It has been suggested repeatedly that the mammalian response to anoxia of hyperpnoea and hypertension may be the survival in air-breathing vertebrates of a phylogenetically primitive mechanism present in fish (Marshall & Rosenfeld, 1936; Schmidt & Comroe, 1940; Schmidt, 1956; Heymans & Neil, 1958). Black (1951), reviewing the results of four previous studies on the effect of deoxygenated sea water on elasmobranch fish, concluded that there was no change in respiration, whilst Lutz (1930) showed that restriction of water flow through the dogfish pharynx (a change likely to cause anoxia) resulted in a slowing both of respiration and heart beat. Thus the evidence, fragmentary though it is, does not suggest that the response of an elasmobranch fish to anoxia is like that of a mammal. In an attempt to re-examine this problem using the common local dogfish Squalus acanthias* L., the work of Lutz has been confirmed and amplified. Reduction of the minute volume of pharyngeal flow caused bradycardia and respiratory slowing. It soon became clear, however, that the responses of the heart and of respiration were mediated by separate mechanisms. Respiratory rate was related primarily to minute volume and was much less influenced by the oxygen content of the water respired. The heart, by contrast, was reflexly slowed by anoxia, whether this was induced by a restricted flow of normally oxygenated water or an ample flow of deoxygenated water.

In this paper the influence of changes in minute volume on the cardiac and respiratory rates will be re-examined and the mechanisms mediating the response to anoxia will be described. The reflexes concerned in the response of respiration to changing flow rate and the receptors from which they arise will be presented in a subsequent paper.

Mounting and perfusion

Thirty-three specimens of S. acanthias varying from 312 to 6 lb. in weight were used. Each fish was secured by two clamps to the trunk and one to the snout in a rectangular tank filled with cooled circulating sea water; the water temperature varied from 8° to 15° C. If the fish was to be dissected, it was anaesthetized by an injection of 3 ml. of 1 % MS 222 Sandoz.

In all experiments except one the mouth was sewn closed and the spiracles were cannulated with specially shaped glass tubes which fitted snugly inside the spiracular rim. These cannulae were connected to a receiving vessel into which sea water poured at a known rate. This rate was changed by a series of resistances that could be cut in and out of circuit by means of taps, providing sixteen known flow rates varying from zero to 985 ml./min. Deoxygenated or CO2-enriched sea water was supplied to the fish by a two-way tap inserted close to the cannula junction. Similar flow rates of normal and deoxygenated water were achieved by ensuring that the polythene connexions between the two-way tap and the two supplies of water were identical in length and diameter, and that the pressure heads were identical. As the deoxygenated water left its container, the space above was filled with nitrogen, maintained at atmospheric pressure with the aid of a manometer.

Deoxygenated sea water was prepared by boiling the water for at least 2 hr. at 160 mm. Hg and bubbling with nitrogen for 2 hr. when cool. CO2-enriched sea water was prepared by equifibrating sea water with ‘Carbogen’ (5% CO2, 95% O2) and returning the pH to 7·9 (the pH of local sea water) with sodium bicarbonate. Normal and deoxygenated sea water were maintained at temperatures that did not differ by more than 2°C., and results were not accepted until they had been demonstrated with the deoxygenated water both above and below the temperature of the normal sea water.

Recording

Records were made on either an Ediswan or a Both four-channel pen writer. The electrocardiogram (e.c.g.) was recorded in the manner described in a previous paper (Satchell, 1961); respiratory movement was recorded using a Statham strain gauge and amplifier. The strain gauge was attached by a thread either to the side of the pharynx above the first gill opening, or, if the fish were mounted upside down, to the region of the mandibular symphysis. The output of the strain gauge was amplified by an a.c.-coupled amplifier (time constant 1 sec.) in earlier experiments; by a d.c.-coupled amplifier in later experiments.

Blood pressure was recorded using the sensitive Statham P23BB pressure transducer coupled by 61 mm. of 2 mm. p.v.c. tubing to a no. 19 needle 1 in. long inserted into the dorsal aorta. This was exposed by cutting off the tail at the level of the posterior dorsal fin. The P23BB transducer does not have a good high-frequency response (9 cyc./sec. with critical damping) but in these experiments it was the systolic and diastolic pressures that were of interest, rather than the wave-form of individual pulse beats. The transducer output was amplified by a carrier-wave amplifier, and calibrated repeatedly during the experiment by recording known pressures; all pressures are given in mm. Hg above the tank water level.

The relation of minute volume to respiratory and cardiac rate

A curve showing the respiratory and heart rates at sixteen different minute volumes ranging from zero (flow turned off) to 985 ml./min. is shown in Fig. 1. Each point is an average of the heart beats or respirations occurring during 1 min. Low rates of flow were not allowed to continue for periods longer than this, and were interspersed with periods of maximum flow.

Fig. 1.

A, cardiac and B, respiratory rates at different minute volumes of pharyngeal water flow. ⬤, Normal, 8° C. ; ○, curarized, 9°C.

Fig. 1.

A, cardiac and B, respiratory rates at different minute volumes of pharyngeal water flow. ⬤, Normal, 8° C. ; ○, curarized, 9°C.

The respiratory rate (Fig. 1B, ○) decreased with decreasing minut evolume throughout the range. The more pronounced slowing as the flow rate dropped below 60 ml./min. was a constant feature ; as the respiration slowed, its amplitude increased. Visual observation showed the fish to be making greater inspiratory efforts, and at rates below 200 ml./min. the closed external gill openings became more obviously indented during the inspiratory phase. Schoenlein & Willem (1895) noted this in the skate and described such respiration as dyspnoeic.

The curve of heart rate (Fig. 1A, ○) tended to flatten out at both low and high flow rates. Though the resting rate varied from one fish to another, the shape of the curve, with the increased steepening at rates below 200 ml./min., was seen in other experiments.

Is the slowing of heart and respiration due to myocardial or medullary anoxia?

Reducing the flow of water through the pharynx usually caused a cough, and this is itself liable to cause cardiac slowing. But even when no cough occurred, the next heart beat following the reduction in flow rate was delayed. The short latency of this response (3−4 sec. depending on the heart rate) argues against a direct action of anoxia on the myocardium or medulla.

Intravenous atropin (0·01 mg./kg.) blocked the action of the cardiac vagus and abolished the short-latency slowing of the heart. When the water flow was stopped no cardiac slowing appeared until 5 min. had elapsed, and even after 12 min. the rate was still twice that observed with zero flow before atropinization. Cutting the cardiac vagi similarly abolished the short-latency cardiac response to decreased minute volume.

The muscles effecting closure of the gill slits have proved to be more resistant to the action of curare than the rest of the branchial musculature, and 0·075 mg./kg. of tubocurarine produced a preparation in which all respiratory movement appeared to have ceased. Re-attaching the thread of the strain gauge to the anterior margin of the first gill flap and increasing the gain of the amplifier showed, however, that some muscle fibres remained active, for a recordable trace resulted. Following curarization to this extent respiratory rate and minute volume ceased to be related. The significance of this finding will be discussed in a subsequent paper, but the observation that the respiratory rate had not fallen after 5 min. of zero flow is relevant here. With continued zero flow the respiratory rate of 26/min. recorded for zero flow before curarization was not achieved until the eighth minute. Thus the evidence indicates that although myocardial and central anoxia do produce cardiac and respiratory slowing if the water flow is turned off for more than 5 min., they cannot explain the shortlatency slowing in response to reduced flow.

Is the slowing of heart and respiration mediated by a common mechanism?

Evidence on this was obtained from experiments in which deoxygenated sea water could be switched into the perfusion circuit. The normal slowing of heart and respiration in response to a reduced or zero flow of normally oxygenated sea water was first recorded; then a flow of deoxygenated sea water was turned on. The respiration promptly speeded up again, and the same range of respiratory rates as had previously been witnessed with different minute volumes of normally oxygenated sea water could be demonstrated by varying the flow rate of deoxygenated sea water. A low or zero flow of deoxygenated sea water produced respiratory slowing: a faster rate caused the respiration to speed up.

The behaviour of the heart contrasted with that of respiration. A flow of deoxygenated sea water caused a slight enhancement of the slowing already resulting from a zero flow of normally oxygenated sea water: alteration of the flow rate of deoxygenated sea water did not change the bradycardia, which persisted as long as deoxygenated sea water remained in the pharynx. Restoration of normally oxygenated water caused the heart to speed once again within one or two beats of the change.

The results suggest that the respiratory response is dependent on flow whether it be of oxygenated or deoxygenated water, whilst the cardiac response is not primarily related to flow, but to the rate at which oxygen is supplied to the fish.

The responses of the fish curarized as described in the previous section are significant here and are shown in Fig. 1. The respiratory rate (Fig. 1B, O) showed no significant changes with alteration in minute volume; statistically the two are unrelated (r = −0·27, P > 0·10). Changes in the amplitude of respiration occurred, but in the absence of normal movement failed to evoke the corresponding changes in rate. The heart, in contrast, still slowed in response to a reduction in minute volume (Fig. 1 A, O), showing that this mechanism was independent of respiratory movement. Similarly, deoxygenated water caused cardiac slowing in the curarized fish.

At this stage in the study it became clear that since respiratory rate was so dependent on the rate of water flow regardless of its oxygen content, any effect that anoxia might have on respiration could only be evaluated if precautions were taken to ensure that the flow rates of normal and deoxygenated sea water were identical. All data presented subsequently in this paper were obtained from experiments in which this was done.

Changes in heart rate and blood pressure in response to anoxia

The period of deoxygenated water flow was set routinely at 2 min. Myocardial anoxia was not likely to be causing slowing of the heart rate since in an atropinized fish the rate showed no change within 2 min., and even after 5 min. flow of deoxygenated water had only fallen from 38 to 35/min. Flow rates of deoxygenated and normally oxygenated water were always kept the same, and varied in different experiments from 1 l./min. to 500 ml./min.

The change in heart rate is shown in Fig. 2. Each rate determination is an average derived from the three preceding beats. The sudden onset and cessation of the slowing is evident. The restoration of normally oxygenated water was always followed by an acceleration to a level above the resting rate, which was not regained until the end of the third minute after the return of normal water. This pattern of change was very constant from one experiment to another, the slowest rate obtaining during the deoxygenated water flow being between 5 and 7/min. The tendency for the slowing to be incompletely maintained and for the rate to rise towards the end of the second minute was a general feature not well shown in Fig. 2, but apparent in Fig. 3 A, ○.

Fig. 2.

Change in heart rate (lower curve), and in diastolic (middle curve, ○) and systolic (upper curve, ⬤) blood pressure, caused by a 2 min. flow of deoxygenated water = black inset. 12°C.

Fig. 2.

Change in heart rate (lower curve), and in diastolic (middle curve, ○) and systolic (upper curve, ⬤) blood pressure, caused by a 2 min. flow of deoxygenated water = black inset. 12°C.

Fig. 3.

Changes in A, cardiac and B respiratory rate caused by a 2 min. flow of deoxygenated water = black inset. ⬤, normal, 10-11°C. ○, after gill deafferentation. 14-15°C.

Fig. 3.

Changes in A, cardiac and B respiratory rate caused by a 2 min. flow of deoxygenated water = black inset. ⬤, normal, 10-11°C. ○, after gill deafferentation. 14-15°C.

Blood-pressure records from the dorsal aorta showed a drop in systolic pressure (Fig. 2, ○) of 10 mm. Hg from the resting systolic level of 36 mm. Hg. The diastolic pressure (Fig. 2, ○) fell 14 mm. from its resting level of 34 mm. Hg. The increase in the pulse pressure from 2·3 to 6 mm. Hg was due to the low diastolic pressure resulting from the less frequent heart beats. Following the return of normally oxygenated water the systolic pressure rose to 40 mm. Hg and this hypertensive overshoot outlasted the period of raised heart rate, suggesting some increase in resistance in the peripheral circulation. The saw-toothed contour of the record of systolic pressure during this period was due to the increase in blood pressure at each expiration periodically coinciding with the systolic peak of pressure.

Respiratory changes in response to anoxia

The changes in respiratory rate evoked by deoxygenated water were not profound ; Fig. 3B, ○, is typical. During the first minute the rate fell from 26·5 to 23·5/min., and then rose again to exceed the resting rate, attaining 29·5/min. in the second minute. With the return of normally oxygenated water the rate did not immediately fall; in some fish a further increase in rate occurred. Sometimes the normal rate was reestablished within 2 min. ; more commonly it slowly returned to the resting rate over 10 min.

The depth of respiration increased during the period of deoxygenated water flow, sometimes steadily throughout the 2 min., sometimes only during the first minute. The changes in depth were never very marked and in some experiments were scarcely discernible. Compared with the changes in depth that could be evoked by varying the flow rate, they were small indeed.

The possibility that closing the mouth and cannulating the spiracles itself masked some respiratory or cardiac response was excluded in an experiment in which the head and fore part of the body were enclosed in a separate container within the main tank. Deoxygenated or normal water was fed into this container, the fish respiring freely with its mouth and spiracles unimpeded. Inevitably the change from normal to deoxygenated sea water was slow and incomplete but the characteristic respiratory responses of decreased rate and increased depth were seen. Cardiac slowing, though less profound, was quite evident.

The mechanisms mediating the response to anoxia

It is now generally agreed that the respiratory hyperpnoea of mammals in response to anoxia is reflexly mediated and originates in the chemoreceptor cells in the carotid and aortic bodies (Adams, 1958; Heymans & Neil, 1958).

Granel (1927), in a detailed morphological study of the pseudobranch in fish, showed that in teleosts there are clusters of acidophil cells surrounding the blood vessels which resemble the glomus tissue and chemoreceptor cells of the carotid body in mammals. The possibility that the response to anoxia in the dogfish depended on receptors in the pseudobranch had therefore to be examined. Experiments failed to confirm this. Fish in which both pseudobranchs were (a) surgically removed, (b) extensively cauterized, (c) frozen by cannulating the spiracles with copper tubes conducting freezing mixture at — 5°C., (d) rendered ischaemic by cutting or clipping the afferent pseudobranchial artery, all showed an unimpaired response to anoxia However, in experiments involving gill deafferentation the pseudobranchs were routinely cauterized in case some small part of the response to anoxia depended on them.

The gills were deafferented by opening the anterior cardinal sinus and cutting the glossopharyngeal, the first three branches of the vagal and the pretrematic branch of the fourth vagal branchial nerves. The post-trematic branch of the fourth vagal branchial nerve carries inhibitory fibres to the heart and was left intact ; the deafferentation was not thereby impaired since there is no gill on the posterior wall of the fifth gill pouch.

In Fig. 3 A the cardiac responses to anoxia before, ○, and after, ◯, deafferentation are compared. Following deafferentation, the bradycardia during the period of deoxygenated water flow was less intense, and its onset was delayed, 90 sec. elapsing before slowing was fully developed ; the speeding of the heart following the return of normally oxygenated water was not significantly changed. The persistence of some cardiac response to anoxia, although of reduced intensity and delayed onset, could be interpreted in one of two ways. Either the response to anoxia was confined to the central nervous system, being dependent on the flow of deoxygenated blood from the gills to the brain. Its impairment might then be ascribed to the blood loss and injury incidental to deafferentation. Or there existed both a peripheral reflex response characterized by its intensity and speed of onset, and a central direct response dependent on the attainment of a critical level of anoxia in the brain and therefore of longer latency.

The first interpretation was excluded by experiments in which the conus arteriosus was exposed and clamped with artery forceps. After the lapse of 30 sec., to allow the blood in the ventral aorta to pass into the peripheral circulation, deoxygenated water was turned on. The heart was inhibited with undiminished intensity (Fig. 4, ◯) and the return of normally oxygenated water caused the heart to speed again. This was significant in excluding the possibility that anoxia of central origin could alone account for the slowing. As will be shown later, such slowing does occur but it is less intense and has a longer latency. Following the removal of the artery forceps there was a second slowing as the banked-up venous blood passed rapidly into the branchial vessels. In a further experiment the conus was clamped and the ventral aorta cut at the junction of the conus with the aorta. Ringer’s fluid was perfused into the dorsal aorta so that it welled out of the cut end of the ventral aorta. Despite this retrograde perfusion of the branchial blood vessels deoxygenated water still caused cardiac slowing and the heart accelerated on the return of normally oxygenated water. This evidence points strongly to the existence of a reflex response to anoxia with the gills as the most likely source of the afferent information in that they are thin walled and the most permeable structures in the path of the deoxygenated water.

Fig. 4.

Changes in heart rate: A, caused by a 2 min. flow of deoxygenated water (black inset); B, the same but with conus occluded (white inset). C, changes caused by conus occlusion alone (stippled inset) after atropinization. 10−11° C.

Fig. 4.

Changes in heart rate: A, caused by a 2 min. flow of deoxygenated water (black inset); B, the same but with conus occluded (white inset). C, changes caused by conus occlusion alone (stippled inset) after atropinization. 10−11° C.

The cardiac response that persisted after deafferentation was presumed to originate in the brain. Evidence supporting this has been obtained in experiments on deafferented fish perfused with normally oxygenated sea water in which the conus has been clamped with artery forceps for a 4 min. period. In Fig. 5 the response is compared with that evoked by a 2 min. period of deoxygenated water flow without the clamp. Despite the disturbances in heart rate associated with closing the artery forceps, the heart slowed steadily from the end of the second minute to the removal of the clamp. In the atropinized fish (Fig. 4) the heart had only slowed from 26·5 to 24/min. after 6 min. of conus occlusion. The intense slowing that occurred after removing the artery forceps was, as in the intact fish, associated with the surge of banked-up venous blood through the gills to the brain. It occurred in other experiments in which the conus was clamped for 6 or 8 min. and was associated with removal of the forceps rather than with a particular duration of conus occlusion. As it occurred in a deafferented fish, it must also be related to a central response to inadequately oxygenated blood.

Fig. 5.

Changes in cardiac rate in fish with gills deafferented caused by: ⬤, 2 min, flow of deoxygenated water (black inset) ; ○, a 4 min. period of conus occlusion (white inset). 14-15°C.

Fig. 5.

Changes in cardiac rate in fish with gills deafferented caused by: ⬤, 2 min, flow of deoxygenated water (black inset) ; ○, a 4 min. period of conus occlusion (white inset). 14-15°C.

When the two traces of Fig. 5 are compared, it is seen that in a deafferented fish a 2 min. flow of deoxygenated water evoked cardiac slowing sooner (50 sec.) than did occluding the conus (2 min.). Presumably perfusion of the brain with blood equilibrated with deoxygenated water lowered the oxygen tension more rapidly and more completely than did the stasis imposed by occluding the conus.

The speeding of the heart following the period of anoxia

As the cardiac acceleration beyond the original rate, following the return of normal water flow, persisted in the deafferented fish (Fig. 3 A), it may be ascribed to some change either in the brain or in the pacemaker of the heart itself. The slowing of the heart and reduction in blood flow during anoxia must result in an increase in carbon dioxide tension and in hydrogen-ion concentration in the medulla and peripheral tissues alike. The response to perfusion of the pharynx with water equilibrated with 5 % CO2 and 95 % O2 was investigated. In the intact fish a profound cardiac and respiratory inhibition resulted. Little significance is attached to this, since such a concentration of CO2 may well be an irritant to the gills. However, in the deafferented fish 5 % CO2 caused acceleration of heart rate from 19 to 25/min. and of respiration from 46 to 48/min., without any early inhibitory episode. Thus a raised tension of CO2 in the brain and tissues caused cardiac acceleration and may account for the speeding following anoxia. It is not possible at present to decide whether the effect is exerted upon the vagal centres or upon the cardiac pacemaker. As the respiratory rate is simultaneously changed in the same direction, it is likely that there is a central response.

Respiratory responses in the deafferented fish

Deafferentation eliminated the respiratory slowing (Fig. 3B, O) and increase in depth associated with deoxygenated water, but the subsequent acceleration remained. The evidence cited above that 5% CO2 caused some respiratory speeding in the deafferented fish suggests that the speeding following deoxygenated water flow may similarly be ascribed to a rise in CO2 tension in the brain. The slowing and the increase in the amplitude both appear to be reflex in origin.

The studies of Schoenlein & Willem (1895) on Scyllium canicula and Torpedo ocellata, of Bethe (1925) on S. canicula and S. catulus and of Ogden (1945) on Mustelus antarcticus agree that deoxygenated water has no effect on either rate or depth of respiration when captive fish are immersed in it. Their findings are not so at variance with those reported here, for the method of perfusion described earlier permitted high rates of deoxygenated water through the pharynx free of admixture with normal sea water. Notwithstanding that perfusion with deoxygenated water must have provided a maximal anoxic stimulus the changes in respiratory rate and depth were not profound. It is not surprising that they were unobserved by earlier workers; it may be doubted whether such changes, as distinct from the changes evoked by varying flow rates, ever occur naturally in these inhabitants of shallow, well-aerated seas.

Evidence has been presented that the cardiac slowing in response to decreased minute volume of water flowing through the pharynx was mediated by a cardioinhibitory response to anoxia. This consisted of an intense reflex inhibition of short latency and a less powerful direct response of slower onset. The evidence from deafferentation suggests that the chemoreceptors are located in the gills; as the receptors responded to deoxygenated water in the absence of branchial blood flow it would appear that they were so superficially placed as to be accessible by diffusion from the gill epithelium. Apart from this, nothing is yet known of their structure, location or function.

The cardiac response to anoxia in mammals contrasts with that in fish; anoxia accelerates the mammalian heart and elevates the blood pressure. Conflicting results have come from the injection of drugs such as cyanide and nicotine into the carotid body; some workers have reported bradycardia to be caused by stimulating the chemoreceptors in this way (Heymans, Bouckaert & Dautrebande, 1931 ; Heymans & Bouckaert, 1941). More recently, Daly & Scott (1958) have shown, in the dog, that when the hyperpnoea of anoxia is prevented by artificial positive pressure ventilation, perfusion of the carotid body with venous blood from a donor dog causes bradycardia. Moreover, in dogs spontaneously breathing room air, denervation of the lungs converts the cardiac acceleration, which is the normal response to such a perfusion, to a bradycardia, despite the hyperpnoea. They suggest that the cardio-acceleratory response to anoxia in mammals is a vagally relayed reflex from the lungs depending on hyperpnoea and masking an underlying bradycardia. The response of the dogfish may thus be regarded as the basic vertebrate response to anoxia, which in mammals is overridden by a phylogenetically more recent reflex related to aerial respiration. Such a conclusion is opposed to the prevailing assumptions outlined in the introduction to this paper.

Any attempt to evaluate the significance of this bradycardia is hindered by the dearth of quantitative data on haemo-respiratory systems of fish. The following suggestions are offered only as an indication of the direction in which, it is hoped, further work will proceed. Krogh (1941) has emphasized that the large mass and low oxygen content of water as compared with air impose physical limitations on the extent to which the volume of oxygen transported to the gills can be increased. Despite the existence of reflexes, as yet incompletely known, whereby the magnitude of the inspiratory effort is inversely related to the volume of water in the pharynx, this limitation in aquatic respiration must exist. Moreover, the easier pharyngeal filling during forward swimming, and the temporary interference with water-intake occasioned by food prehension and swallowing, presumably must alter pharyngeal flow. The haemorespiratory apparatus can be regarded as a mechanism by which the oxygen in solution in the incoming water is transferred to combine with haemoglobin in an outflowing stream of blood. The slowing of the heart in response to a reduction in pharyngeal flow, it is suggested, serves to regulate the cardiac output, so that the quantity of blood leaving the gills is no more than is necessary to permit the haemoglobin in it to be saturated by the oxygen obtainable from the amount of water presented to the gills at that time. Dill, Edwards & Florkin (1932) have shown that the blood leaving the gills of a skate is 93 % saturated, closely approximating to that of man (95 %). It is reasonable to infer that the cardiac slowing and fall of blood pressure do indicate a fall in cardiac output, since no system of vasomotor control has been reported in fish, and the blood-pressure changes do not occur in atropinized fish. More debatable is the implication that changes in the minute volume of flow through the pharynx would result in changes in saturation of the blood leaving the gills if the cardiac output remained unchanged. If the amount of water pumped through the gills were much in excess of what was necessary to oxygenate the blood there would be a margin of safety that would make such a regulation of cardiac output superfluous. But the high coefficient of utilization of oxygen in fish (50 − 80%, van Dam, 1938; Hazelhoff & Evenhuis, 1952) suggests that this is not so, and that a reduction in the minute volume of water without a corresponding change in cardiac output would result in a diminished saturation. The only way in which this could be avoided would be by an increase in the coefficient of utilization, and this clearly has an upper limit. Moreover, the high cost in pumping work done per unit of oxygen acquired makes it, a priori, unlikely that fish normally operate much below the maximum efficiency that the respiratory system permits.

These speculations may have relevance to the well established phenomenon of respiratory dependence in fish, reviewed by Fry (1957). In teleost fish active metabolism, as opposed to vital processes, decreases as the oxygen content of the water is reduced. Ferguson (1957) has published graphs relating the swimming speed of perch to the oxygen content of their water; as this falls their activity, as indicated by swimming speed, decreases. The response occurs before tensions low enough to affect the oxygen saturation of their haemoglobin are reached ; its mechanism is unknown. If it is that a decrease in oxygen uptake by the gills reflexly results in a reduction both in cardiac output and in motor activity, a mechanism whereby the needs of the tissues for oxygen are attuned to the ability of the respiratory system to supply it is evident. Such a mechanism would be very different from that in air-breathing vertebrates, yet appropriate to the physical limitations that aquatic respiration imposes.

  1. Curves relating the cardiac and respiratory rates of Squalus acanthias L. to the minute volume of water passing through the pharynx are presented ; decreased minute volume caused respiratory and cardiac slowing.

  2. The change in heart rate was dependent on anoxia; the respiratory response was largely independent of the oxygen content of the inspired water.

  3. The respiratory and cardiac changes caused by deoxygenated water flow are described. Anoxia produced cardiac inhibition and a fall in blood pressure; respiration slowed at first and then accelerated to a level above the resting rate.

  4. Gill deafferentation distinguished a short-latency reflex cardiac inhibition from a weaker long-latency slowing of central origin. The reflex, it is suggested, was mediated by branchial chemoreceptors. The existence of the central response was confirmed by occluding the conus with a clamp.

  5. It is suggested that bradycardia in response to anoxia has a significance in relating cardiac output to the minute volume of water flow, thus ensuring adequate loading of the haemoglobin in the blood leaving the gills.

I am most indebted to the Nuffield foundation for a grant that has enabled equipment and technical assistance for this project to be purchased. In addition, thanks are due to the Medical Research Council of New Zealand who have supported this work. To Mr Vic Hansen I am perennially grateful for the supply of live dogfish. To my departmental colleagues, and in particular to Associate Professor J. R. Robinson, I am indebted for much helpful discussion.

Adams
,
W. E.
(
1958
).
The Comparative Morphology of the Carotid Body and Carotid Sinus
,
272
pp.
Springfield
:
Charles C. Thomas
.
Bethe
,
A.
(
1925
).
Atmung; Allgemeines und Vergleichendes
.
Handb. norm. path. Physiol
.
3
,
1
36
.
Black
,
E. C.
(
1951
).
Respiration in fishes
.
Univ. Toronto Stud. biol
.
59
,
91
111
.
Daly
,
M.
De
Burgh
&
Scott
,
M. J.
(
1958
).
The effects of stimulation of the carotid body chemoreceptors on heart rate in the dog
.
J. Physiol
.
144
,
48
166
.
Van Dam
,
L.
(
1938
).
On the utilization of oxygen and regulation of breathing in some aquatic animals
.
Dissertation. Grôningen
.
Dill
,
D. B.
,
Edwards
,
H. T.
&
Florkin
,
M.
(
1932
).
Properties of the blood of the skate (Raia oicillata)
.
Biol. Bull., Woodt Hole
,
62
,
23
36
.
Ferguson
,
R. G.
(
1957
).
The Physiology of Fithet
(ed.
Brown
,
M. E.
), vol.
1
, p.
39
.
New York
:
Academic Press Inc
.
Fry
,
F. E. J.
(
1957
).
The Physiology of Fithet
(ed.
Brown
,
M. E.
), vol.
1
, pp.
1
63
.
New York
:
Academic Press Inc
.
Granel
,
F.
(
1927
).
La pseudobranchie des poissons
.
Arch. Anat. micr
.
23
,
13
317
.
Hazelhoff
,
E. H.
&
Evenhuis
,
H. H.
(
1952
).
Importance of the counter current principle for oxygen uptake in fishes
.
Nature, Lond
.,
169
,
77
.
Heymans
,
C.
&
Bouckaert
,
J. J.
(
1941
).
Au sujet des influences de l’alphanicotine et de la bêtanicotine sur la respiration, la fréquence cardiaque et la pression artérielle
.
Arch. int. Pharmacodyn
.
65
,
196
205
.
Heymans
,
C.
,
Bouckaert
,
J. J.
&
Dautrebandb
,
L.
(
1931
).
Au sujet du mécanisme de la bradycardie provoquée par la nicotine, la lobéline, le cyanure le sulfure de sodium, les nitrites et la morphine, et de la bradycardie asphyxique
.
Arch. int. Pharmacodyn
.
41
,
261
89
.
Heymans
,
C.
&
Neil
,
E.
(
1958
).
Refiexogenic Areas of the Cardiovascular System
,
271
pp.
London
:
J. and A. Churchill Ltd
.
Krogh
,
A.
(
1941
).
The Comparative Physiology of Respiratory Mechanisms
,
172
pp.
Philadelphia
:
University of Pennsylvania Press
.
Lutz
,
B. R.
(
1930
).
Respiratory rhythm in the Elasmobranch Scyllium canicula
.
Biol. Bull., Woods Hole
,
59
,
179
86
.
Marshall
,
E. K.
&
Rosenfeld
,
M.
(
1936
).
Depression of respiration by oxygen
,
J. Pharmacol
.
57
,
437
57
.
Ogden
,
E.
(
1945
).
Respiratory flow in Mustelus
.
Amer. J. Physiol
.
145
,
134
9
.
Satchell
,
G. H.
(
1960
).
The reflex co-ordination of the heart beat with respiration in the dogfish
.
J. Exp. Biol
.,
37
,
719
31
.
Schmidt
,
C. F.
(
1956
).
Medical Physiology
(ed. by
Bard
,
P.
),
1421
pp.
St Louis
:
The C. V. Mosby Company
.
Schmidt
,
C. F.
&
Comroe
,
J. H.
(
1940
).
Functions of the carotid and aortic bodies
.
Physiol. Rev
.
20
.
115
57
.
Schoenlein
,
K.
&
Willem
,
V.
(
1895
).
Beobachtungen fiber Blutkreislauf und Respiration bei einigen Fischen
.
Z. Biol
.
32
,
511
47
.
*

I am indebted to Dr J. Garrick, of the Dominion Museum, Wellington, for pointing out that Squalus lebruni Vaillant, the name previously used by me for this species, is a synonym of S. acanthias L.