In ascidians the heart is a mere contractile sac open at both ends, and drives the fluid alternately in opposite directions. In snails it is a simple sac of protoplasm without differentiated nerves, but it drives the nutritive fluid in one direction. In the amphioxus there is no special heart, but only numerous contractile dilatations in the chief blood-vessels. In fishes the heart may be said to consist of three parts - the auricle, ventricle, and arterial bulb. The heart of the frog has already been described, and that of mammals requires no description.

Even the complicated mammalian heart may be regarded as a special development of the simple contractile tube endowed with the power of peristaltic contraction. The direction in which the contraction occurs is probably determined at first by slight differences in the stimuli to which the two ends of the tube are subjected, and the direction may be altered by altering the stimulus. Thus in the heart of a fish the contraction usually proceeds from the auricle to the ventricle and bulb, but by irritating the bulb the direction may be reversed so that the bulb contracts first and the auricle last, and this reversal of rhythm may persist for some time.1 In the mammalian heart it is not perhaps so easy to reverse the rhythm by simple irritation, and probably some interference with the cardiac nervous system is also requisite, but by introducing tincture of opium into the mammalian ventricle the rhythm may be reversed so that the beats of the auricle follow instead of preceding those of the ventricle.2

The cause of rhythmical pulsation in the heart is usually supposed to be the motor ganglia which it contains. Of late years numerous researches have shown that, although these are very important indeed, yet they are not to be looked upon as the exclusive originators of the rhythm. The heart of the snail, although it consists of simple protoplasm without nerves, beats rhythmically, and when a ligature is tied across the venous sinus in the frog the venae cavae and upper part of the sinus continue to beat although they possess no special ganglia, while the rest of the heart remains motionless although it contains both Bidder's and Remak's ganglia. From this experiment one would be inclined at first to say that the initiation of rhythm in the heart is due to the muscular tissue of the venae cavae and sinus,

1 Gaskell, Journ. of Physiol., vol. iv. p. 78.

2 Ludwig, Physiologie, 1861, vol. ii. p. 88.

and might be inclined to regard the nervous system of the heart as an apparatus for merely conducting- stimuli from the sinus to the auricles and ventricle.

Other experiments would seem to deprive the nerves even of this function, for Engelmannl and Gaskell have shown that when Bidder's ganglia are excised, or the nerves cut through as they traverse the auricles, contractions still pass from the venous sinus to the ventricle, and continue to do so when the nerves have not only been divided but most of the muscular tissue of the auricle has been cut through and only a narrow bridge remains behind. This may seem to prove that the muscular tissue of the heart conducts the motor stimuli from the venous sinus to the auricle and ventricle, which cause them to contract, and may appear to show that the cardiac nerves are entirely superfluous. A similar mode of reasoning, however, would lead us to say that the ganglia in medusae are also superfluous because the contractile tissue will pulsate rhythmically after they have been cut off, if it be placed in acidulated water.

In regard to the conduction of stimuli, the fact probably is that under favourable conditions they may be conveyed by the muscular tissue alone from the sinus to the ventricle, but under ordinary circumstances they are conveyed in part, at least, by the nerves.

Ganglionic tissue is more sensitive than contractile tissue, and the stimuli which act on the ganglia of the medusa, under the conditions in which it lives, are insufficient to excite contractile tissue. When the ganglia are paralysed by a poison, the effect is the same as if they were cut off, and pulsation is arrested. A similar condition appears to occur in the ventricle. The muscular tissue forming the apex of the frog's heart under ordinary circumstances will not beat when separated from the rest unless an extra stimulus be applied to it. The ventricle containing Bidder's ganglia will usually pulsate rhythmically, and if its apex be dipped in a solution of chloral no effect is produced, but if its base be dipped in the solution so that the drug acts upon the ganglia, the pulsations are arrested apparently by paralysis of the ganglia (Harnack).

We may consider, then, that ganglia are more susceptible to stimuli than muscular fibre, and have the function of making it pulsate rhythmically when it otherwise would not.

It is probable also that they serve to prevent the occurrence of blocks at the junction between the different cavities of the heart which might occur if the stimuli were transmitted from each cavity by muscular tissue alone.

When the heart is dying-, and when we may fairly assume that its nerves are losing their functional activity, such blocks actually take place, and the ventricle may beat only once for every two or three or more beats of the auricle.

The cardiac muscle is also without doubt losing its functional activity, yet it still retains it to such an extent that each cavity can contract powerfully. The same thing occurs when the heart is poisoned with chloral, iodal, or other members of the same group, which, as already mentioned, paralyse the cardiac ganglia.2 In the present state of our knowledge it is difficult to make any absolute statement regarding the function of the cardiac ganglia, but I think we may fairly assume them to have two functions, (1) to originate rhythmical pulsations in the heart when the muscular fibre alone, although capable of independent rhythmical pulsation, would not pulsate under the conditions which may be present; (2) to transmit and receive stimuli from one cavity of the heart to the other, and thus prevent the occurrence of blocks at the junction of the cavities and consequent irregular action which might occur if the stimuli were transmitted only by the muscular fibre.

1 Pfliiger's Archiv, xi. p. 465.

2 Harnack and Witkowski, Arch. f. exp. Path. und Pharm., vol. xi. p. 15.