Principles of General Physiology

It is somewhat difficult to state what is actually the function of the cell body of the neurone, apart from being the meeting place of fibres from various other neurones and thus allowing for the connection of a number of afferent arcs with the same " final common path." It seems that it must act in reinforcing impulses which might be too weak to set up a disturbance in a fresh neurone. The refractory period, no doubt, plays a part and we must also suppose that changes in the cell body are able to prevent the reception of impulses coining from sources extraneous to those connected with the particular act in which the neurone is engaged at a given moment. In certain cases it seems that it is not necessary that the impulses should pass through the actual cell body itself. When this lies, as it were, on a side branch of the nerve fibre, the continuation of this fibre may branch and act as dendrites to form connection with another neurone. This fact has only been clearly shown for one case, namely, that of the crab, in which the cell bodies lie on the surface of the ganglion mass, and Bethe (1897 and 1898) has succeeded in cutting them off, leaving the reflex to be conveyed through the neuropile. After a time the reflex disappears, presumably because the trophic action of the cell body with its nucleus has gone.

A less convincing experiment of the same kind was made by Steinach (1899) on the dorsal root ganglia of the frog. By separating these ganglia from their blood supply, it was found that the cells degenerated after about fourteen days, but that the sensory impulses were still transmitted through the ganglia. If these results apply to neurones in general, it must be admitted that the actual cell substance itself has very little function, except that of nutrition, and the main physiological activities must be relegated to the synapses.

The cell bodies have the usual needs of cells undergoing metabolic changes, especially the need for oxygen. The question of the respiratory exchange of the central nervous system has not been thoroughly investigated. Leonard Hill and Nabarro (1895) found the oxygen consumption and the carbon dioxide evolution to be considerably less than that of muscle, as would be expected. The rate of the blood current was not taken especially into account in these experiments, so that it is difficult to compare the absolute respiration of the brain with that of other organs. The dark colour of the blood in the cerebral veins suggests fairly considerable consumption of oxygen and the immediate loss of consciousness, produced in man by cessation of the normal supply of arterial blood, shows the dependence on constant supply of oxygen. If the actual consumption is small, this latter fact shows that it must be present with a high tension in order to act efficiently.

In the work of Alexander and Cserna (1913) the rate of flow was determined and the conclusion was arrived at that the small values of Hill and Nabarro were due to the animals being narcotised. If allowed to escape from the influence of deep anaesthesia, the oxygen consumption was found to be very considerable, namely, 0'36 c.c. per gram per minute. If this is not a misprint for 0*036 c.c., it seems to throw some doubt on the accuracy of the method used, since Barcroft and Dixon only obtained O'Oll for the heart muscle and Barcroft 0'028 for the salivary glands (see Bancroft's article, 1908, p. 757).

It has not been possible to prove satisfactorily the existence of any other metabolic change in nervous tissue. Of course, the products of the breaking down of lecithin and other constituents might be found in the blood, or elsewhere, when extensive degeneration processes are taking place. It is obvious that there must be physiological continuity between constituent units of the nerve centres in all cases. In all animals above the Coslenterate condition, however, there is not direct structural protoplasmic continuity between the neurones ; we have already spoken of the syiiaptic membrane intervening, and we shall have to discuss its properties presently. In the Coelenterates, and possibly in certain parts of the higher animals, there is a kind of nervous tissue, possessing certain properties of central nature, in which a direct anatomical connection appears to exist between the various nerve fibres and cells. This is associated with the power to conduct impulses in all directions, as we saw in the case of the swimming movements of the jelly-fish. Here the ability to perform co-ordinated movements depends upon the refractory phase, so that impulses arriving at the centre when in a state of activity are inoperative. Further details may be found in the paper by Romanes (1876) and in Bethe's book (1903, pp. 78-124).

Von Uexkiill (1909, p. 81) finds it necessary to introduce the conception of "interrupter"' or " representant " as intervening between the nerve net and the muscle fibre, whose property it is to accumulate and give out "excitation." This excitation flows from a place of higher pressure in the network to one of lower pressure, so that a stretched muscle becomes stimulated and vice versa (see also p. 135 of von Uexkiill's book). In higher invertebrates, such as the earthworm, the neuropile of the central nervous system has been stated to consist of a network of the processes of the afferent and efferent cells. It is a matter of much difficulty to be certain as to whether there is direct anatomical continuity, and, in fact, Retzius (1892, p. 14) says that connection is by contact.

Even in vertebrate animals, we find in peripheral parts nervous interlacings which appear to be networks. A figure of such an arrangement in the palate of the frog is to be found on p. 79 of Bethe's book (1903). Similar structures exist on the walls of blood vessels, and in connection with smooth muscle generally (see Fig. 109, page 402, after Retzius, 1892). With regard to these "networks," although there are local thickenings to be seen in stained preparations, especially at places where branches are given off, there is no evidence that they possess the properties of centres, such as that of reflex action. Moreover, if the figures given by Retzius (1892) be looked at carefully, the impression is given that there is no anastomosis between branches of different nerve fibres. At the same time, in such cases as those of the vasornotor nerves, there is no necessity for the separate stimulation of different muscle cells, as in the delicate adjustments of voluntary

muscle ; a considerable number of smooth muscle fibres are required to be in action at the same time. But, even if they exist, it is obvious that such networks are not the same thing as the nerve centres of the lower invertebrates. An exception must, perhaps, be made in the case of Auerbach's plexus in the alimentary canal. This does actually possess reflex functions as we have seen (page 367), and ganglion cells are readily to be detected in it. This system, however, seems to possess the properties of a synaptic system, such as excitation and inhibition in definite directions, rather than the indifference of the typical network.

It is of interest in this connection that Meek (1911) has found that the plexus connections are regenerated in about four months after t^ansection of the intestine. The muscular and epithelial regeneration is completed considerably earlier than the capacity of transmitting a wave of inhibition, thus giving additional evidence that the " my enteric reflex" is of nervous origin. Hofmann (1907), as the result of extensive and detailed investigation of certain peripheral networks, especially those in the heart muscle and those innervating the chromatophores of the Cephalopods, comes to the following conclusions, which he finds to agree with the observations made by other workers, with regard to the smooth muscle of vertebrates in general. The nerve bundles supplying such muscular structures form a primary plexus by branching and by division of the coarse nerve fibres contained in them. This plexus is, as regards the direction of its fibres, independent of that of the muscular strands, and often passes transversely across them. These fibres do not anastomose with each other. In fact, Hofmann had previously shown (1904) that each nerve fibre has its own area of distribution and that there is no spreading of excitation, such as we meet with in the Medusae. From this plexus, again, fibres are given off which form another, " terminal," plexus, whose single fibres run close beside the muscle cells, not forming definite " endings " in or upon them, but finally looping backwards and joining other branches of the same fibre, or perhaps other fibres of the same nerve, although it was not possible to decide whether the latter was really the case and whether a continuous network of the branches of the same nerve bundle may exist. The possibility of fine short fibrils passing to the muscle substance from the loops is not excluded by these results and it does not seem unlikely that there is some further connection between the nerve fibre and the muscle cell than the mere contiguity of the loops. There is nothing of the nature of ganglion cells present in these plexuses. What have been taken for them can be seen in good preparations to be separate from the nerve fibres and appear to be nuclei of connective tissue sheaths to the nerves.

The importance of these facts with respect to the heart will be seen later. Peripheral networks of the kind described above clearly serve only for conduction and not for origination of nervous impulses, nor as reflex centres. Owing to the fact that nerve fibres conduct in both directions, it will be seen that, if one branch of a nerve fibre, which has divided, is put into excitation by some means, the impulses will spread over all the other branches of the same fibre and to any other fibres with which these may be connected in a network. We have, in such a case, an " axone reflex" as defined by Langley (see page 425). Excitation must spread to any effector cells supplied by any of the fibres. I refer to this fact here on account of certain curious phenomena connected with the vaso-dilator mechanism of the skin. When the peripheral ends of dorsal roots are stimulated, the region which has its sensory innervation in the particular root stimulated shows that its arterioles have dilated. I was able to show (1901, 2) that the nerve fibres concerned have the same kind of anatomical connections, and show the same degeneration relations, as the ordinary afferent fibres. It is well known that certain substances, such as mustard, placed on the skin, produce inflammation there. Spiess (1906) observed that such inflammation does not occur if the area of skin is anaesthetised, as by cocaine, and thought that the inflammatory process was brought about by a reflex from the spinal cord to vaso-dilator nerves. Ninian Bruce (1910), however, found that the inflammation was still produced after mere section of the dorsal roots, so that it could not be a spinal reflex. It was not present, on the other hand, if sufficient time had been allowed for the

nerve fibres to degenerate. The vaso-dilatation produced by mustard depends 011 the integrity of sensory nerve fibres, although not an ordinary reflex. There seems to be no other explanation except that given by Bruce himself, namely, that the phenomenon is of the nature of an axone reflex. The sensory fibres must be supposed to branch at the periphery, one part supplying receptors in the skin, the other supplying the muscular coat of arterioles and acting there as efferent inhibitory fibres. When the receptor branch is stimulated, excitation spreads to the main fibre, and back along the vascular branch to the blood vessels. It may be that this latter enters a peripheral network, and it is not impossible that the particular receptors stimulated by the irritant substance may also form some part of this network (see the diagram in Fig. 145).

The anatomical unit of the higher central nervous systems is, as we have seen, the neurone. Perhaps the clearest proof of the structural discontinuity of the individual neurones is afforded by the fact that the degeneration which takes place in a nerve fibre, when it is cut off from the rest of its neurone, only proceeds as far as its contact (synapse) with the processes (dendrites) of another neurone. Although physiological continuity must exist, there is evidently an absence of

protoplasmic or nutritive continuity. As Sherrington points out (1906, p. 17), such a contact surface is of great functional importance since " it might restrain diffusion, back up osmotic pressure, restrict the movement of ions, accumulate electric charges, support a double electric layer, FIG. 145. DIAGRAM OF REFLEX AKTIDROMIC VASCULAR alter in shape and surface DILATATION AS AXONE REFLEX. tension with changes in differ- A stimulus at A gives rise to an impulse passing along fibre B to the i-«. • <• .• i -.1

spinal cord. A branch from this fibre A is given off at C, which Qinerence OI potential With ends in the walls of the arteriole D. Stimulation of the sensory /.Vionfroc! in ^m-t'-i,-,. fnneir»n r»r end organ of the fibre at A gives rise to an impulse which passes to the arteriole along the branch C in addition to reaching the in shape, Or intervene as a spinal cord, if the main fibre B is inUct membrane between dilute solu- (Bainbridge and Menzies, " Essentials tions of electrolytes of different

of charge " ; in fact, all the numerous phenomena which the earlier chapters of the present book have shown to be of fundamental importance in the mechanism of the cell. We have also manifold possibilities of excitation and inhibition in the use of one and the same neurone in different nervous acts and the consequent advantages to the organism in economy of machinery. Some points regarding the properties of the synaptic membrane have been already alluded to, but may be recapitulated here.

Since there is no structural continuity, the possibility of actual retraction owing to increase of surface tension must be admitted. The action of electrolytes (page 218), chloroform, and strychnine (page 427) is, no doubt, exercised on the synaptic membrane, which, like other cell membranes, is, presumably, a colloidal system. The summation of a series of ineffective stimuli, so that a reflex is ultimately produced, is a common property of nerve centres. Also "facilitation," as it is called by Sherrington, in which an effective stimulus leaves the mechanism for a time capable of excitation by stimuli which were previously too weak, seems to be a further aspect of the same phenomenon. The work of Adrian and Lucas (1912, p. 121) has been already referred to.

consequences for the central nervous system, since it shows that a disturbance cannot be permanently altered in strength by passing through some region of decrement. We cannot assume that it can be made in this way too small to pass through the synapse, which must itself be looked upon as a place of decrement (see page 426). This fact shows the importance of the actual connections of a particular neurone ; in other words, the anatomy of tracts and the centres which they bring into relation with one another is of essential importance. At the same time, the effect of strychnine shows that, in the spinal cord, there is potential communication, at the least, between a receptor and all the motor neurones. A localised stimulus sets into activity the whole of the muscles of the body.

Irreciprocal Conduction. — -It seems to be a very usual property of the synaptic membrane to allow impulses to pass in one direction only. Thus Gotch and Horsley (1891, p. 485) found that stimulation of the central end of an efferent root caused no electrical change in the spinal cord above, although that of an afferent root did so. On the other hand, the discharge of a spinal centre flows, in part, backwards down the other afferent, dorsal roots. Veszi (1909), by an ingenious form of experiment, has shown that continued stimulation of a motor nerve produces no fatigue in the reflex centres ; the excitatory process does not spread inwards as far as the place where central fatigue occurs. Some further facts will be found in the next chapter (page 491).

We have seen (page 141) how permeability of a membrane to one only of the ions of a salt may allow an electrical current to pass in one direction only. Irreciprocal permeability is, therefore, a state experimentally realisable. Fatigue. — We have already seen that a motor centre may be fatigued for one reflex but remain unaffected for another (page 423). This state of fatigue is, accordingly, situated in some synapses, not in the efferent neurone itself. Excessive fatigue has been found to result in changes in the cell substance, as the experiments of Dolley, referred to on page 16 above, and of other observers, show.

In the most primitive condition, an effector may be excited directly by a receptor cell, or its prolongation, as in the sea anemone. But this arrangement cannot be called a central nervous system. Although the neurone is the anatomical unit of such systems, the reflex is the functional unit, having as its anatomical basis the reflex arc. This, in its simplest possible form, consists of at least two neurones in addition to the effector cell itself. The receptor neurone forms a synapse with the motor neurone, whose cell body is in the central nervous mass and whose axone passes out to excite an effector. Put in another way, the mechanism consists of three parts, receptive, conductive (including nerve fibre and central cell), and effective (the peripheral organ set in action). Even in the vertebrate, the cell body of the receptor neurone, although moved up to the dorsal root ganglion, as stated above, is still outside the central nervous system. But there is, in any central nervous system, except the very simplest, a synapse between at least two neurones. The economy, as well as the integrative efficiency, resulting from the use of one motor neurone by several receptors could not otherwise be obtained. This is the principle which is called by Sherrington (1906, p. 55) that of the "final common path."

The simplest kind of reflex arc is to be found in the stellar ganglion of the Cephalopod, according to the work of Frohlich (1909, 1). That this ganglion has central functions is shown by the fact that stimulation of a point of the mantle causes contraction over a wide area if the ganglion is intact, but if it is removed, contraction is limited to the spot stimulated. The application of strychnine has no effect, hence the conclusion is drawn that the intermediate neurone, or its synapse, on which this alkaloid acts, is absent and the arc consists only of the receptor neurone forming a synapse with the motor neurone in the ganglion. That is, there are two neurones and one synapse.

According to the experiments of Jolly (1910), the "knee-jerk" that is, the contraction of the extensor muscles of the knee evoked by tapping the tendon below the patella, has a "synapse time" of 0'002 second, while that of the flexion reflex is 0-004 second. If the latter involves two synapses and three neurones, it appears that the knee-jerk only involves one synapse and two neurones, an afferent one from the tendon, and the motor neurone to the muscle fibre. The experiments were made by the use of the string galvanometer. By this means it is possible to record electrical changes in the afferent fibres and in the efferent fibres, when the tendon is struck, and the various times making up the total latency can be determined. The knee-jerk must, perhaps, be regarded as an exceptional form of reflex in the higher vertebrates, although it shows the possibility of an arc of two neurones only. The following measurements from Jolly's paper are of interest : —

Total latency - Afferent endings - Conduction time - Motor endings Subtracting the sum of the second, third, and fourth numbers from the total latency, we have the time spent in passing synapses, as given in the last line. FIG. 146. DIAGRAM OF THE SPINAL ARCS INVOLVED IN THE SCRATCH REFLEX IN THE DOG. Ra, Rfi, Receptive nerve paths from hairs in different regions of the dorsal skin of the left side. /•'(.'. The final common path, in this case the motor neurone to a flexor muscle of the hip.

In actual experience, there is usually to be found at least one additional, intermediate neurone, the whole of which is contained within the nerve centre. Thus the " scratch reflex " of the dog consists of the following elements (Sherrington, 1906, p. 54), (Fig. 146). 1. The receptor neurone from the skin of the back to the grey matter of a certain spinal segment in the shoulder region. This forms a synapse in the grey matter with 2. A long neurone in the spinal cord itself (proprio-spinal), which passes backwards to the segments of the hind leg. Here it forms a synapse with

3. The motor neurone, whose axone supplies a flexor muscle of the leg performing the scratching movement. There are thus three neurones and two synapses. There may, indeed, be more intraspinal neurones and synapses between (1) and (2) and between (2) and (3), but the arrangement described is the simplest possible for such a reflex as the one in question, having its receptors and effectors in parts of the body distant from one another. The motor neurone (3) is the final common path, and the rest of the arc up to it from the receptor is the afferent arc.

The complete mechanism of the scratch reflex includes, of course, also a cutaneous sense organ, which intensifies such a stimulus as the bite of a flea, for example, so as to produce a propagated disturbance in the receptor neurone ; and at the other end, the fibres of the flexor muscle, the actual effector, are a part of the mechanism. There are some important general characteristics of reflex action which will best be deferred till the following chapter.

The increase in complexity and effectiveness of the central mechanism, due to the addition of longer intermediate neurones, as well as the addition of more and more intermediate neurones to form cross connections, has been referred to above. At the anterior end of an animal, we find the gradual evolution of the "head" with its specialised and elaborate receptor system, notably that of the " distance receptors," as they are called by Sherrington. These enable the organism to be affected by occurrences which do not themselves come into actual contact with it ; such occurrences are those affecting organs sensitive to light or to sound.

In connection with these distance receptors, the highest part of the central nervous system, the " brain," is developed. This part of the system has control over all the rest. In ourselves, we know that it is associated, in some way, with consciousness. So far is it the case that the brain is to be regarded as the central ganglion of the distance receptors, that, as Langendorff has pointed out (Sherrington, 1906, p. 349), a blinded frog is like one with its cerebral hemispheres removed ; a shark without olfactory lobes behaves as if it had lost its fore-brain.

Now Pavlov has worked out a method, that of " conditioned reflexes" to be described in the next chapter, by which it is possible to investigate the mechanism of the highest centres without appeal to consciousness. Results of great value are to be obtained by this method. Pavlov himself regards the introduction of psychological modes of expression as obscuring the problems. No doubt, the physiologist is not, as such, concerned with phenomena of consciousness, and the introduction of a method in which they can be omitted is a great advance. At the same time, the phenomena of the higher sense organs necessitate the use of methods of investigation involving the introduction of consciousness, although it is used as an indicator only.

The methods of comparative psychology may also be referred to as illustrating the possibility of investigation of the higher functions of the nervous system without the necessary introduction of consciousness. The book by Margaret Washburn on "The Animal Mind" (1909) may be consulted by those interested. A description, in any detail, of the functions of the higher parts of the central nervous system would require much more space than can be given in a book with the scope of the present one. Fig. 147, compared with Fig. 143 (page 468), may serve to give some idea of the kind of connections that are found in the higher nerve centres. There are, however, a few facts of general interest that may with advantage be given here.