly exhausted with boiling water.

The filtrate and washings from the barium carbonate were con-
centrated to crystallization, and, after standing over night, the
product that separated was filtered out, washed with cold water,
and dried. It weighed 4.2 gm. This was dissolved in about 800 c.c.
of boiling water, the solution decolorized with a little bone-black,
the latter extracted thoroughly with boiling water, and the solution
concentrated to crystallization. After twenty-four hours 2.95 gm.
of tyrosine separated. The filtrate on further concentration yielded
0.3 gm. more, making 3.25 gm. of tyrosine, or 3.55 per cent.

ARGININE, HISTIDINE, AND LYSINE.

A quantity of the “ gluten-meal’’ weighing 100 gm. was heated
for three hours on the water bath with three times its weight of
sulphuric acid and six times its weight of water and then boiled for
twelve hours on an oil bath. The products of hydrolysis were
worked up for bases according to the method of Kossel and Patten
with the following results: |

The solution containing the histidine was made up to 500 c.c.
and nitrogen determined in it.

Nitrogen, 100 c.c. sol. required 1.26 c.c. 5/7 N—HCl = 0.0126 gm. N =
0.0630 gm. N in 500 c.c. = 0.2322 gm. histidine, or 0.24 per cent.

488 Thomas B. Osborne and S. H. Clapp.

The solution of the arginine was made up to 500 c.c. and nitro-
gen determined in it.

Nitrogen, 50 c.c. sol. required 1.61 c.c. 5/7 N—HCl=o0.0161 gm. N =
0.1610 gm. in 500 C.c. = 0.5000 gm. arginine, or, allowing for the solu-
bility of the arginine silver, 0.6180 gm., or 0.67 per cent.

The filtrate from the first silver precipitate of arginine and his-
tidine was examined carefully for lysine according to the method of
Kossel and Kutscher, but none was found.

As the results of these determinations were so low, we repeated
them, using Kossel’s later method for separating histidine from
arginine, which depends on heating the neutral or slightly acid
solution of the silver salt with barium carbonate.?°

One hundred grams of zein from the “ gluten-meal’’ were hy-
drolyzed as before, and, after removing the sulphuric acid, the solu-
tion containing histidine was made up to 500 c.c. and nitrogen
determined in it.

Nitrogen, 100 c.c. solution required 2.14 c.c. 5/7 N—HCl = 0.0214 gm. N =
0.1070 gm. N in 500 c.c. = 0.3589 gm. histidine = 0.43 per cent.

The solution containing the arginine was made up to 500 c.c.
and nitrogen determined in it.

Nitrogen, roo c.c. solution required 6.1 c.c. 5/7 N—HCl = 0.0610 gm. N =
0.305 gm. N in 500 c.c. = 0.9477 gm. arginine. Adding 0.1180 gm.
for the solubility of the arginine silver gives 1.0657 gm., or 1.16 per
cent.

These results are somewhat higher than those first obtained, but
fall decidedly below those published by Kossel and Kutscher,!!
namely, histidine 0.81, arginine 1.85 per cent.

HYDROLYSIS OF THE ALKALI-SOLUBLE PROTEIN.

The preparation used for this hydrolysis was that made from the
seeds ground in the laboratory. Unfortunately the amount of this
material made it necessary to use much less of this protein for the
hydrolysis than we would have used had our supply been greater.

10 Cf. WEISS, Zeitschrift fiir physiologische Chemie, 1907, lii, p. 107.
1! KosseEL and KuTscHeEr: Zeitschrift fiir physiologische Chemie, 1900, xxxi,
p- 165.

flydrolysis of the Proteins of Maize. 489

Two hundred and fifty grams of the protein (containing ash equal
to 1.45 per cent, moisture equal to 10.63 per cent, ether-soluble sub-
stance equal to 0.81 per cent) were suspended in a mixture of
250 c.c. of water and 250 c.c. of hydrochloric acid of specific
gravity I.19 and warmed at 100° for about five hours. The hy-
drolysis solution was then boiled in the oil bath for eighteen and a
half hours.

A preliminary removal of glutaminic acid yielded 23.66 gm. of
the free acid, or 10.87 per cent of the protein.

Carbon and hydrogen, 0.1654 gm. subst. gave 0.2484 gm. CO, and 0.0916 gm.

H,O.
Calculated for C;H,O,N = C 40.81; H 6.12 per cent.
Bagttiieees . ce ==’ 40.95 5 H 6.15 “ ‘“

The substance melted at about 202°.

The filtrate from glutaminic acid hydrochloride was then freed
from water by evaporation under reduced pressure and the residue
esterified precisely as in the case of zein. By distillation under
diminished pressure the following fractions were obtained:

Temp. of bath

Fraction. up to Pressure. Weight.
I 85° 12.00 mm. 15.62 gm.

II 95° 5.00) ** ¥69.47 7;

Ill 80° 2.60) '& By eee

IV th fo o.7oo say f-

V 200° 0.36 * 37..8Qhae
EO se hae ea ee Sl en, oe OSE

The undistilled residue weighed 39 gm.

Fraction I. — This fraction yielded 1.00 gm. of the hydrochloride
of glycocoll ethyl ester. The melting-point was 144°.

Chlorine, 0.2390 gm. subst. gave 0.2411 gm. AgCl.
Calculated for CyH;,)0. NCI = Cl 25.45 per cent.
Monn toa ts. 32 Se og gag FS)

The filtrate from the glycocoll was added to the corresponding
filtrate of Fraction II.

Fraction 11.— From this fraction there were obtained 2.6 gm.
of leucine, while the presence of neither glycocoll nor alanine could

490 Thomas B. Osborne and S. H. Clapp.

be definitely established. The fraction further contained a not
inappreciable quantity of proline which was worked up conjointly
with that from Fractions III and 1V. The leucine gave the follow-
ing analysis:

Carbon and hydrogen, 0.1304 gm. subst. gave 0.2621 gm. CO, and 0.1180 gm.

H,0.
Calculated for C,H,;;0.N = C 54.96; H 9.92 per cent.
Found | (2) 29° ae Sel ro.05

Fraction III and Iv. — From Fractions III and IV there were
further isolated 10.98 gm. of leucine.

Carbon and hydrogen, 0.1266 gm. subst. gave 0.2541 gm. CO, and 0.1167 gm.
H,0.
Calculated for CsH;;0.N = C 54.96; H 9.92 per cent.
Found: )\)o 20 2° 1054.7 45 Ee

The proline extracts from Fractions II, III, and 1V were united.
By concentrating somewhat under reduced pressure and precipitat-
ing with ether, the proline was obtained as a somewhat colored
crystalline mass, which proved to be readily soluble in alcohol.
After drying to constancy over sulphuric acid, it weighed 10.86 gm.
For identification the phenyl-hydantoine of the levo modification
was employed.

Carbon and hydrogen, 0.1504 gm. subst. gave 0.3674 gm. CO, and 0.0792 gm.

H,O.
Calculated for C;2H;,.0.N. = C 66.67 ; H 5.56 per cent.
Hounds; <<). a GG Oa Tne ee
The melting-point was 143°.
Fraction V.— From this fraction there were isolated by the usual

method 4.63 gm. of phenylalanine hydrochloride, 0.60 gm. of
aspartic acid as the barium salt, and 1.58 gm. of air-dry copper
aspartate, while no glutaminic acid hydrochloride was obtained.
The phenylalanine gave, on boiling with dilute sulphuric acid and
potassium bichromate, the characteristic odor of phenylacetaldehyde.
but the analysis indicated a considerable admixture.

Carbon and hydrogen, 0.1667 gm. subst. gave 0.3922 gm. CO; and 0.1031 gm.
H,O.
Calculated for CjH,,0.N = C 65.45 ; H 6.66 per cent.
Found." . > 0. == eaabS G7

Hydrolysis of the Proteins of Maize. 4gI
The aspartic acid reddened but did not decompose at 300°.

Carbon and hydrogen, c.0g66 gm. subst. gave 0.1288 gm. CO, and 0.0515 gm.

Eo
Calculated for C,H,0,N = C 36.09; H 5.26 per cent.
Hound). 2 in os. —=.c 36.36 5 H 5.92 Kou ek

TYROSINE.

Fifty grams of the protein, equal to 43.55 gm. moisture, ash,
and fat free, were hydrolyzed by boiling in an oil bath for twelve
hours with a mixture of 150 gm. sulphuric acid and 300 c.c. of
water. After removing the sulphuric acid with baryta the hydroly-
sis solution was concentrated until crystallization began and then
allowed to stand for twenty-four hours. The substance that had
separated was filtered out, washed with cold water, and recrystal-
lized. There were thus obtained 1.51 gm. of tyrosine in the charac-
teristic needles, equal to 3.44 per cent. By similar treatment of a
preparation from the “ gluten-meal” there were obtained from
100 gm. equal to 86.4 gm. moisture, ash, and fat free, 3.30 gm.
recrystallized tyrosine equal to 3.82 per cent.

LVitrogen, 0.2340 gm., dried at 110°, required 1.82 c.c. 5/7 N—HCI.
Calculated for C,H;,O;N = N 7.73 per cent.
omnia ee eae yg

HISTIDINE.

The solution, washings, and mother liquor, from which the tyro-
sine first described had separated, were concentrated and worked
up for bases according to the method of Kossel and Patten. The
solution of the histidine was made up to 500 c.c. and nitrogen deter-
mined in it.

Nitrogen, 100 ¢.c. solution, required 7.09 c.c. 5/7 N—HCl = 0.0709 gm. N =

0.3545 gm. N in 500 c.c. = 1.3063 gm. histidine, or 3.00 per cent.

The histidine was converted into the dichloride. The character-
istic prisms decomposed at about 230° and gave on warming a
pronounced biuret reaction.

492 Thomas B. Osborne and S. Hl. Clapp.

ARGININE.

The solution of the arginine was made up to 1000 c.c. and nitro-
gen determined in it.

Nitrogen, 50 c.c. solution, required 4.84 c.c. 5/7 N—HCl = 0.0484 gm. N =
0.968 gm. in 1000 c.c. = 3.0047 gm. arginine. Adding 0.072 gm. for
the solubility of silver arginine gives 3.0767 gm., or 7.06 per cent.

The arginine was identified as the copper nitrate double salt.

Water, 0.2893 gm. subst., air dry, lost 0.0296 gm. H,O at 110°.
Calculated for Cy,H,;0,N,Cu oo » 3 H,O = H.O g.16 per cent.
Found = s< » O20: 23) ee
Copper, 0.1822 gm. are dried at mice. gave 0.0267 gm. CuO.
Calculated for Cj,H On Cu pera = Ol 11.57 per Cenk.
Found . . : PN SUS ay

LYSINE.

The lysine was isolated as the picrate, of which there were ob-
tained 3.27 gm. equal to 1.2730 gm. of lysine, or 2.93 per cent.
The lysine picrate gave the following determination of nitrogen:

LVitrogen, 0.0980 gm. subst. gave 16.8 c.c. moist N» at 28° and 756.6 mm.
Calculated for C,H,,O.N.2 * CsH;O,;N; = N 18.67 per cent.
Found... - se NBEO i7t mmo

GLUTAMINIC ACID.

Fifty grams, equal to 43.09 gm. ash, moisture, and fat free pro-
tein, were hydrolyzed with 200 c.c. of hydrochloric acid sp. gr. 1.11
by boiling on an oil bath for fifteen hours. The solution was con-
centrated to about 75 c.c., saturated with hydrochloric acid gas,
and after standing on ice for some time the substance that had
separated was filtered out and recrystallized after decolorizing with
animal charcoal. The weight of the glutaminic acid hydrochloride
thus obtained was 6.85 gm. after deducting the ammonium chlo-
ride which it contained. This is equal to 5.48 gm. glutaminic acid,
Or 12.72 per. cent.

TRYPTOPHANE.

A qualitative test for tryptophane with glyoxylic acid gave a
strong reaction for this substance.

flydrolysis of the Proteins of Mazze. 493

AMMONIA.

One gram protein equal to 0.8711 gm. moisture, ash, and fat
free substance was hydrolyzed by boiling for eight hours with
50 c.c. hydrochloric acid, during which time the solution was finally
concentrated to about 3 c.c. The residual solution was then taken
up in about 300 c.c. water and distilled with magnesium oxide.
The ammonia that was liberated was equivalent to 1.52 cc. 5/7
N—HCI equal to 0.0152 gm. N, or 2.12 per cent of ammonia.

The results of these hydrolyses are given in the following table:

Alkali soluble Alkali soluble
Zein. Protein. Zein. Protein.
per cent. per cent. per cent. per cent.
Glycecoll. . . 0:00 0.25 penne. . . . ©5797 net isolated
MiaGaiees . . 2.23- notisolated Tyrosine. . . 3.55 3-78
Wamner) 2. .. 0.29 «notisolated Arginine . . . 1.16 7.06
Heneme . . = ° 18.60 6.22 Fistidines” = 2 0.48 3-00
2) rr 4-99 Eysie “7. Yee 2.93
Phenylalanine . 4.87 1.74 Amimnoniae) 13) 6.3.69"? ‘Tete
Aspartic acid . 1.41 0.63 Tryptophane . 0.00% present
Glutaminic acid 18.28 L272 Total’. 3. s2)2— (653 45-44

These figures show that zein, like the other alcohol soluble pro-
teins, is characterized by yielding a very small percentage of argin-
ine and histidine, no lysine, and much ammonia and proline. The
proportion of glutaminic acid is much less than that found in the
other alcohol soluble proteins, hordein and gliadin, while the pro-
portion of leucine is very much greater. Unfortunately the amount
of the alkali soluble protein which was available for hydrolysis was
too small to enable us to obtain satisfactory results for quantitative
comparison. It is interesting, however, to note that those amino
acids which are lacking in zein are all present in notable proportions
in this protein, so that the mixture of the proteins as they occur
in this seed yields all of the amino acids usually obtained from
protein substances.

12 OsBORNE and HARRIS: Journal American Chemical Society, 1903, xxv,
Pp-3 23:
18 QsBORNE and HarRIs, /é2d., 1903, xxv, p. 853.

THE HYDROLYSIS OF GLIADIN FROM RYZE

By THOMAS B. OSBORNE anp S. H. CLAPP.

[From the Laboratory of the Connecticut Agricultural Experiment Station.]

LCOHOL extracts from rye flour a protein substance which
very closely resembles gliadin obtained from wheat flour under
similar conditions. The gliadin from rye has been the subject of
extensive study in this laboratory,” and a strict comparison in respect
to composition and reactions has been made between it and the
gliadin from wheat without revealing any differences. This com-
parison has now been supplemented by a determination of the pro-
portion of the several decomposition products which the gliadin of
rye yields when hydrolyzed. The material for this hydrolysis was
obtained by extracting rye flour, ground in this laboratory, with
cold 75 per cent (by volume) alcohol, concentrating the perfectly
clear extract under reduced pressure to a syrup, and precipitating
the gliadin by pouring the solution into ice water. The precipitate
thus obtained was redissolved in 85 per cent (by volume) alcohol,
and the solution again poured into several volumes of ice water.
The gliadin that separated was then washed with distilled water,
redissolved in 85 per cent alcohol, and the clear solution precipitated
by pouring in a thin stream into a large volume of absolute alcohol.
After dehydrating by long digestion with absolute alcohol, the
gliadin was dried over sulphuric acid, ground to a fine powder, and
moisture, ash, and ether soluble matter determined in it.

Four hundred and fifteen grams of this preparation, equal to
362.67 gm. moisture, ash, and fat free, were suspended in a mixture
of 415 c.c. of water and 415 c.c. of hydrochloric acid of sp. gr. 1.19.
After warming for three hours at 100° the hydrolysis solution was
boiled in a bath of oil for eighteen and a half hours.

1 The expenses of this investigation were shared by the Connecticut Agricul-
tural Experiment Station and the Carnegie Institution of Washington, D. C.
? OSBORNE: Journal of the American Chemical Society, 1895, xvii, p. 429.
494

The Hydrolysis of Gliadin from Rye. 495

After concentrating to about two thirds of the original volume,
the solution was saturated with hydrochloric acid gas and allowed
to stand at 0° for four days.

The precipitate of glutaminic acid hydrochloride when recrystal-
lized from strong hydrochloric acid and freed from ammonium
chloride weighed 124.44 gm., equivalent to 99.68 gm. of free glu-
taminic acid, or 28.24 per cent of the protein.

The filtrate from the glutaminic acid hydrochloride was freed
from water as completely as possible by evaporating under reduced
pressure, and the residue esterified with alcohol and dry- hydro-
chloric acid gas, as often described.

The esters were liberated and shaken out with ether, and the
aqueous layer made strongly acid with hydrochloric acid, freed
from inorganic salts, and the esterification repeated according to
the usual procedure.

By distillation under diminished pressure the following fractions

were obtained:
Temp. of bath

Fraction. up to Pressure. Weight.
i 100° 18 mm. 14.61 gm.
II 70° os 47-44. ©
Ill 100° 0550. Azan. ©
A 170° 0.35.) % 54.300
IV
B 200° GAs 16.21
ROGU beet a ase «| Gite <P SOO.agneua

The undistilled residue weighed 68 gm.

Fraction I.— From this fraction and from the ether distilled from
the esters on the water bath at atmospheric pressure, there were
obtained 0.87 gm. of pure glycocoll ester hydrochloride, which is
equivalent to 0.47 gm. of glycocoll, or 0.13 per cent of the protein.
The melting-point was 144°. The remainder of Fraction I con-
sisted essentially of alanine, of which 3.6 gm. were isolated.

Carbon and hydrogen, 0.1011 gm. subst. gave 0.1502 gm. CO, and 0.0756 gm.
H,O.
Calculated for C;H,O,N = C 40.45 ; H 7.86 per cent.
Bente, «= C. 40.525 FH S.4n

Fraction II.— This fraction yielded by the customary methods
10.54 gm. of leucine and 1.08 gm. of alanine, while no valine was
obtained.

496 Thomas B. Osborne and S. 1. Clapp.

The large quantity of proline contained in Fraction II was worked
up conjointly with that from Fraction III. No glycocoll could be
brought to separation in this fraction as the ethyl ester hydrochloride.

Fraction 111.— The yield of leucine was 11.71 gm. The substance
decomposed at about 298° and gave the following analysis:

Carbon and hydrogen, 0.1755 gm. subst. gave 0.3524 gm. CO, and 0.1575 gm.

H,O.
Calculated for C;H,;0,N = C 54.96; H 9.92 per cent.
Found ~ Ae 5 = (Gs 54-76 ; H 9-97 “ ‘cc

The proline from Fractions II and III, when freed as completely
as possible from substances insoluble in absolute alcohol and dried
to constancy over sulphuric acid, weighed 34.67 gm. By redissolv-
ing in absolute alcohol the substance separated in the characteristic
prisms melting at about 206°.

Carbon and hydrogen, 0.1760 gm. subst. gave 0.3348 gm. CO, and 0.1307 gm.

H,O.
Calculated for C;H,O,.N = C 52.18; H 7.83 per cent.
Found. 27 is wee GCL See sen os

For identification the substance was converted to the copper salt
and the levo separated from the racemized with absolute alcohol
in the usual way.

The racemic copper salt crystallized from water in the character-
istic plates containing two molecules of water-of-crystallization.

Water, 0.1365 gm. subst. (air dry) lost 0.0146 gm. H,O at 110°.
Calculated for Cj)H,,0,N.Cu * 2 HzO = H,O 10.99 per cent.
Found We | Sheer = F100 1e:sot We

Copper, 0.1177 gm. subst. dried at 110°, gave 0.0318 gm. CuO.
Calculated for C1)H;,0,N2Cu = Cu 21.81 per cent.
Pound .)°.° 4.1 5 oe ete TG ee

The levo proline was converted to the characteristic phenyl-
hydantoine. The melting-point was 142°.

Carbon and hydrogen, 0.2716 gm. subst. gave 0.6604 gm. CO, and o.1401 gm.
H,O.
Calculated for C,.H,;,0.N2 = C 66.67; H 5.57 per cent.
Found 0. <5. ogee Oost Maen ee

The Hydrolysis of Gliadin from Rye. 497

Fraction IV.— The yield of phenylalanine hydrochloride from this
fraction was 11.62 gm. The free substance was employed for the
analysis.

Carbon and hydrogen, 0.1347 gm. subst. gave 0.3226 gm. CO, and 0.0833 gm.

H.O.
Calculated for C,H,;,O.N = C 65.45; H 6.66 per cent.
Hound Voea on ce a= ges: EL 6.670 Sire

There were further isolated from this fraction 0.87 gm. of pure
aspartic acid as the barium salt. The substance crystallized in the
characteristic form, and reddened but did not decompose at 300°.
Unfortunately no elementary analysis was obtained, as the prepara-
tion was lost. The filtrate from the barium aspartate was freed
from barium and examined for glutaminic acid. There were isolated
9.89 gm. of the hydrochloride. The free substance decomposed at
about 203°.

Carbon and hydrogen, 0.2687 gm. subst. gave 0.4049 gm. CO, and 0.1524 gm.

H,0.
Calculated for C;sH,O,N = C 40.81 ; H 6.12 per cent.
Hosea. </> -< = Geareg: HiG.2zq * “

In the filtrate from the glutaminic acid hydrochloride no copper
salt of aspartic acid could be obtained.

There was further isolated from Fraction IV 0.2 gm. of pure
serine. The substance browned at about 215° and decomposed at
about 238°, with effervescence, to a brownish mass.

Carbon and hydrogen, 0.1212 gm. subst. gave 0.1536 gm. CO, and 0.0790 gm.

H,O.
Calculated for C;H;,O;N = C 34.29 ; H 6.67 per cent.
Hone « aks oC 44.56; H z2gee “

TYROSINE.

Fifty grams of rye gliadin, equal to 42.98 gm. moisture, fat,
and ash free, were hydrolyzed by heating with a mixture of 150 gm.
sulphuric acid and 300 c.c. water for two and a half hours on a
water bath and boiling for twelve hours on an oil bath. After
removing the sulphuric acid with an equivalent quantity of baryta
and boiling out the barium sulphate several times with water the
solution was concentrated to crystallization. After standing over
night the substance that had separated was recrystallized from

498 Thomas B. Osborne and S. H. Clapp.

water, and 0.51 gm. of tyrosine in characteristic needles was
obtained.
LVitrogen, 0.2045 gm. subst. required 1.63 c.c. 5/7 N—HCl = 0.0163 gm. N.

Calculated for C,H,,0;N = N 7.73 per cent.
Found = Neyor

HISTIDINE, ARGININE, AND LYSINE.

Fifty grams of the rye gliadin, equal to 47.37 gm. moisture, fat,
and ash free, were hydrolyzed and the bases determined as Kossel
and Patten direct. The solution of the histidine was made up to
500 c.c. and nitrogen determined in 100 c.c. of it.

Nitrogen, too c.c, solution required 1.00 c.c. 5/7 N— HCl = 0.0100 gm.
N = 0.0500 gm. N in 500 c.c. = 0.1843 gm. histidine, or 0.39 per cent.

The solution containing the arginine was made up to 1000 c.c.
and nitrogen determined in 50 c.c of it.

Nitrogen, 50 c.c. solution required 1.50 c.c. 5/7 N—HCI = 0.0158 gm. N =

0.3160 gm. N in 1000 c.c. = 0.9809 gm. arginine. Adding 0.0720 gm.

for solubility of silver arginine gives 1.0529 gm. arginine, or 2.22 per cent.

The results of this hydrolysis are given in the following table,
and for comparison are also given those which we have obtained
with the other alcohol soluble proteins:

Gliadin, Gliadin, Hordein, Zein,
Rye Wheat Barley Maize
per cent per cent per cent per cent
Giyeocell 2... 5 “838 0.02 0.00 0.00
PAanURES 5" ici dai JF A Bees 2.00 0.43 2.23
Valine not isolated 0.21 0.13 0.29
Leucine . 6.30 5-61 5-67 18.60
Proline 9.82 7.06 13-43 6.53
Phenylalanine 2.70 2.25 5-03 4.87
Aspartic acid 0.25 0.58 not isolated 1.41
Glutaminic acid » 33:51 733 36.35 18.28
Serine - 63:06 0.13 not isolated 0.57
Tyrosine 1.19 1.20 1.67 3-55
Arginine . 2.22 3-16 2.16 1.16
Lysine 0.00 0.00 0.00 0.00
Histidine 0.39 0.61 1.28 0.43
Ammonia 5k 5-11 4.87 3-61
Tryptophane. present present present 0.00
Cystine not determined 0.45 not determined not determined
Total 64.31 65.81 G1.32 6153

The Hydrolysis of Ghadin from Rye. 499

The agreement between the analyses of the gliadin from wheat
and rye is so close that the conclusion that differences exist between
the preparations from these two seeds is not justified. Between
hordein, zein, and gliadin, however, such distinct differences exist
that, taken in connection with the differences in ultimate composition
and properties, there can be no question that these are distinctly
different proteins. These hydrolyses show that the alcohol-soluble
proteins of the cereals form a distinctly characterized group which
differ from all the other protein substances thus far analyzed.
These differences are especially shown in their high content of
proline, glutaminic acid, and ammonia, and their low content of
arginine and histidine and absence of lysine.* Zein is especially
worthy of note, as it lacks glycocoll, lysine, and tryptophane, which
are obtained from nearly all the other proteins. |

8 Cf. KosseL and KuTSCHER, Zeitschrift fiir physiologische Chemie, 1900,
xxxi, p. 165.

FURTHER DATA REGARDING THE CONDITION OF
THE VASOMOTOR NEURONS IN? “SHOCK.

BY W. T. PORTER ann W. C.“OUINBY:
[From the Laboratory of Comparative Physiology in the Harvard Medical School.]

i

1 1903 the writers pointed out that the quantitative method of

Porter and Beyer * might be used to determine the condition of the
vasomotor neurons in the symptom complex termed shock. The de-
pressor nerve is afferent to the bulbar vasomotor centre. The fall in
blood pressure produced by stimulating this nerve can be measured
with exactness and its percentile value recorded. If this percentile
value be as great, or almost as great, in shock as in the normal state,
it is certain that ie vasomotor cells concerned in the reaction are not
“ exhausted ” or “ depressed.”

In the experiments reported very briefly i in 1903 “ the normal fall
of blood pressure produced by stimuli of uniform intensity applied
to the central end of the depressor nerve was measured in the rabbit
and the cat. In the same animals shock was then brought on and the
measurements repeated.” It was found that the percentile fall in the
blood pressure obtained during shock was little, if any, less es that
obtained before shock appeared.

A fuller report has been reserved until certain studies of vasomotor
reflexes carried on in this laboratory should be, at least in part, com-
pleted. These studies confirm our former statements.

i

The methods employed in the experiments of 1903 and the results
obtained are shown in the following protocols.

1 W. T. PorTER and H. G. BEYER: This journal, 1900, iv, pp. 283-299.
500

Condition of the Vasomotor Neurons in “ Shock.” 501

Carotid blood pressure.

Operative procedures. : :
P P During

stimu- Fall.

EXPERIMENT, SEPTEMBER 15, 1903.

Rabbit tracheotomized. .Hg. -Hg. | per cent.

Left depressor nerve stimulated . . . . . 39

Stimulation of depressor nerve

Stimulation of depressor nerve

Sciatic nerves laid bare.

Boiling water to hind limbs.

Left leg thoroughly burned in Bunsen flame.
The blood pressure rose.

The other leg was burned.

Depressor stimulation

Depressor stimulation

Opened abdomen ; withdrew intestines:
burned loops with saturated zinc sulphate
and then with concentrated nitric acid.

| Depressor stimulation eee

EXPERIMENT, SE PTEMBER 16,

Spinal cord of rabbit exposed at i ane
vertebra
Stimulation of depressor nerve

Stimulation of depressor nerve .

Thoroughly painted both hind limbs with
concentrated nitric acid.

No fall in blood pressure

Section of spinai cord in lumbar region.
Stimulation of central end of cord caused
the blood pressure to rise.

Depressor stimulation

Depressor stimulation :

Injected 60 c.c. normal saline solution into
external jugular vein; without effect on
the blood pressure.

Stimulation of depressor nerve

EXPERIMENT, SEPTEMBER

Rabbit tracheotomized. Both vagicut. Blood
pressure 80

Stimulated depressor nerve.

Exposed intestines.

Ligated mesenteric artery. Applied concen-
trated nitric acid to intestine; blood pres-
sure rises.

Stimulation of mesenteric nerves at mesen-
teric artery causes a fall followed by a rise
in blood pressure.

_ Rectal temperature 26°. No anesthetic for
many hours.

Stimulation of depressor nerve '

Stimulation of depressor nerve

Stimulation of depressor nerve

Stimulation of depressor nerve

502 W. 7. Porter and W. C. Quinby.

In all these experiments the clinical signs of shock were present:
the blood pressure was very low, the temperature was subnormal, the
heart beat weak and often irregular, and the irritability of the nervous

9.15 A.M. Rectal temperature, 38°. Speed, one centimetre in 24 seconds.

FicurE ]1.— The stimulation of the depressor nerve in the rabbit causes the blood pres-
sure to fall 46 per cent (from 65 to 35 mm. Hg). Both vagi have been cut. Compare
with Fig. 2 from the same rabbit about eight hours later.

system apparently much reduced, so that the animal required no
anesthetic in spite of the most serious injuries.

The conclusion that the vasomotor cells are not exhausted in the
symptom complex termed shock is confirmed by the stimulation of

4 P.M. 4.50 P.M. 5.15 P.M.

id

ih Bi Nyy i Ai uf ig a yi pet a yin at

Rectal temperature, 26°. Speed, one centimetre in 24 seconds.

FIGURE 2.— From the same rabbit as Fig. 1, after lying about eight hours with exposed
intestines painted with nitric acid. For more than two hours the rectal temperature
had been 11° C. below normal. Stimulation of the depressor nerve lowered the
blood pressure 43, 45, and 34 per cent respectively. The last figure will be 41 per
cent if the measurement be taken from the lowest point in the curve.

the brachial, sciatic, and depressor nerves after gross injuries of the
brain.” It is further confirmed by the analysis of 765 blood-pressure
records obtained by stimulating these nerves after injuries to the

2 W. T. PorTER and T. A. SToREY: This journal, 1907, xviii, pp. 181-199.

Condition of the Vasomotor Neurons in “ Shock.” 503

brain, section of the spinal cord, exposure and mechanical injury of
the abdominal viscera, application of nitric acid and zinc sulphate to
the peritoneum, cauterization of the skin of the limbs, hemorrhage,
and section of the splanchnic nerves.?*

III.

In reflecting upon these experiments it is necessary to consider, first
of all, whether the condition produced in the experimental animal was
really shock. As tothe symptoms, or rather the signs, of shock, there
is general agreement, and it should be easy to determine how far the
condition here produced is open to criticism.

Clinicians usually state that the fall in blood pressure is the most
significant as well as the most constant symptom of shock. Opinion
as to the extent to which the blood pressure must fall to bring the
case within the category of shock can be gained by taking the average
of the observations made by a clinician who has experimented upon
this subject. In the first fifty pages of a recent treatise * are recorded
twenty-eight experiments on dogs in which the blood pressure at the
beginning of the experiment averaged 132 mm., while the blood
pressure after shock is said to be present averaged 57 mm. Hg.’ It is
evident that in our experiments the blood pressure was even lower
than that usually taken to be symptomatic of shock.

Again, it is sometimes urged that in shock the blood pressure falls
instead of rising on stimulation of afferent nerves. This abnormal
reaction was observed in several of our experiments.®

Finally, it may be objected, by those who are not well versed in
experimental physiology, that the symptoms of shock in the cat or
rabbit cannot have the diagnostic value of the identical symptoms in
man, because of the differences between man and these lower animals.
These differences are marked indeed, but they should not be made the

* W. T. PorTER: This journal, 1907, xx, pp. 399-405.

4 G. W. CRILE: Blood pressure in surgery, 1903.

5 The initial blood pressure was mentioned in nineteen instances and the blood
pressure during shock in twenty-five. _

6 Even a fall would indicate that the vasomotor cells were not exhausted, though
it would point to a disturbance of their normal equilibrium. This fall, however,
often occurs when the blood pressure is at the normal level and when signs of
shock are absent. * It can be produced by strychnine, chloral, or curare.

504 W. 7. Porter and W. C. Quinby.

basis of a hasty generalization. It is conceded that skilled move-
ments, for example, are much more highly developed in man and in
the anthropoid apes than in such animals as the rabbit and cat. But
experience suggests that the maintenance of blood pressure, like the
respiration, belongs to those fundamental functions that are singularly
alike in all the higher animals. As this point is vital to the application
of our experiments, we are glad to present the conclusions reached in
a comparative study of the blood pressure about to be published by
Dr. Porter and Mr. Richardson. The same electrical stimulus was
applied to the sciatic and the brachial nerves in the dog, cat, rabbit,
guinea pig, rat, and hen, and the rise in blood pressure recorded. It
is found that the reaction is quantitatively very similar in these widely
separated types. In the cat, rabbit, rat, and hen it is, in fact, identical.
As the difference in structure between the cat and the hen, for ex-
ample, is greater than the difference between the cat and man, it would
seem safe to conclude that the vasomotor reactions in man are essen-
tially like those in other high mammals.‘

There can indeed be no question that the experimental animals in
the several investigations reported from this laboratory exhibited the
clinical picture termed shock, and unquestionably the hundreds of
measurements we have now collected are evidence enough that
the vasomotor cells in these animals were neither exhausted nor
depressed.

Ly.

It will be noted that this paper deals with the symptoms of shock
rather than with shock itself. The distinction is important. The
symptoms of shock are a clinical entity about which there can be little
dispute; shock, on the contrary, is a pathological state, the data of
which are at present hypothetical.

The hypothesis which constitutes the hitherto generally accepted
definition of shock declares that the vasomotor cells are depressed,
exhausted, or inhibited by excessive stimulation of afferent nerves;
the fall in blood pressure and the accompanying symptoms are the
result of this depression. The experiments cited in this paper demon-
strate that the vasomotor cells are not thus depressed or inhibited, and

7 It is noteworthy that of all these animals the dog is found to be the least
adapted for vasomotor studies. : ;

Condition of the Vasomotor Neurons in “ Shock.” 505

experiments published in the last number of this journal § show that
excessive stimulation of afferent nerves does not materially lessen
the blood pressure. The present hypothetical basis of shock is thus
removed.

The thoughtful reader will hardly quarrel with this conclusion; he
will remember that there is as yet no evidence that either the respira-
tion or the temperature can long be altered by afferent impulses.

8 W. T. PorTER, H. K. Marks, and J. B. Swirt, Jr.: This journal, 1907, xx,
PP- 444-449.

ay

ae

INDEX” TO VOL. XX.

BBOTT, F. M.
ABBOTT, I.
Amandin, hydrolysis of, 470.
Animal holder, 362.
Annunciator for use in metabolism experi-
ments, 358.
AUER, J. See MELTZER and AUER, 259.
Automatism of respiratory and cardiac
mechanisms, 407.

ECHT, F. C. See Carison, GREER,
and BECHT, 180.
Blood pressure, during gastric and perito-
neal stimulation, 74.
——, in shock, §00.
, related to respiratory movements, 451.
Bone ash, in diet, 343.
Brown, E. D. See SOLLMANN, BROWN,
and WILLIAMS, 74.

ANNON, W.B. The acid control of
the pylorus, 283.

Cartson, A. J., J. R. Greer, and F. C.
Becut. The relation between the blood
supply to the submaxillary gland and the
character of the chorda and the sympa-
thetic saliva in the dog and the cat, 180.

CARLSON, A. J., and F.C. McLEAN. Fur-
ther studies on the relation of the oxygen
supply of the salivary glands to the com-
position of the saliva, 457.

Cerebral vessels, action of drugs on, 206.

Crapp, S. H. See OsporneE and CLapp,
479, 477, 494.

Colloidal solutions, osmotic pressure of, 127.

| Se as affect osmotic pres-
sure of collodial solutions, 127.

Embryo, enzymes in, 81, 97.

—, glycogen in, 117.

Enzymes, inverting of alimentary tract, 81.

5°7

See LOMBARD and ! no sometimes caused by xanthin,

439-
Fibrin, swelling, 330.
FISCHER, M. H., and G. Moore.
swelling of fibrin, 330.
Frog, muscles of thigh, r.

On the

ELATIN, nutritive value, 234.
GiEs, W. J. See STEEL and GIEs, 343,
378.

Gliadin, hydrolysis from rye, 494.

Glycogen, in embryo pig, 117.

GREER, J. R. See CARLSON, GREER, and
BECHT, 180.

GUTHRIE, C. C., and F. H. PIKe. Fur-
ther observations on the relation between
blood pressure and respiratory move-
ments, 451.

El test. affected by magnesium sul-
phate, 323.
Heart, after isolation from extrinsic nerve
impulses, 407.

EAVENWORTH,C.S. See MENDEL
and LEAVENWORTH, 117.

LEE, F. S. The action of normal fatigue
substances on muscle, 170.

LILLIE, R. S. The influence of electrolytes
and of certain other conditions on the os-
motic pressure of colloidal solutions, 127.

LOMBARD, W. P.,and F. M. ABRoTT. The
mechanical effects produced by the con-
traction of individual muscles of the thigh
of the frog, I.

NM CLEAN, F. C. See CARLSON and
MCLEAN, 457.

MACNIDER, W. DE B., and S. A. MarT-
THEWS. A further study of the action

of magnesium sulphate on the heart, 323.

508

Magnesium sulphate, action on the heart,
323.

MANDEL, A. R. Xanthin as a cause of
fever and its neutralization by salicylates,

439-
MarKS, H. K. See PORTER, MARKS, and
SWIFT, JR., 444.

MATTHEWS, S. A. See MACNIDER and
MATTHEWS, 323.

MELTZER, S. J., and J. AUER.
rush, 2509.

MENDEL, L. B., and C. S. LEAVENWORTH.
Chemical studies on growth.— III. The
occurrence of glycogen in the embryo
pig, 117.

MENDEL, L. B., and P. H. MITCHELL.
Chemical studies on growth.—I. The
inverting enzymes of the alimentary tract,
especially in the embryo, 81.

MENDEL, L. B., and P. H. MITCHELL.
Chemical studies on growth.—II. The
enzymes involved in purine metabolism
in the embryo, 97.

MeEyeErR,G.M. Animproved animal holder,
362.

MEYER, G. M.
366.

Metabolism, bone ash in experiments, 343.

, proteid, 234.

, purine, 97.

MiImciuirnht. ee EL
MITCHELL, 81, 97.

Moore,G. See FISCHER and MooRE, 330.

Morin, J. R. The nutritive value of gela-
tin. —II. Significance of glycocoll and
carbohydrate in sparing the body’s pro-
teid, 234.

Muscle, fatigue, 170.

Muscular contraction, its mechanical ef-
fects, Yr.

Peristaltic

See SALANT and MEYER,

See MENDEL and

SBORNE, T. B., and S. H. CLapp.
Hydrolysis of amandin from the
almond, 470.

OsBORNE, T. B., and S. H. CLapp. Hy-
drolysis of the proteins of maize, Zea
mays, 477.

OsBoRNE, T. B., and S. H. Capp.
hydrolysis of gliadin from rye, 494.

Osmotic pressure of colloidal solutions, 127.

The

pee ees 378.
Peristaltic rush, 259.
PIKE, F. H. See GUTHRIE and PIKE,

451.
PIKE, F. H. See STEWART and PIKE, 61.

L[ndex.

PORTER, W.T. The effect of uniform af-
ferent impulses upon the blood pressure
at different levels, 399.

PorTER, W. T., H. K. MARKs, and J. B.
SwiFr, JR. The relation of afferent
impulses to fatigue of the vasomotor
centre, 444.

PorTER, W. T., and W. C. QuinBy. Fur-
ther data regarding the condition of the
vasomotor neurons in shock, 500.

Proteins of maize, hydrolysis of, 477.

Purine metabolism, 97.

Pylorus, acid control of, 283.

UINBY, W. C. See PORTER and

QUINBY, 500.

ADIUM, elimination from normal and
nephrectomized animals, 366.
Respiration, after isolation from extrinsic
nerve impulses, 407.

—— , during gastric and peritoneal stimu-
lation, 74.

Respiratory movements affect blood pres-
sure, 451.

Respiratory nervous mechanism, resuscita-
tion of, 61.

ALANT, W., and G. M. Meyer. The
elimination of radium from normal
and nephrectomized animals, 366.

Saliva, composition varies with oxygen sup-
ply of submaxillary glands, 180, 457.

Shock, 500.

SOLLMANN, T., E. D. BRowN, and W. W.
WituiaMs. The acute effects of gastric
and peritoneal cauterization and irritation
on the blood pressure and respiration, 74.

STEEL, M., and W. J. Gigs. On the use
of bone ash with the diet, in metabolism
experiments on dogs, 343.

STEEL, M., and W. J. GiEs. On the chem-
ical nature of paranucleoprotagon, a new
product from brain, 378.

STEWART, G. N., and F. H. PIKE. Fur-
ther observations on the resuscitation of
the respiratory nervous mechanism, 61.

Stewart, G. N. Some observations on
the behavior of the automatic respira-
tory and cardiac mechanisms after com-
plete and partial isolation from extrinsic
nerve impulses, 407.

Submaxillary gland, blood supply affects.
composition of saliva, 180, 457.

SwiFT, Jr. J. B. See PorTER, MARKS,
and SWIFT, JR., 444.

Index.

ASOMOTOR centre, relation to affer-
ent impulses, 399.
Vasomotor fatigue, 444.
Vasomotor nerves of cerebral vessels,
206.
Vasomotor neurons in shock, 500.

ELKER, W.H. A simple electrical
annunciator for use in metabolism
experiments, and in connection with fil-

909

tration, distillation, and similar opera-
tions, 358.

WIGGERs, C. J. The innervation of the
cerebral vessels as indicated by the ac-
tion of drugs, 206.

WILLIAMS, W. W. See SOLLMANN,
Brown, and WILLIAMS, 74.

So as a cause of fever, 439.

BINDING SECT. MAN 4 1066

4) American Journal of Physiology

45
v.20
COPed

Biological
& Medical
Serials

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY

in Pa hee ee —e

_ ery "7 ¥ . — eee eae
A Tne ae eS — — ~

; , my * a

; ype, = — , sete ot yet eee
or wt athe : .- Porson oP iene = 8

ras renee set oe ees Sas . : *

ears wpe
a

ee ee
marley se

Prerts weep ee

TT ale ee ee

nie.

nt en Neg iN

BR Adnked

wee lower
ne my

roe Pree

o> ee

imanyce ee we
ieee <= . nas ~ 4 . r = nual. 2

Ne oe Ra ag . 7 ‘ seins Hae = em Me mae : : ey ‘ oietas vow Sree

~ wag i , : ene 4s SAB eh nar maser =e . - Ewe on reco Ms st » ' : ae He 5 ‘ mace ey LOI ee as coe.
“’ : tons pts wn : o lavainn é 3

a Rhowcbare tices

mone
. Sf rin age

a aS . ‘ a “+ ree
a 2 yen ene sw
Spee Plena s 4 “ 4 cs 7 ny a wai é we sw eee,
aks se : eS - ve Per:

ve we