FURTHER RESEARCHES ON THE CLOSURE 
OF THE CORONARY ARTERIES 
BY 
W. T. PORTER 
From 
THE JOURNAL OF EXPERIMENTAL MEDICINE 
Vol. I, No. i, 1896 FURTHER RESEARCHES OH THE CLOSURE OF THE 
CORONARY ARTERIES. 
{From the Laboratory of Physiology in the Harvard Medical School.) 
W. T. PORTER. 
The purpose of this communication is to examine the chief phe- 
nomena which follow closure of the coronary arteries, and to inquire 
into the cause of these phenomena, in the hope of clearing up, if 
possible, a confusing and not altogether necessary controversy. 
I. The Frequency with which Arrest of the Heart follows 
Closure of the Coronary Arteries. 
The closure of the coronary arteries was accomplished in my ex- 
periments by stopping the mouth of the artery in the sinus of 
Yalsalva, or by plugging the branches with emboli composed of 
lycopodium spores, or by ligation of the arteries on the surface of 
the heart. 
The first two methods were employed in twenty dogs. In each 
case the heart was promptly arrested. 
Closure by ligation was done in sixty-seven dogs. The frequency 
with which the heart was arrested is shown in the following table: 
Artekt mgated. 
SERIES I. 
SERIES II. 
SERIES III. 
SERIES IV. 
Arrest. 
No 
arrest. 
Arrest. 
No 
arrest. 
Arrest. 
No 
arrest. 
Arrest. 
No 
arrest. 
Ramus circumflexus 
4 
1 
0 
2 
0 
1 
3 
0 
Kamus descendens 
4 
4 
0 
V 
2 
16 
5 
1 
Ramus septi 
0 
2 
0 
3 
Ooronaria dextra 
2 
9 
0 
3 
stated after the consideration of the ligations as a whole 
The reason for the division of the experiments into series will be 
The summary of the four series of experiments discloses two 
facts of great interest. The first is that the frequency of arrest is in 
proportion to the size of the artery ligated. The smallest artery of 
the four is the arteria septi, the ligation of which has not caused 
Copyright, 1896, by D. Appleton and Company 2 
Further Researches on the Closure of the Coronary Arteries 
arrest; the next in size is the coronaria dextra, fourteen per cent of 
its ligations having been followed by arrest; then comes the still 
larger descendens, with twenty-eight per cent; and finally the cir- 
curnflexus, the largest artery of a 11, with sixty-four per cent. 
Summary of Ligations. 
Artbbt ligated. 
No. of cases. 
Arrest. 
No arrest. 
Percentage 
of arrests. 
Ramus circumflexus 
11 
V 
4 
64 
Ramus descendens 
39 
11 
28 
28 
Ramus septi 
5 
0 
5 
0 
Coronaria dextra 
14 
2 
12 
14 
The second fact of interest is that arrest is least frequent after 
the ligation of the arteries the 'preparation of which most injures the 
cardiac tissues. In my experience, the preparation of the circum- 
flex is made with less injury to the heart than the preparation of any 
other artery. It is done without loss of blood, without touching the 
cardiac muscle, and without any necessary injury of the ventricular 
nerves, excepting those in the artery itself, presumably of vasomotor 
nature.* The descendens is less easy to prepare, although this 
artery, like the circumflex, can usually be laid bare by merely incis- 
ing the visceral pericardium and connective tissue. The position of 
the coronaria dextra, lying in a mass of fat which is richly provided 
with blood vessels and traversed by nerves, makes its isolation much 
less easy. The ligation of the arteria septi is really a difficult opera- 
tion. This artery is never seen on the surface of the heart. It is 
the only artery of the four the preparation of which inevitably 
lacerates the cardiac muscle. Yet the ligation of this artery never 
stopped the heart, while the ligation of the circumflex, least difficult 
of all to prepare without injuring the heart, caused arrest eleven 
times in eighteen cases, j* 
Equally interesting facts are revealed by a further analysis of 
these statistics. It will be noticed that the ligations are presented 
* Heymanns and Demoor. Etude de I’innervation du coeur des vertebres k I’aide de 
la methode de Golgi. Arch, biol., 1895, xiii, pp. 619-676. W. T. Porter. The Vaso- 
motor Nerves of the Heart. The Boston Med. and Surg. Journal, 1896, p. 39. 
•j- The difficulty of reconciling these facts with the idea that the arrest of the heart 
after ligation of a coronary artery is due to mechanical injury of the heart will be insisted 
upon in the discussion of the cause of standstill. W. T. Porter 
3 
in four series. The first series was made in Berlin,* the dogs being 
given curare and, in four cases, morphine ; the second was made in 
St. Louis,f ether alone being used ; the third and fourth were done 
in Boston. In the third, one dog was anaesthetized with morphine 
and ether throughout the operation, the remainder being etherized 
until the destruction of the bulb at its junction with the spinal cord 
made further etherization superfluous. In the fourth series the 
treatment was as follows : 
Date. 
Treatment. 
Artery ligated. 
Result of ligation. 
1895. 
Jan. 17 
Bulb cut; peptone. \ 
Morphine; curare; peptone. 
Desccndens, 
No arrest. 
“ 31 
“ 
Arrest. 
Feb. 4 
u u u 
U 
U 
“ 6 
“ 
“ 
U 
“ 7 
u u 
“ 
a 
“ 12 
“ “ peptone. 
“ bulb cut; “ 
“ 
No arrest. 
“ 26 
Circumflex. 
Arrest. 
“ 28 
U ll u 
“ 
U 
Mar. 1 
U 
a 
The four series may therefore be divided into two groups. In 
the first group, comprising Series I and IY, morphine or curare or 
both were employed ; in the second group, comprising Series II and 
111, neither of these agents was used, with the exception of a single 
case in Series 111, the animals being narcotized with ether or the 
brain cut off by section of the bulb. The condition of the heart 
evidently can not have been the same in the two groups, for morphine 
and curare do not affect the heart as ether or bulbar section do. 
Compare now the results obligation in the two groups. It will be 
seen that the results are not alike. 
The number of ligations of the circumflex, arteria septi, and right 
coronary arteries in Group II is so small that I will not use them for 
comparison with those of Group 111 further than to point out that 
they make it probable that the condition of the heart is a factor of im- 
*On the Results of Ligation of the Coronary Arteries. The Journal of Physiology, 
1894, xv, pp. 121-138. 
f See Ueber die Frage eines Coordinations-centrums im Herzventrikel. Arch. f. d. ges. 
Physiol., 1894, Iv, pp. 366-371. 
\ The peptone was employed to prevent the coagulation of the blood during the meas- 
urement of the blood discharged from the left ventricle before and after the ligation of a 
coronary artery. 4 
Further Researches on the Closure of the Coronary Arteries 
portance in the results of ligation. This probability is made a cer- 
tainty by the comparison of the arrests from ligation of the descen- 
dens, the arrests being sixty-four per cent where morphine and curare 
are used, and only eight per cent where they are not used. The 
frequency of arrest is greatly increased hy morphine and curare. 
Group I, 
Comprising Series I and IV. 
Group II, 
Comprising Series II and III. 
Morphine, curare, or both ; sometimes peptone. 
Ether or section of bulb. 
Per 
Per 
No. of 
Ar- 
No 
cent 
No. of 
Ar- 
No 
cent 
cases. 
rest. 
arrest. 
ar- 
cases. 
rest. 
arrest. 
ar- 
rested. 
rested. 
8 
7 
1 
88 
Circumflex 
3 
0 
3 
0 
Descendens.... 
14 
9 
5 
64 
Descendens... . 
25 
2 
23 
8 
Septi 
2 
0 
2 
0 
Septi 
3 
0 
3 
0 
Coronaria dextra 
11 
2 
y 
14 
Coronaria dextra 
3 
0 
3 
0 
11. The Effect of Closure of the Coronary Arteries on the 
Blood Pressure and on the Force and Frequency of the 
Heart Beat. 
In my paper On the Results of Ligation of the Coronary Ar- 
teries the conclusions were reached that “ a gradual and continuous 
decrease in the height of the intraventricular upstroke appears 
almost immediately after ligation. When ligation does not cause 
standstill this lowering of the upstroke is absent or transient. 
When standstill follows ligation the diastolic intraventricular pres- 
sure is invariably increased.” 
Further experience has supported these statements. 
The diastolic rise of pressure is particularly important. The 
normal mean pressure in the auricles and in the pulmonary veins is 
known to be very low.* It is probable, therefore, that the pressure 
in the coronary veins near their mouths is also low. A rise of a few 
millimetres in the auricular pressure would in that case interrupt 
the entire coronary circulation, unless overcome by a compensatory 
rise of pressure in the coronary arteries. This rise in the auricular 
pressure is certainly present after the ligations which are followed 
by arrest, for the rise of diastolic pressure which has been shown to 
*W. T. Porter. The Journal of Physiology, 1892, xiii, pp. 513-553. W. T. Porter 
take place in the ventricle must be accompanied by a similar rise in 
the auricle, owing to the free communication between the two 
chambers during the greater part of diastole. The rise in auricular 
pressure after ligation is, moreover, not compensated by increased 
pressure in the coronary arteries. On the contrary, the pressure in 
the arteries is falling while that in the auricles is rising. 
It must be acknowledged, then, that a rising auricular pressure 
after ligation may at length put a stop to the whole blood supply of 
the cardiac muscle, and, as this rise is often occasioned by the closure 
of a single vessel, it is plain that the entire coronary circulation can 
in fact be interrupted by the ligation of one coronary artery. 
This possibly explains the fact that both ventricles, if arrested at 
all, usually stop at the same instant, no matter which ventricle the 
ligated artery supplied. But in our present ignorance of the nature 
of the cardiac co ordinating mechanism it would be hazardous to 
insist on this explanation. 
The frequency of the ventricular beat is seldom changed when 
ligation is not followed by standstill. When, however, ligation is 
followed by standstill, the frequency is usually sooner or later altered. 
But some hearts beat with unchanged rhythm to the last. In hearts 
isolated by a modification of Ludwig’s method, to be described in 
the next section, the frequency after ligation is less regular, the ex- 
planation of the difference doubtless being that the hearts thus iso- 
lated are placed under very unusual conditions. 
111. The Effect of Closure of the Coronary Arteries on the 
Quantity of Blood discharged by the Left Ventricle. 
The method employed to measure the outflow from the heart was 
in the main that used by Pawlow *in Ludwig’s laboratory. The 
differences were the exposure of the heart, in order to reach the 
coronary arteries ; the closure of the aorta by ligation above the root 
of the left lung instead of closure by distending a bag introduced 
into the aorta through the subclavian artery; and the omission of 
the floats in the measuring cylinders, the change of current which 
* Archiv fur Physiologic, 1887, pp. 452-468. 6 
Further Researches on the Closure of the Coronary A rteries 
they would have produced by making a contact at the top of the 
cylinders being secured in ray experiments by turning a double 
Fig. I.—Showing fall in arterial pressure and diminished output of left ventricle in consequence of the ligation of the circum- 
flex artery. The curve reads from left to right. It is one half the original size. The upper curve is the pressure in the 
carotid artery. The unbroken line is atmospheric pressure. The next curve is the measurement of the outflow from the 
left ventricle, each rise and each fall indicating the passage of 50 c. c. of blood into the aorta. The lower line is a time 
curve in seconds. At * the circumflex artery was ligated. 
key ( when the blood 
rose to a certain mark on 
the cylinder. This method 
of turning the current into 
first one and then the other 
of the magnets which con- 
trol the inflow and outflow 
from the measuring cjdinders 
was found to be sufficiently 
exact for the business in 
hand. The turns of the 
double key were recorded 
electrically on the smoked 
paper. The quantity of 
blood thrown out by the left 
ventricle was not essentially 
changed by ligations that 
were not followed by arrest. 
On the other hand, the quan- 
tity of blood thrown out by 
the left ventricle steadily 
diminished after the liga- 
tions that were followed by 
arrest. (Fig. 1.) The details 
of the experiments are shown 
in the tabular view on page 7. 
The anaesthetics and pep- 
tone employed in the experi- 
ments are given on page 3, 
in the account of the fourth 
series of ligations. 
The results of these experiments are in accord with the measure- 
ments of intracardiac pressure already mentioned. The fall of 
pressure in systole and the rise in diastole are accompanied by a W. T. Porter 
7 
The Number of Cubic Centimetres of Blood thrown out by the Left Ventricle per Fifty 
Seconds before and after the Ligation of a Coronary Artery. 
Date. 
Jan. 
17. 
Feb. 12. 
Jan. 
31. 
Feb. 6. 
Feb. 26. 
Feb. 28. 
March 1. 
Seconds. 
C. cm. 
C. cm. 
Pulse. 
0. cm. 
C. cm. 
Pulse. 
C. cm. 
Pulse. 
C. cm. 
Pulse. 
C. cm. 
Pulse. 
1- 50.. . . 
38*7 
482 
87 
392 
405 
93 
720 
130 
725 
43 
315 
83 
51-100.... 
387 
477 
87 
333 
468 
97 
675 
105 
702 
47 
293 
80 
101-150... . 
342 
432 
293 
446 
98 
720 
93 
743 
45 
315 
81 
151-200... . 
333 
392 
87 
293 
423 
96 
765 
96 
720 
45 
315 
82 
Ligation of 
circumflex. 
201-250... . 
342 
378 
87 
329 
455 
93 
765 
94 
720 
46 
297 
76 
Ligat. 
Ligation of 
Lifjat. 
Ligation of 
Ligation of 
des’n. 
descendens. 
des’n. 
circumflex. 
circumflex. 
251-300... . 
351 
482 
87 
446 
441 
95 
648 
126 
608 
37 
270 
78 
301-360 
324 
468 
82 
243 
441 
96 
558 
123 
486 
36 
234 
75 
Arrest 7 
sec. later. 
351-400... . 
320 
441 
84 
180 
423 
94 
419 
36 
221 
76 
Ar- 
Ligation of 
Arrest 20 
rest. 
descendens. 
sec. later.* 
401-450.. . . 
329 
405 
482 
99 
167 
67 
Arrest 35 
sec. 
later. |1 
451-500.... 
315 
482f 
83 
405 
94 
501-550... . 
324* 
387 
91 
Arrest 30 
sec. later. J 
diminished outflow from the left ventricle, while the frequency of 
beat is usually little changed. These changes in the heart beat, to- 
gether with the slowing of the systole demonstrated in my first paper, 
are the fundamental alterations in the activity of the heart which 
usher in the loss of co-ordinating power and final arrest of the heart 
after the ligation of the coronary arteries. 
We must now inquire the cause of these phenomena. Two ex- 
planations have been offered. The changes in the heart beat are 
produced either by the sudden arrest of the nutrition of a sufficiently 
large part of the heart or by the mechanical injury done the heart 
in the operation of ligation. I shall show that the closure of a coro- 
nary artery puts an end to the nutrition of the area which it sup- 
* The experiment of January 17th continued 2,190 seconds longer. Ligation was not 
followed by arrest. 
f The record of outflow continued until the fourteen hundredth second. No important 
change could be perceived. The heart beat, in all, 2,900 seconds, and was then stopped by 
the clotting of the blood in the measuring cylinders. 
\ During these 30 seconds 225 cubic centimetres were thrown out. 
# During these 20 seconds 135 cubic centimetres were thrown out. 
| During these 35 seconds 113 cubic centimetres were thrown out. 8 
further Researches on the Closure of the Coronary Arteries 
plied and that the phenomena which follow the ligation of the coro- 
nary arteries are not due to the mechanical injury of the heart. 
IV. The Ligation of a Coronary Artery puts an End to the 
Nutrition of the Area which it supplied. 
That the closure of a coronary artery puts an end to the nutrition 
of the area which it supplied is an accepted truth in pathology, so 
well established by observation and direct experiment that it would 
be unnecessary to mention it here had not some recent writers neg- 
lected the obvious bearing of this fact on the correctness of tlieir 
own hypotheses. The present state of pathological opinion regard- 
ing this matter is assuredly fairly expressed in such works as Zieg- 
ler’s Allgemeine Pathologie, eighth edition, 1895. I find there the 
following statements: “ Anaemic infarcts occur principally in the 
spleen, heart, kidneys, and retina” (page 175). “ Anaemic infarcts 
are termed when the vascular area closed possesses no communications 
through which blood can be received; the area then remains blood- 
less and dies” (page 173), These words certainly represent the gen- 
eral beliefof pathologists regarding the anaemic infarcts of the heart. 
Haemorrhagic infarcts are also sometimes, though very rarely, 
present in the heart. The infarcted area is then infiltrated with 
blood. Investigators differ as to the source of this blood, some be- 
lieving it to be derived from neighbouring capillaries, others through 
back flow from the veins. But investigators are agreed that in no 
case is the circulation re-established. On this point there is no dis- 
pute. The blood which slowly works in from neighbouring vessels 
“ stagnates” * and can not preserve the life of the part. 
The present unanimity of opinion as to the entire interruption 
of the circulation in the area of infarction rests, it must be remem- 
bered, not merely on post-mortem observations, but also on the con- 
sequences of the experimental closure of terminal arteries in the 
living animal. So also is the speedy death of the infarcted area a 
matter of direct observation. The experimental closure of terminal 
arteries in the living animal for the purpose of studying the infarcts 
* Ziegler, loc. citp. 145. W. T. Porter 
9 
formed has until very recently been done on organs other than the 
heart—the kidney, for example. The experimental production of in- 
farcts in the heart has been attempted, so far as I am aware, only by 
Kolster and myself, in each case with complete success. 
Kolster * tied a small branch of the ramus descendens in the dog 
and kept the animal alive from one day to seventeen months. The 
microscopical examination of the hearts of dogs killed twenty-four 
hours after ligation showed a “ typical coagulation necrosis in the 
area of the artery ligated.” Kolster remarks that “ nothing else 
could have been expected, for Weigert (Virchow’s Archiv, Bd. Ixxix) 
took these very cardiac infarcts as a type of coagulation necrosis. 
All the alterations which Weigert gives as characteristic of the ne- 
crosis established by him as the consequence of cutting oft'the blood 
supply are to be found here” (page 24). In dogs kept longer alive 
the progressive alterations were traced from day to day and from 
month to month. 
My own experiments f were made on the main coronary trunks, 
the descendens, circumflex, or arteria septi, and in some cases more 
than one of these arteries, being ligated close to their origin, and the 
animals allowed to live. Two dogs in which the descendens was 
ligated a few millimetres from its origin lived respectively four days 
and fourteen days and a half. Characteristic anaemic infarcts were 
found occupying the anterior part of the septum and that part of 
the anterior wTall of the left ventricle which adjoins the interven- 
tricular furrow. The infarcted areas were examined macroscopic- 
ally, thin sections being made with a sharp knife.:}; The area in- 
farcted was completely degenerated, except close to its borders. 
These experiments show (1) that the arteries ligated are terminal 
arteries;* (2) that the rapid closure of a coronary artery is followed 
Arch. f. d. ges. Physiol., 1894, ly, p. 866-3'71. 
* SJcand. Arch,, f. Physiol., 1893, iv, pp. 1-43. 
£ The well-known descriptions of the microscopic appearance of infarcts in the human 
heart and the very minute description of infarcts in the dog’s heart, published by Kolster 
a few months before my work, seemed to make a microscopical examination superfluous. 
# I do not speak here of the immediate object of the experiments published in the 
Archiv f. d. ges. Physiologic—namely, the demonstration of the absence in the infarcted 
area of any nerve centre essential to the co-ordination of the contraction of the heart. 10 
Further Researches on the Closure of the Coronary Arteries 
by the death of the part which it supplies; and (3) that each part of 
the heart wall receives its blood supply essentially from one main 
artery; for example, the branches of the circumflex do not contrib- 
ute any supply of importance to the area embraced by the branches 
of the descendens. This last fact is in accord with Cohnheim’s 
statement that each coronary artery in the dog’s heart keeps to its 
owfi boundaries, and does not, as a rule, pass into the field of another 
artery. Kolster’s observation that the infarcted area contained spots 
of normal heart tissue depends probably on his having tied a small 
branch and not a main trunk. 
The attempts made by Michael is* and others to overthrow the 
great mass of pathological observations and the unequivocal testi- 
mony of infarcts experimentally produced will hardly commend 
themselves. Michaelis concludes that the coronary arteries are not 
terminal arteries, because a skilful injector (Wickersheimer) has 
succeeded in making an injection liquid pass from one artery, 
through communicating branches, to another, Ho one, so far as I 
know, has ever denied that one coronary artery could be injected 
from another. Cohnheim himself makes special mention of the 
passage of liquids through the communicating branches under 
pressure.f Every one agrees that the coronary arteries anastomose. 
It is not the absence, but the character of the anastomosis that is the 
basis of the present pathological teaching. The incontestable fact is 
that the anastomosis is too slight to permit a collateral circulation 
sufficient to keep a vascular area alive after the closure of the artery 
which supplies it. 
The argument of those who rely on injections to prove the ab- 
sence of terminal arteries appears to premise that one artery can be 
injected from another only when arterial communications exist, and 
that terminal arteries do not have arterial communications. The 
first premise is incorrect. Thin injections can be forced through 
any blood vessels. The second premise is inexact. Terminal arteries 
may possess arterial anastomoses, provided the resistance in the com- 
municating vessels is high. The idea of terminal arteries is physio- 
* Zeitschrift fur klinische Medicin, 1894, xxiv, pp. 2'70-294. 
f Virchow’s Archiv, 1881, Ixxxv, p. 509. W. T. Porter 
11 
logical, not anatomical. Terminal arteries differ from other arteries 
in that the peripheral resistance in the anastomosing vessels is too 
high to be overcome by the normal blood pressure in any of the 
arteries of which the communicating vessels are branches. Hence 
the rapid closure of any terminal artery cuts off the nutrition of its 
own capillary area because sufficient blood for the life of the area 
can not be sent through the communicating vessels on account of the 
high resistance in them. The resistance in the communicating ves- 
sels, and not their size, is the factor of first importance. The circum- 
stance on which Michaelis lays so much stress—namely, that one 
coronary artery can be injected from another—has long been known, 
and has no force against the pathological and experimental evidence 
of the terminal nature of coronary arteries. 
I conclude, then, that the rapid closure of a coronary artery puts 
an end to the nutrition of the area which it supplied. 
If now it can be proved that the changes in the heart beat pro- 
duced by the ligation of a coronary artery are not the result of the 
mechanical injuries of the operation, it will follow that these changes 
are the consequence of a sudden interruption of the nutrition of the 
heart. 
I shall determine, first, the consequences of extensive mechanical 
injuries of the tissues surrounding the coronary arteries ; second, 
the effects of the mechanical injury done the heart in preparing a 
coronary artery for ligation ; and last, the results which follow the 
closure of the coronary arteries without mechanical injury. 
Y. The Consequences of Extensive Mechanical Injuries of the 
Tissues surrounding the Coronary Arteries. 
The experiments* about to be described made on dogs 
poisoned with curare. I have already shown that curare increases 
the frequency of arrest after ligation (page 4). Notwithstanding 
the unfavourable influence of curare, arrest took place but once in 
ten cases of sudden and violent crushing of the periarterial tissues. 
The method of operating will be seen in the following example: 
* These experiments were done in the University of Berlin in 1892, 12 
Further Researches on the Closure of the Coronary Arteries 
Experiment July 7, 1892.—Dog; 8,100 grammes; curare; artificial respi- 
ration. The descendens was quickly and carefully freed from its bed at a 
point thirteen millimetres from the aortic origin of the left coronary artery. 
Only the visceral pericardium and the connective tissue over the artery 
were injured. A strongly curved, threaded needle was now passed into the 
cardiac muscle, entering on one side of the interventricular furrow opposite 
the place where the artery lay bare and emerging on the opposite side. The 
distance between the jaoint of entry and of exit was about twelve milli- 
metres. The needle, thanks to its curve, passed through the cardiac muscle 
three millimetres beneath the artery. The thread was now brought back 
underneath the artery. Thus the periarterial tissues were included in the 
ligature, but the artery was not included. This ligature was drawn tight, 
crushing the tissues embraced in it and dragging, doubtless, on many ven- 
tricular nerves. The intraventricular pulse, recorded by a Gad-Cowl ma- 
nometer, was counted during four hundred and sixty-five seconds. Before 
the ligation, the frequency per ten seconds was twenty-three; after the liga- 
tion, the frequency varied from twenty-one to twenty per ten seconds. There 
was no arrest. 
The number of experiments after this method was ten. In five 
the tissues about the descendens alone were crushed, in three those 
about the circumflex alone, and in two those about both arteries. In 
nine of the ten animals the heart continued to beat; slight irregu- 
larity is noted in three, and in one an increase both in force and 
frequency. The following case of double crushing is especially 
noteworthy: 
Experiment July 27,1892.—Prepared periarterial tissues around the cir- 
cumflexus about six millimetres from the aortic origin of the left coronary 
artery and about the descendens on a line with the anterior edge of the pul 
monary artery. Drew both ligatures tight. No great irregularity. 
In the case in which arrest followed the crushing, the periarterial 
tissues about both circumflex and descendens were prepared. As in 
all experiments, the tissues were included for at least ten millimetres, 
the thread at its deepest point being about three millimetres below 
the surface. On drawing the circumflex periarterial ligature tight, 
the heart fell into fibrillary contractions. 
The result of this series of deliberate, extensive injuries to the 
cardiac muscle and nerves lying about the coronary arteries is most 
instructive. It shows clearly that too much has been made of the 
supposed effects of such injuries on the heart, and that they by no 
means explain the consequences of ligation of the coronary arteries. 
This conclusion will become more and more probable as we proceed. W. T. Porter 
13 
YI. Mechanical Injury of the Heart in preparing the 
Artery for Ligation. 
The mechanical injuries said to be the cause of the phenomena 
which often follow the closure of a coronary artery are produced by 
the preparation of the artery for ligation. They can hardly be pro- 
duced by the injury to the artery itself in the act of ligation, for not 
even the most ardent advocate of mechanical injury would, I presume, 
venture to claim that the crushing of the arterial wall alone seriously 
affects the heart beat. The extent of the injuries produced by the 
preparation of the artery have surely been greatly exaggerated. The 
coronary arteries, excepting the arteria septi, can be prepared with- 
out the injury of a single muscle fibre, or even in many cases the loss 
of a single drop of blood. The superficial nerves which are some- 
times pulled upon or torn during the operation are, so far as is 
known, centripetal nerves, the section and direct stimulation of 
which does not cause the effects the explanation of which we seek.* 
The statistics that are available do not show that the preparation 
of the artery has any serious consequence whatever. It will be re- 
membered that Fenoglio and Drogoul f clamped or ligated a coro- 
nary artery in fifty dogs without once seeing the phenomena that 
precede arrest. The notes of nearly one hundred ligations made by 
me show that the only effect of the preparation of an artery is the 
occasional dropping of a beat. The changes in intracardiac pressure 
and the engorgement of the heart which form the unvarying intro- 
duction to the arrest that often follows ligation have never been 
observed. Several operations were made during the continuous 
recording of the blood pressure in the carotid artery by a Hurthle 
manometer. The number of dropped beats and the state of the 
blood pressure were thus accurately determined. One of the curves 
is here reproduced, (Fig. 2.) The records thus secured confirmed 
the observations with the unaided eye. An example of these ex- 
periments will be given. 
* Wooldridge, L, Archiv fur Physiol., 1883, pp. 522-541. McWilliams, Journal of 
Physiology, 1887, viii, p. 298. 
f Archives italiennes de biologic, 1888, ix, p. 49. 14 
Further Researches on the Closure of the Coronary Arteries 
Experiment September 20, 1895.—D0g-; ether; tracheotomy; division of 
bulb near spinal cord; artificial respiration; continuous curve of carotid 
pulse and pressure. Prepared descendens for ligation thirteen millimetres 
from the aortic origin of the left coronary artery. Occasionally a dropped 
beat during the preparation; nothing more. After seventy-three seconds 
the ligature was drawn tight. Speedy arrest of the heart with tumultu- 
ous fibrillary contractions and strong auricular pumping. 
In addition to the facts just mentioned, the reader will remember 
those demonstrated in the first section of this paper—namely, that 
the frequency of arrest is in proportion to the size of the artery 
ligated, and is least frequent after the ligation of arteries the prepa- 
ration of which most injures the cardiac tissues. These observations 
can not be reconciled with the mechanical-injury explanation. 
I am therefore able to conclude that the necessary mechanical 
injury of the heart in the preparation of an artery for ligation may 
Fig. 2.—Showing the character of the irregularity sometimes occurring during the prepara- 
tion of a coronary artery for ligation. The curve reads from left to right, Upper 
line is carotid pressure and pulse curve. Lower line is atmospheric pressure and time 
in seconds. At the preparation of the descendens for ligation was begun. 
cause a slight irregularity in the ventricular stroke, but that this 
irregularity is almost invariably transient and unimportant. 
YII. The Closure of the Coronary Arteries without 
Mechanical Injury.* 
It has now been shown that extensive sudden injuries of the tis- 
sues lying about the coronary arteries rarely arrest the heart, or do 
* The first mention of my method for this purpose was made in the Centralhlatt f. 
Physiol, 1895, No. 16, p. 481. See also the reply to R. Tigerstedt in a later number of the 
same year. W. T. Porter 
15 
more than cause a passing irregularity. It has been shown also that 
in nearly one hundred consecutive ligations of the coronary arteries 
not one ease was found in which the preparation of the artery for 
ligation materially affected the beat of the heart. The methods thus 
far employed can not, however, wholly exclude the influence of 
mechanical injury, because ligation can not be accomplished with- 
out such injury. It is possible, nevertheless, to close a coronary 
artery without the slightest injury either of the artery itself or of 
the surrounding tissues. This is done by passing a glass rod through 
the aorta into the aortic opening of the coronary artery, or by plug- 
ging the branches of the coronary arteries by means of emboli com- 
posed ot lycopodium spores. The flrst method is carried out as 
follows; 
The dog is etherized and tracheotomized, the bulb destroyed 
near its junction with the spinal cord, and artificial respiration 
begun. The left first rib is exposed near the sternum, and a small 
opening made into the chest in the angle formed by the rib and 
the sterno-cleido-mastoid muscle. Through this opening, the left 
internal mammary artery is easily secured. The first five ribs on 
the right side and the first three on the left side are now exposed 
hJ a straight incision along the margins of the sternum. Stout 
ligatures are passed with a large bent aneurism hook around the 
eight exposed ribs some distance from the sternum, and around the 
left fourth and fifth and the right first, second, and third ribs near 
the sternum. Another strong ligature is tied around the sternum 
between the third and fourth pairs of ribs. The ligated ribs and the 
upper part of the sternum may now be removed. It is best to begin 
with the fifth rib on the left side, cutting along the outer line of 
ligatures, raising the ribs as soon as possible, and securing the right 
internal mammary artery by reaching underneath the sternum. The 
pericardium should be slit a little to the right of the left phrenic 
nerve, and sewed to the margins of the chest opening with three or 
four stitches. The innominate artery is found beneath the inner 
border of the innominate vein, ligated as near the neck as possible, 
a rubber band placed under it near the cardiac side of the ligature, a 
compression forceps applied about fifteen millimetres below the 16 
Further Researches on the Closure of the Coronary Arteries 
ligature, and the previously oiled glass rod introduced into the artery 
through as small a hole as possible. The artery should now be 
compressed about the rod by tying the rubber 
band, which should press tightly enough to pre- 
vent bleeding along the rod, but not so tightly as 
to hinder the operator in moving the rod. The 
rod is passed down toward the left anterior sinus 
of Yalsalva. As soon as the resistance of the 
aortic valve is felt, the rod is withdrawn a little 
and its head made to traverse the aortic wall near 
the bottom of the sinus. The opening of the left 
coronary artery will presently be found. Into this 
the rod is passed and fastened in place by tying a 
Fig. 3.—Lower por- 
tion of glass rod 
(actual size) used 
to plug the left 
coronary artery. 
ligature around the innominate artery, including the rod. A finger 
of the left hand placed very gently over the left coronary artery will 
readily determine whether the rod is in the right place. The rod is 
210 millimetres long, the head and neck being of the size and shape 
represented in the accompanying figure. (Fig. 3.) The position of 
the rod was always carefully examined post mortem by the method 
to be described in the account of the following experiment ot 
September 20, 1895. 
Plugging the left coronary artery during life is not a very diffi- 
cult operation. I have attempted it in nineteen dogs, each time 
with complete success. The closure of the artery was always 
promptly followed by arrest. Fibrillary contractions were present 
in every case but one. 
It seems hardly possible that the gentle introduction of the glass 
rod into the artery could in any way injure the artery, and, in view 
of the facts presented in former- sections of this paper, it is wholly 
improbable that any slight injury which might be thus inflicted 
would stop the heart. Nevertheless, even this remote possibility can 
be excluded. 
Testimony against injury by the rod is furnished by the experi- 
ment of September 20: W. T. Porter 
Experiment September 20, 1895.—D0g; 9,500 grammes; ether; trache- 
otomy; artificial respiration; section of bulb at its junction with the spinal 
cord. 
The heart being exposed, the glass rod was passed through the innomi- 
nate artery into the supposed trunk of the left coronary artery. The head 
of the rod was distinctly felt by a finger placed over the circumflex artery. 
The handle of the rod was now tied in place. The expected standstill did 
not follow, nor could any irregularity in the heart beat be detected. After 
waiting some time for any sign of heart failure, and perceiving none, the 
descendens was laid bare. The descendens pulsated. It was now ligated 
thirteen millimetres from the aorta. Speedy standstill followed, with 
tumultuous fibrillary contractions and strong auricular pumping. The 
heart was greatly distended. 
Autopsy: Find head of rod in the circumflex artery, nine millimetres 
from the aorta. Open circumflex just beyond the head of the rod. No blood 
escapes. Raise the pressure in the aorta. Still no blood escapes; closure 
perfect. Open descendens on cardiac side of ligature. Blood immediately 
runs out in a stream. Outflow hastened hy raising the pressure in the 
aorta. Descendens not closed by rod. Find that each artery has a sepa- 
rate origin, the circumflex opening being situated between the middle and 
left end of the sinus of Valsalva, the descendens near the right end. The 
circumflex was tightly plugged, the descendens not at all. 
In this experiment, then, the plugging of the circumflex artery 
caused no disturbance of the heart beat. The operation of plugging 
was precisely the same in this case as in others in which closure was 
followed by arrest with fibrillary contractions. The constant arrest 
after complete closure and the nearly constant fibrillary contractions 
can not therefore be explained by the mechanical injury done by 
the rod. 
The next experiment is even more instructive. 
Experiment September 23, 1895.—Dog ; 11,000 grammes; ether; trache- 
otomy; section of bulb near spinal cord; artificial respiration. The search 
for the opening of the left coronary artery lasted forty-four seconds. Not 
a heart beat fell out. Immediately after closure of the artery, the crural 
blood pressure began to sink, the individual heart beats grew weaker and 
weaker, and the heart began to swell. At the sixty-fifth second after 
closure, the blood pressure had fallen from 60 millimetres, Hg., to 25 milli- 
metres. The rod was now drawn back into the aorta. The pressure con- 
tinued to fall a few millimetres during twenty-two seconds, then rose 
gradually, reached its former height in fifty-five seconds, and then rose 
farther to 85 millimetres, after which a slight fall occurred. As the blood 
pressure rose, the individual beats increased in strength and the heart re- 
turned to its normal size. (Fig, 4.) 18 
Further Researches on the Closure of the Coronary Arteries 
Fig. 4.—lllustrating experiment of September 23d. The curve reads from left to right. It is one half the original size. The upper line is 
the pressure in the crural artery. The lower line marks atmospheric pressure and time in seconds. At * the left coronary artery was 
plugged. At f the plug was withdrawn. 
The artery was now closed again. The re- 
sult was quite as before. The blood pressure 
sank, the heart swelled, and the beats were 
small. After forty-six seconds, the rod was 
once more withdrawn. All the threatening 
symptoms disappeared. In fifty-six seconds 
the heart was working as well as before. 
A third and a fourth time was this repeated. 
The fourth time the artery remained closed 
fifty seconds. The heart swelled enormously. 
Its beats were scarcely perceptible in the curve. 
During 127 seconds, it went from had to worse. 
Finally, it remained seven seconds without a 
beat. Strong massage was then employed, the 
heart being cpmpressed six times in seven sec- 
onds. The beat returned. The massage was 
twice repeated, each time with good results. 
Finally, the heart beat again strongly and regu- 
larly. 
Once more the artery was closed and the 
entire series of these interesting phenomena 
called back. The artery was shut for fifty 
seconds. This time, however, the heart beat 
did not return after the withdrawal of the plug. 
Massage helped some, but the highly swollen 
heart could be made to contract hut a few times. 
If the introduction of the rod had 
caused the arrest in this experiment through 
mechanical injury, the arrest would have 
been permanent. The injury done by the 
introduction of the rod would not have 
been undone by its •withdrawal. 
The observations now to be described 
appear to be conclusive; 
Experiment October 14,1895.—Dog ; 11,500 
grammes; ether ; tracheotomy; section of bulb 
near spinal cord. A glass tube, slightly conical 
at the end, with a diameter of 2’75 millimetres 
at the outlet, was connected with a bottle of 
warmed defibrinated ox blood, diluted one half 
with normal saline solution (0-6 per cent). 
The bottle was 150 centimetres above the heart. 
The blood was warmed nearly to the body temperature, and kept thoroughly 
oxygenated by a constant stream of oxygen. The glass tube was now in- W. T. Porter 
19 
troduced into the aortic opening of the left coronary artery. As usual, it 
entered the circumflex, closing the mouth of the descendens. The blood 
was at once turned on and allowed to flow for eight minutes. During this 
time 318 cubic centimetres flowed through the circumflex. The heart con- 
tinued to beat. There were no symptoms of arrest. At the end of eight 
minutes the blood was shut off. Arrest followed in 120 seconds. The left 
coronary artery was found tightly plugged by the glass tube, which ex- 
tended into the circumflex, stopping the descendens with its side. 
Experiment October 17, 1895. Dog ; 9,500 grammes ; ether ; trache- 
otomy; section of bulb. At 11.39 A. M. the descendens was ligated about 
one centimetre from the aorta. No arrest. At 12.15 P. M. the glass tube 
used in the preceding experiment was passed through the aorta into the 
circumflex artery, and warm defibrinated, oxygenated, diluted ox blood 
was sent through the tube at a pressure of 140 centimetres (blood). The 
heart was at first stimulated, the blood pressure rising. The blood pres- 
sure returned then to its former level or a little less. Every alternate 
beat was weak. After four minutes, 114 cubic centimetres of blood mix- 
ture having run through, the oxygenated blood was suddenly replaced 
with carbon-monoxide blood. The heart became within four seconds 
very irregular. After twenty-eight seconds the carbon-monoxide blood 
was turned off and oxygenated blood allowed to flow again, but it was too 
late. Four seconds thereafter arrest with fibrillary contractions occurred. 
The blood was prepared by dividing 960 cubic centimetres of defibri- 
nated ox blood into two parts and diluting each with an equal volume of 
normal saline solution (o‘6 per cent). Through one part carbon monoxide 
was passed during forty-five minutes; through the other, oxygen. Both 
mixtures were afterward warmed in stoppered flasks to near the body tem- 
perature. The carbon monoxide was made from oxalic and sulphuric acids, 
and the carbon dioxide removed by passing the gas through two flasks 
filled with a strong solution of sodium hydrate. The gas was then con- 
ducted through a wash flask, and finally into the blood. Passing the puri- 
fied carbon dioxide through barium hydrate gave no precipitate, showing 
that the carbon dioxide had been wholly removed. 
The experiments just cited show that the presence of the tube in 
the coronary artery is not the cause of the arrest with fibrillary con- 
tractions observed after closure of the artery. Were this arrest due 
to a mechanical injury by the tube, it could not have been prevented 
by circulating oxygenated blood through the tube. If the glass tube 
did not cause arrest though mechanical injury, the arrest that fol- 
lowed the introduction of the glass rod could not have been due to 
mechanical injury, for the tube was fully as large as the rod. 
Arrest with fibrillary contractions follows the stopping of the 
coronary arteries with lycopodium spores. The method is illustrated 
by the following experiment :  Further Researches on the Closure of the Coronary Arteries 
Experiment December 17, 1895.—A dog was etherized, the bulb severed, 
and artificial respiration begun. The left carotid and left subclavian ar- 
teries were tied near their origin. Ligatures were placed around the com- 
mon origin of the right subclavian and carotid arteries and about the aorta 
below the left subclavian artery, but were not drawn tight. The cannula 
of a syringe filled with the warmed, defibrinated, oxygenated blood from 
the same dog was tied into the left carotid artery. This blood contained a 
great quantity of lycopodium spores. The hitherto open ligatures around 
the right subclavian and carotid arteries and the aorta were now drawn 
tight, and at the same moment the lycopodium blood was driven through 
the carotid into the aorta. There was now hut one outlet for the blood in 
the arch of the aorta—namely, the outlet through the coronary arteries. 
These vessels were at once distended with the blood mixed with locy- 
podium. At this instant the ligatures were loosed again, forty-two seconds 
after they had been drawn tight. The circulation was free to go on as 
before. About twenty seconds later the heart, although beating strongly, 
suddenly became irregular, stopped short, and fell almost at once into 
violent fibrillary contractions. At the autopsy, the smaller coronary 
branches ramifying over the surface of both ventricles were seen to be 
plugged with masses of lycopodium. 
In this experiment the heart was not once touched by the oper- 
ator. The heart itself closed the coronary arteries. It is well 
known that a brief ligation of the aorta is borne by the dog’s heart 
without apparent injury. In this dog the heart still contracted 
strongly after the aortic ligature was loosed. There can therefore 
be no thought of mechanical injury in this experiment. 
The inquiries regarding the relation between mechanical injury 
of the heart and the results that follow the rapid closure of a coro- 
nary artery have now been answered. It has been shown (1) that 
sudden extensive injuries of the tissues lying about the coronary ar- 
teries rarely arrest the heart or do more than cause some irregular- 
ity ; (2) that in almost one hundred consecutive ligations of the coro- 
nary arteries not one case was found in which the preparation of the 
artery for ligation materially affected the beat of the heart; and (3) 
that closure of a coronary arter}r without mechanical injury produces 
all the phenomena that have been ascribed to the mechanical injury 
of the heart. Hence, of the two conceivable reasons for the changes in 
the action of the heart after closure of the coronary arteries, one—the 
mechanical injury of the heart—is excluded. The phenomena brought 
on by closure of the coronary arteries are due, therefore, to the sud- 
den stopping of the nutrition of a considerable part of the heart. W. T. Porter 
21 
YIII. The Changes in the Heart Beat that follow the 
Opening of the Great Arteries. 
The principal factor in the creation of the prevailing difference 
of opinion regarding the cause of the phenomena that follow closure 
of the arteries has been a serious misconception—namely, that the 
phenomena in question differ essentially from the changes produced 
by the sudden stopping of the supply of blood to the heart muscle, 
when the great arteries are opened freely. When the carotid arter- 
ies or other large branches of the aorta are opened, the co-ordinated 
contractions of the heart grow weaker and weaker until they finally 
disappear. Very different from this is the picture that Cohnheim 
and others have drawn of the changes following the closure of the 
coronary arteries. In the widely circulated paper of Cohnheim and 
von Schulthess-Bechberg,* it is stated that the ligation of a prin- 
cipal branch of the left coronary artery has no immediate influence 
on the action of the heart, neither the frequency nor the character 
of the pulse nor the blood pressure being altered during the first 
thirty to forty seconds, and often still longer after ligation. Only 
as arhythmia becomes pronounced does the blood pressure sink, and 
then but slightly. And this is the more surprising, for the strongly 
beating although irregular heart now suddenly stops and the arterial 
pressure falls swiftly to the abscissa (page 513), 
I have already shown the incorrectness of Cohnheim’s general- 
izations. The force of the heart beat, the height of the blood 
pressure, the power of the ventricle to discharge its contents, and 
the amount of blood thrown into the aorta, are all progressively 
diminished by those ligations that are followed by arrest, and this 
chain of alterations is seen to begin within a few seconds after the 
closure of a sufficiently large branch. They do not begin instantly 
because the capillaries are not instantly emptied, the more so for the 
reason that the pressure from behind is largely cut off by the tying 
of the arterjL The heart, then, is not struck down by a single blow 
in the midst of full activity. On the contrary, the fatal result is 
* Virchow’s Archiv, 1881, Ixxxv, pp. 503-536. 22 
Further Researches on the Closure of the Coronary Arteries 
always preceded by tlie changes mentioned above. The irritability 
of some hearts, however, is so great that they are brought to a stop 
before the characteristic alterations have gone very far. The small- 
ness of the alteration observed by Cohnheirn is probably to be ex- 
plained by the high irritability and consequent speedy arrest pro- 
duced by the curare that he used, a drug which increases the 
sensitiveness of the heart to closure of the coronary arteries, as I 
have pointed out above. 
The rise of diastolic pressure so constantly present after the 
closure of the coronary vessels is not seen when the anaemia of the 
heart muscle is brought about by bleeding from the great arteries. 
In the former case the high peripheral resistance against which the 
ventricle must pump prevents the enfeebled organ from emptying 
its contents into the aorta ; in the latter the peripheral resistance is 
almost nothing, and even weak contractions expel the contents of 
the ventricle with ease. The ventricle, moreover, receives less than 
the usual quantity of blood through the auricle. The difference in 
the height of diastolic pressure can not therefore be used in support 
of the idea that the results of anaemia due to closure of the coronary 
arteries are essentially different from those due to anaemia from great 
haemorrhage; and the alleged essential difference in the other symp- 
toms that precede arrest does not exist, as I have already clearly 
demonstrated. 
A second argument in favour of a difference between anaemia 
from haemorrhage and anaemia from closure of the coronary arteries 
is sought in the statement of Cohnheirn and von Schulthess-Rech- 
berg that “no sort of stimulus will make the dog’s heart give even 
a single contraction after standstill has occurred ” (page 516). The 
impossibility of restoring the heart beat after standstill has once 
occurred, the authors insist, shows most conclusively that the arrest 
can not be due to anaemia. According lo them, the ligature about 
the coronary artery can be loosed as soon as standstill has occurred 
and blood forced through the coronary arteries by massage of the 
ventricles, but the heart does not beat again (pages 521-525). This 
argument has no longer any force, for it has been shown by Arnaud,* 
* Arch, de physiol., 1891, pp. 396—400. W. T. Porter 
Hedon and Grilis,* Langendorff,f and myself:}; that the heart brought 
to a standstill by bleeding from the great vessels can be made to 
beat again for long periods by forcing blood through the coronary 
arteries, and I have proved in the preceding section of this paper that 
a heart arrested by plugging the coronary arteries can recover its 
power of co-ordinated contraction when the plug is withdrawn and 
the heart is massaged. There is then no essential difference here. 
It is charged, especially by Langendorff,# that fibrillary contrac- 
tions are absent when anaemia of the heart muscle is produced by 
opening the great vessels. 
Fibrillary contractions are simply inco-ordinated, irregular con- 
tractions of individual cardiac muscle cells or groups of cells. When 
a dog is bled from the carotid arteries, the contractions of the heart, 
as has been already said, grow feebler and feebler, until they finally 
disappear. Immediately or very soon thereafter three kinds of con- 
tractions can be seen in the heart—first, contractions in various parts 
of the ventricles, very often not co-ordinated with each other, and of 
small area; second, small, sharply marked, narrowly circumscribed 
twitchings here and there on the surface of the heart. These two 
kinds of movements are apparently the last traces of the action of a 
co-ordinating mechanism that is no longer master of the whole heart. 
Third, slight fibrillary contractions, commonly very faintly, sometimes 
almost or altogether unrecognizable,at other times quite unmistakable. 
There is, I believe, no fundamental difference between these 
various contractions and the pronounced fibrillary contractions here- 
tofore considered in this paper. They are identical in that they are 
all circumscribed contractions of a part of the heart wall appearing 
after a temporary or permanent loss of the power of general co-ordi- 
nation, They are further alike in that they can be made to give 
place to co-ordinated contractions. The restoration of the co-ordi- 
nated beat is easy in all mammals yet investigated, except the dog. 
Tigerstedt, Michaelis, and others declare that fibrillary contractions 
of the dog’s heart are invariably fatal. In taking this position, they 
* Comptes Rendus Soc. de biol., 1892, pp. 760, 761. 
\ Arch. f. d. ges. Physiol., 1895, Ixi, pp. 291-332. 
\ The Boston Med. and Surg. Journal, 1896, p. 39. 
# Loc. cit.  Further Researches on the Closure of the Coronary Arteries 
24 
assume a fundamental difference between the heart of the dog and 
the hearts of nearly related mammals. They oppose, moreover, their 
negative results to the positive observations of McWilliam,* who 
more than once has seen fibrillary contractions in the dog’s heart 
give place to regular co-ordinated beats ; and they do this in the 
face of the fact that no one, so far as I am aware, has attempted 
to restore the dog’s heart, after fibrillary contractions have begun, 
by the systematic employment of an artificial circulation through 
the coronary arteries. It is evident that without this, recoveries will 
be rare indeed. 
It can not be admitted, then, that fibrillary contractions differ in 
kind from the irregular, inco-ordinated contractions that appear 
after standstill brought on by opening the great vessels. 
The difference in degree is certainly great. The cause of this 
difference lies probably in the fact that after closure of the coronary 
arteries the heart continues to work against a high peripheral resist- 
ance. The co-ordinating mechanism of the strongly working anaemic 
heart soon gives way, while the less highly specialized contractility 
of the muscle fibres is still considerable. Pronounced fibrillary con- 
tractions are then possible. The heart made anaemic by oj:)ening the 
arteries, on the contrary, contracts against almost no peripheral re- 
sistance. It does little or no wcrk. The co-ordinating mechanism 
is only gradually borne down. The contractility of the cardiac mus- 
cle cell sinks with it, and when arrest has finally come the contrac- 
tility remaining is not sufficient for outspoken fibrillary contractions. 
An example of this is seen in the single case in which my closure 
of the left coronary artery with the glass rod produced arrest without 
the ordinary tumultuous fibrillary contractions. The heart of this 
dog was brought slowly to standstill, and the contractility, or, per- 
haps one should say, the irritability, of the muscle fibres was proba- 
bly so much reduced thereby that decided fibrillary contractions were 
no longer possible. The changes in the heart beat that follow the 
opening of the great arteries are therefore, in my opinion, not funda- 
mentally different from those produced by the sudden closure of the 
coronary arteries. 
* Loc. cit., p. 299.