ON THE TEMPERATURE LIMITS OP THE VI- 
TALITY OP THE MAMMALIAN HEART. By 
H. NEWELL A., M. D., D. Sc., F. E. S., and 
E. 0. A., Fellow of the Johns Hopkins 
University. With Plates XXIII, XXIY, XXV. 
Some years ago one of us1 published the results of a research 
showing that between the limits 27° C. and 41° C., the heart of 
the dog, when isolated from all other organs but the lungs, beats 
quicker the higher its temperature, so that by heating or cooling 
the blood supplied to the isolated organ through the superior 
vena cava, the pulse rate could be controlled. 
These earlier experiments made it clear for the range of temper- 
atures above stated, that the tendency to beat quicker at a higher 
temperature was a fundamental property of the heart, and that 
the cardiac muscle of the dog agreed in the dependence of its 
rhythm upon temperature changes with the histologically very 
different muscle tissue of the frog’s heart. A question which 
then presented itself was, What are the limits of temperature 
within which the heart of a mammal will beat at all? It also 
seemed of interest to ascertain whether mammalian cardiac 
muscle agreed with amphibian in having an optimum tempera- 
ture at which its rate of beat was quickest, beyond which any 
increase would slow the rhythm, without necessarily killing the 
heart. That there was some such optimum for the dog’s heart 
was rendered probable by the fact that in some of the earlier 
experiments3 it appeared that heating the organ above 42° C. 
slowed its beat. 
Attempts made three or four years ago to solve these problems 
were futile, because near the liminal temperatures, high or low, 
the heart beat with so little force that its death seemed rather 
due to deficient circulation through the coronary capillaries than 
to direct influence of heat or cold on the cardiac muscular or ner- 
vous tissues. Our problem was, therefore, to keep the cardiac 
1 H. N. M. Phil. Trans. Roy. Soc. 1883, Pt. II, p. 663, 
2 Phil. Trans. 1883, Pt. II, table on p. 681. 276 
H. NEWELL MARTIN AND E. C. APPLEGARTH. 
vessels well supplied with blood whether the heart beat feebly or 
strongly, and at the same time to vary its temperature at will. 
The first requirement it seemed we might attain by connecting 
the aortic stump of the isolated heart with a Mariotte flask filled 
with blood and kept at a constant level above the organ. Under 
such circumstances a constant pressure would be maintained in 
the coronary arteries, quite independent of the force of the 
heart’s beat. If this could be done and the heart kept alive, it 
would be easy to add arrangements for varying the temperature. 
A preliminary experiment or two having showed that the plan 
was feasible, wre arranged our apparatus essentially as indicated 
diagrammaticly in Plate XXIII. 
The heart was isolated in the manner described in previous 
numbers of this Journal, and throughout the experiment was 
contained in a large moist chamber, A ; part of the roof and of 
one end of this are indicated at a and a' in the figure. This 
chamber has no floor; it sits in a shallow iron trough containing 
water, which, during an experiment, is heated by Bunsen’s 
burners placed under the trough so as to keep it and the chamber 
warm, and the air within the chamber saturated with moisture. 
This air is (so far as necessary occasional opening the doors of the 
chamber will permit) kept at a temperature ranging between 38° 
and 40° C. The chamber used was that described in the first 
number of the fourth volume of these Studies (p. 41), and figured 
in Plate Y accompanying that number. Most accessories had, 
however, to be changed when the experiments described in this 
article were undertaken ; the Mariotte flasks, for instance, were 
placed outside the chamber instead of inside it; and a new form 
of aortic cannula had to be devised. 
The animal (cats were used in all cases) was rendered insen- 
sible by inhalation of ether or by subcutaneous injection of paral- 
dehyde, and then given curare in some cases, in others not. 
Paraldehyde in doses of about 5 cc. we found on the whole the 
most satisfactory. Tracheotomy was next performed (as a pre- 
liminary to opening the thorax), and the wind-pipe connected 
with the apparatus for artificial respiration. The thorax having 
been opened, and most vessels of the systemic circulation tied 
essentially as described previously,1 the glass cannula shown in 
from Biol. Lab., Yol. IV, p. 36. VITALITY OF THE MAMMALIAN HEART. 
277 
Plate XXIII was tied into the distal end of the aortic arch. 
This cannula1 has three side-pieces; through the one farthest from 
the heart it is supplied with defibrinated blood from the Mar- 
iotte flasks through the rubber tubej?; through the next the 
most of this blood is carried off through the tube q and poured 
into a funnel; the height of this funnel above the heart of course 
determines the pressure in the aortic stump. A thermometer was 
tied into the third branch of the cannula, as shown in the diagram. 
All branches of the aortic arch except the coronaries were of course 
closed. The blood supplied to the aortic stump could thus escape 
only through q into the funnel shown in the plate, or through the 
circuit «/, indicated by dotted lines, and consisting of the coronary 
vessels. The blood taking the coronary circuit, on reaching the 
right auricle proceeded to the corresponding ventricle, and from 
it through the lungs to the left auricle. This blood (that taking 
circuit q') was therefore the only blood entering the cavities of 
the heart or passing through the lungs, unless there were some 
inefficiency of the aortic semilunar valves. That the cavities of 
the heart were not distended with more blood, we found not to 
influence the normal character of its beat, which continued in 
many cases forcibly and rhythmically for three or four hours. 
The beat of the frog’s heart, as well known, is much promoted 
by moderately high intracardiac pressure; that this is not, at 
least to so great an extent, the case in mammals seems easily 
explicable. The frog’s heart, having no capillaries, depends for 
its nourishment or the washing out of its wastes, on the forcing 
of liquid under pressure into the spongy network of the ventricle, 
while in the mammal the nourishment of the heart depends on 
pressure in the aortic arch, from which the coronary system is 
supplied. 
The side tube q is designed to get rid of the difficulty of 
insufficient aeration of the blood; were it not present, the only 
flow from the aorta would be through the coronary system, 
and the flow would be so slow that it would be impossible 
to renew the blood in the cannula fast enough to keep it from 
using up its own oxygen and becoming very venous, and unfitted 
to keep the heart at work. But by a free outflow through q, the 
1 The cannula was modified in the later experiments, as it was found more 
easy to manipulate when made in two pieces (p. 280). 278 
H. NEWELL MARTIN AND E. 0. APPLEGARTH. 
blood in the cannula is quickly changed, and, moreover, so rapid a 
bubbling of air through the supplying Mariotte flask secured, that 
the flask takes the place of a lung and supplies arterial blood. The 
blood flowing from q into the funnel is collected in a beaker and 
poured every minute or two into the receiving Mariotte bottle, 
until the supplying bottle is nearly empty and the other full, then 
the stop-cocks are reversed and the receiving becomes the sup- 
plying bottle, and vice versa. 
The preceding sketch of the general plan of an experiment will probably be 
sufficient for most readers. In practice many unforeseen difficulties had to be 
overcome, which may be of interest to those desiring to further investigate the 
subject. A primary difficulty, which it took us long to meet successfully, was, that 
in a progressive heating or cooling experiment, the blood during its circuit changed 
its temperature so much that we could not, on interchanging the supplying and 
feeding Mariotte bottles, get a continuous series of observations. This was over- 
come by an arrangement which poured water of desired temperature into the trough 
containing the Mariotte bottles; thence it flowed through wide tubes forming 
water-jackets for the tubes p and q, PI. XXIII, and then into a water-jacket 
surrounding the funnel receiving from q, and emptying into the tank containing 
the beakers, from which the circulating blood was returned to one of the Mariotte 
bottles. Another difficulty which we soon discovered was that the temperature 
of the heart was by no means always that of the blood supplied to it; so a ther- 
mometer inserted through the precava, and with its bulb reaching into the right 
ventricle, was used to indicate the actual temperature of the heart, instead of 
the thermometer in the cannula. In the tables which follow, it is the record 
of the thermometer placed inside the right heart which is given. 
The moist chamber (A, Plate XXIY) had on its top the two Mariotte flasks 
securely fastened in a tin box, B, serving as a water-jacket. One side of B is of 
glass and turned towards the assistant to enable him to see readily when a flask 
is nearly empty, so that by reversing the stop-cocks he may convert it from the 
supplying into the receiving flask. The flasks each hold about 700 cub. cent. 
From near the bottom of each passes the outflow tube, which is conveyed 
through the spouts h of the tin box into the warm chamber, and is continued 
by the rubber tube T (the left one is for the greater part omitted in the plate), 
to an aperture in the end of the warm chamber through which it passes to one 
leg of the Y-piece q, Plate XXV. Each upper leg,/, of the Y-pieee has a stop- 
cock on it (6s), so that the supply may be taken from either Mariotte flask at 
pleasure. 
Each Mariotte flask is closed air-tight by a rubber cork pierced by three glass 
tubes (Plate XXIV) ; one of these, o, is connected by a Y-pieee with the funnel 
P, from which it gets blood when acting as receiving flask ; a second, n, serves 
to let air out as blood enters the flask ; both n and o are of course closed when 
the flask is acting as supply bottle to the heart. The third tube, m, reaches to 
near the bottom of the flask, and through it air bubbles in when the flask is 
acting as supply bottle. Prom its upper end passes the rubber tube, seen again VITALITY OF THE MAMMALIAN HEART. 
279 
at m in Fig. 1, Plate XXV, where only the tube of one side is drawn ; of course 
in actual use there must be a like tube attached to the other Mariotte flask. The 
long rubber tube m passes through the cork of a tightly closed test-tube, Tb, 
containing a little water. Another tube, at. passes to near the bottom of the 
test-tube and is open. When the Mariotte flask is at work all the air entering 
it must first bubble through the water in the test-tube, which, hanging outside 
the chest and the water-jacket, is easily observed, and indicates how fast the 
blood from the flask is flowing, or if the bubbling stop, that something has gone 
wrong and the supply bottle has ceased to act as a Mariotte flask. It is not 
necessary to describe in detail the stop-cocks which lie on all the tubes so that 
they can be in a moment closed or opened when the functions of the two flasks 
are to be reversed. 
All the system of blood-carrying tubes is inclosed in a set of water tubes con- 
tinuous with the spouts hh of B, and having warm or cold water from B flowing 
steadily through them during an experiment. Over those ends of each of the 
spouts which open into the warm chamber is fitted thin rubber tubing inches 
in diameter ; these tubes surround the blood-conveying tubes TT, and finally 
end (Pig. 3, Plate XXV) in J and J, which open into a funnel-shaped tin water- 
jacket. Inside this jacket the Y-piece q (Pig. 1, Plate XXV) is placed, and 
its upper limbs// pass through JJto join TT. On the sides of the tin jacket 
are openings through which pass the handles (bdr, Fig. 2) of the stop-cocks 
bs bs (Fig. 1). 
The lower limb Q, Pig. 2, Plate XXV, of the tin jacket surrounds q, Pig. 1 
(which conveys the blood again into the warm chamber) and is continued over C. 
This tube ((7, Plate XXIV) leads to the aortic cannula a, and up to about six 
inches from that point is jacketed by a wide rubber tube. At this point the 
jacketing tube is securely closed by a rubber stopper, through which a glass can- 
nula, a, continues the blood-carrying system to the aortic stump. The reason for 
stopping the jacketing for a short distance at this point is to facilitate insertion 
of the cannula. 
So far, then, it will be seen that water supplied to B, Plate XXIV, will flow 
out through the wide tubing surrounding TT through the jacket JJ (Plate XXV) 
to the tube around C, keeping the blood under essentially the same temperature 
conditions during its whole journey from the supply bottle to the heart. Prom 
near the rubber cork closing the jacket of G the water finds its exit through a 
side branch, a wide rubber tube surrounding e, and so to the wide glass funnel 
f, Plate XXIV. This funnel has a short stem into which is accurately fitted the 
stem of the blood-receiving funnel (R) by means of rubber tubing. From/ 
the water issues through the outflow piece through the jacketing hose I into the 
tin kettle y (Plate XXV) containing the beakers for catching the blood; and 
finally into the waste bucket through the tube Wsf. 
The only parts of the blood-conveying system, then, which are not jacketed 
are the funnel P, the tubes oo leading from it to the Mariotte bottles, and the 
cannulas near the heart with a few inches of adjacent tubing. With the 
exception of the funnel P, however, most of these, including the innominate and 
aortic cannulas, were packed in a layer of cotton wool, thus checking any rapid 
loss or gain of heat. 280 
H. NEWELL MARTIN AND E. C. APPLEGARTH. 
All the water-jackets, or rather the whole jacket, for the system is but one, 
are supplied directly from the water pipe of the room. This pipe by means of a 
T piece is connected with an air chamber whose capacity is 18 gals., and which 
is capable of standing a high pressure to the square inch. From the bottom of 
this chamber a delivery pipe, W, Pig. 1, Plate XXV, runs to terminate in the 
stop-cock D, which stands in connection with a two-way stop-cock C, by means 
of which the current of water can be shunted either to the heater, R, or in the 
case of a cooling experiment to a worm (not represented in the figure) placed in 
a freezing mixture and substituted for the heater. In this way a constant stream 
of any desired temperature may be obtained. 
The heater consists of about 30 feet of small brass pipe wound into a coil and 
placed over a small Bunsen gas-stove. The stove and coil are entirely sur- 
rounded by the screen (Scr) provided with its chimney, H. Before the water 
passes to the heater, however, it is made to go through Ww to a worm of 25 ft. 
of £-inch bore lead pipe put in the hot water of the pan which forms the floor 
of the moist chamber. After traversing this worm it comes to the heater 
through Rw. In this manner the stream of water can be warmed even before 
it reaches the heater. The temperature of the water may be regulated by with- 
drawing the worm totally or in part from the water in the floor of the chamber, 
by the amount of gas burned, and by the rate of flow of the water. Of course 
in a cooling experiment this second worm was not used. 
After issuing from the heater (or, in the case of cooling experiments, the 
worm in the freezing mixture) the stream is directed by Rw through the regis- 
tering apparatus (Rg) where its temperature is ascertained. This apparatus 
is simply a T piece, one leg of which carries a thermometer. From this the 
water passes up Bw (Plate XXV) to the jacket .B (Plate XXIV) and is then distri- 
buted over the whole system just described. 
Dr (Fig. 1, Plate XXV) is a small sliding door which gives the assistant 
access to the interior of the case during the experiment. While an experiment 
is going on, the main doors of the moist chamber are opened as little as possible. 
As will be seen, the cannula in Plate XXIV differs in detail from that shown 
in the diagram Plate XXIII. It is made in two pieces. One, a, is inserted in 
the aortic stump, and receives blood from the Mariotte flask. By its side branch 
k it was in some cases connected with the manometer of a kymograph. The 
other piece, i, is inserted in the innominate artery and contains a thermometer 
bulb. From it the tube e carries off the excess blood to the funnel R. 
An attempt was made at first to get a graphic record of the pulse, by con- 
necting a manometer with the aortic cannula; but the quantity of blood 
reaching the left ventricle was so small that the pulse waves due to its beat 
were indistinct and sometimes imperceptible. We had therefore to resort to 
counting by direct observation of the heart, which may have led to slight errors 
when the pulse was very fast and very feeble, but as our object was not to study 
the absolute pulse rate, small errors in regard to it were of no great importance. 
Moreover, by always taking two counts immediately following one another the 
chances of error were greatly lessened. First, one of us held the watch for 
thirty seconds while the other counted the heart beats. Then the duties were 
reversed and the other made the count, still being ignorant of the number VITA LITY OF THE MA MM A LI AN HEART. 
281 
arrived at by the previous counter. If the two counts did not agree, but 
differed only by three or four in a minute, their average was put down in the 
tables as the correct rate ; if the difference was greater (as was sometimes the 
case when rapid changes of temperature were occurring) a fresh count was 
made or the observation rejected altogether. 
In all cases at least half an hour was allowed to elapse after 
complete isolation of the heart before the actual experiment 
commenced. This was in order to eliminate possible disturbing 
elements, as irritation of the vagi or accelerators, due to the 
operative procedure, or the first effects on the heart itself of the 
defibrinated blood. The general results of the experiments can 
be stated in few words. 
First, as to cooling. The isolated heart of the cat may be cooled 
down to a temperature of 16.5° C. (as indicated by a thermome- 
ter introduced into the right heart) and yet not be killed, as it 
revives if soon warmed again (see Table 1, p. 282), but it usually 
dies at about 17° or 18° C. 
As the cooling proceeds, the pulse becomes slower and slower; 
this has two causes. As pointed out in previous papers, after 
about the first half-hour the pulse rate of the isolated heart tends 
to become slower the longer its isolation ; so that a dog’s heart 
isolated for one hour and beating, say, 220 times a minute at a 
temperature of 39° C., will at the end of two hours and at the 
same temperature beat only 200, or perhaps 190 times a 
minute. But the slowing in such cases is very gradual, and the 
heart will beat vigorously for four or five hours if not cooled. 
When the influence of cold is added, the fall in the pulse rate is 
far more rapid, and death of the isolated organ, if the cold be 
continued, occurs much sooner than it otherwise would. 382 
H. NEWELL MARTIN AND E. C. APPLEGARTH. 
Cooling experiment with recovery, June 13, 1888. 
Adult cat narcotized with paraldehyde during isolation of 
the heart. Isolation completed at 3.30 P. M. 
Table I. 
Time P. M. 
Temperature 
C. in Bight 
Heart. 
Pulse perl'. 
Bemarks. 
h. in. 
4.08 
34.0 
183 
4.11 
33.7 
175 
4.14 
32.7 
158 
4.17 
30.9 
138 
4.20 
29.9 
124 
4.23 
28.5 
111 
4.26 
28.5 
92 
4.29 
26.7 
80 
4.32 
25.7 
74 
4.35 
24.7 
66 
4.38 
24.0 
61 
4.41 
23.5 
54 
4.44 
23.0 
53 
4.47 
22.3 
50 
4.50 
21.7 
40 
4.52 
20.5 
J 
4.56 
19.9 
? 
5.00 
19.5 
12 
5.04 
19.2 
11 
5.06 
18.9 
9 
5.12 
18.6 
9 
5.15 
18.0 
9 
5.20 
16.5 
5 
Heating now commenced. 
5.40 
21.0 
14 
5.45 
24.0 
20 
5.50 
27.3 
78 
Contractions becoming very feeble. 
5.56 
29.7 
85 
6.06 
31.7 
90 
The results obtained on heating the heart to or near to its 
death point are more complex than those observed on cooling, 
but also more interesting. 
In the first place we found that there ■was an optimum temper- 
ature at which the isolated cat’s heart, like that of the frog, beat 
quickest, any rise beyond this slowing the beat. It by no means 
follows, of course, that this optimum is the temperature at which 
it would do most work. Averaging the results of thirteen sepa- 
rate experiments, this optimum is 41.3° C., ranging between 
43.3 and 40.6. 
If one steadily and slowly heats the heart to the highest 
temperature to which it can be raised without dying, this lethal VITALITY OF THE MAMMALIAN HEART. 
283 
temperature lies between 44.5° and 45° C. in the great majority 
of cases; but sometimes by working with care, one can raise this 
maximum as well as the optimum temperature. If as soon as the 
heat is beginning to slow the pulse, or the heart shows other signs 
of weakening, the organ be cooled a little for a short time and then 
heated again, it can often be raised to a higher temperature without 
being weakened, and the optimum temperature also raised. Table 
II will serve as an example of a simple maximum and optimum 
experiment, while in Table III is given an example in which the 
optimum is raised by slightly cooling the heart from near the 
lethal temperature and then very slowly heating it again. *This 
power of the heart to rapidly accommodate itself to an abnormally 
high temperature was so surprising to us that we made repeated 
observations on the matter. Since then we have been informed 
by friends engaged in the practice of medicine that something 
similar is frequently seen in persons suffering from remittent fevers. 
A rise of temperature associated with considerable quickening of 
the pulse may, later in the disease, be associated with a less marked 
quickening. 
When the temperature of the heart rises to about 40° C., even 
though the optimum be not yet reached, small increments of 
temperature often have no effect on the rate of beat of the 
isolated heart, or the pulse may even become slower: this no 
doubt is due to that natural gradual slowing of the pulse of the 
isolated heart, always occurring as time goes on, and referred to 
above. 
Another phenomenon often observed in heating experiments 
is what we may call a long latent period. An increase of 
temperature may not show any effect on the pulse rate for a 
minute or two, then it comes on, though the heart may in the 
meantime have been slightly cooled. For this reason (even 
before the optimum is reached) variations in the pulse curve tend 
to lag a little behind the temperature variations which led to 
them. 284 
H. NEWELL MARTIN AND E. C. APPLEGARTH. 
Table II. 
Simple heating experiment to show optimum and maximum. 
June 7, 1888. Adult cat. Paraldehyde. Isolation of heart 
completed at 11.15 A. M. 
Time. 
Temperature 
C. in Right 
Heart. 
Pulse per V. 
Remarks. 
h. m. 
11.51 
34.1 
168 
11.54 
34.8 
174 
11*57 
35.3 
186 
13.00 
36.5 
188 
12.03 
36.9 
186 
12.06 
38.3 
189 
13.09 
38.8 
190 
12.13 
40.0 
200 
13.15 
40.3 
201 
12.18 
40.5 
300 
12.21 
40.5 
200 
12.24 
41.2 
203 
12.27 
41.3 
303 
12.30 
41.0 
200 
12.33 
41.8 
206 
13.36 
43.5 
309 
13.39 
43.3 
210 
Optimum or probably rather above it. 
13.43 
44.7 
160 
12.44 
45.7 
? 
Pauses alternating with periods of 
13.47 
46.0 
very rapid beats impossible to count. 
Fibrillar contractions. 
Optimum temperature raised during an experiment. January 
31, 1889. Young cat. Ether and curare. Defibrinated blood 
used to feed the heart, diluted with one-third its volume of normal 
saline. Isolation completed at 2.55 P. M. 
Table III. 
Time. 
Temperature 
C. in«ide 
Right Heart. 
Pulse per V. 
Remarks. 
h. m. 
3.51 
39.3 
234 
3.55 
39.5 
236 
3.58 
39.5 
220 
4.01 
39.8 
245 
4.03 
40.3 
240 
4.06 
41.0 
292 
Temperature now raised with great 
4.09 
41.0 
280 
care, as heart is near its critical 
4.12 
41.0 
272 
point. 
4.15 
41.5 
304 
First optimum. 
4.18 
41.5 
? 
Irregular and not countable. VITALITY OF THE MAMMALIAN HEART. 
285 
Time. 
Temperature 
C. inside 
Right Heart. 
Pulse per V 
Remarks. 
4.22 
41.5 
246 
Regular. Heart apparently getting 
4.25 
41.0 
248 
used to this temperature. 
4.28 
41.0 
204 
4.31 
40.5 
208 
4.34 
41.0 
204 
4.37 
41.8 
224 
4.40 
42.0 
? 
4.43 
42.0 
228 
4.46 
42.5 
234 
Second optimum 1° C. above the first. 
4.48 
42.8 
188 
4.51 
42.3 
198 
4.54 
43.0 
200 (?) 
Somewhat irregular. 
4.57 
43.5 
178 
Regular. 
5.00 
43.0 
178 
5.04 
42.5 
176 
5.07 
42.8 
176 
5.10 
43.0 
168 
5.13 
44.5 
Very irregular. 
5.16 
44.0 
176 
Regular. 
5.19 
5.21 
48.0 
172 
Very irregular but not feeble beats ; 
the irregularity probably due to 
the rapid change of temperature. 
Only right auricle beating. 
It is, of course, not possible to say at what temperature the 
heart of Table III would have died had the heat been slowly 
pushed on at 4h 18m, instead of stopping and cooling from 4h 25m 
to 4U 34m, thus giving the organ time to accommodate itself to 
the high temperature. But from many other observations in 
which the heating was slowly pushed on, we feel sure that all 
pulsation would have ceased at a temperature several degrees 
below 48° C. 
It will be observed that though the temperature of the second 
optimum is higher than that of the first, the pulse is slower, this 
being no doubt due to the increasing malnutrition of the heart as 
time goes on. STUDIES FROM THE BIOL LAB 
VOL IV N°5. PI ATE XXIII VOL IV N°5. PLATE XXIV 
STUDIES FROM THE BIOL LAB STUDIES FROM THE BIOL LAB 
VOL IV N°5. PLATE XXV