OUP user menu

A HISTORY OF NERVE FUNCTIONS: FROM ANIMAL SPIRITS TO MOLECULAR MECHANISMS Sidney Ochs 2004. Cambridge: Cambridge University Press Price £65.00. ISBN 052124742X

Ian McDonald
DOI: http://dx.doi.org/10.1093/brain/awh359 227-231 First published online: 13 December 2004

Impulses good and bad

Two quite different cases seen when I was a medical student raised similar questions, and have provided points of reference for much of the work with which I have been involved over the past 45 years. The initial was in 1953 and was the first patient I saw with multiple sclerosis. I was struck by the degree of recovery possible from even a severe relapse. The second, seen in 1957, was a youth who had been accidentally shot in the thigh (a well known hazard of the antipodean shooting season). He was paralysed below the knee. However, after about a week, to the surprise of all, movement began to return at the ankle. It is obvious to us now that the deficits in these two patients and their recoveries must have been mediated by changes in the electrical properties of axons traversing the lesions. But this explanation was not always so apparent and, in the timescale of attempts to understand paralysis and recovery, is very recent. The unfolding of this story (and much more besides) is told in Sidney Ochs's A history of nerve functions: from animal spirits to molecular mechanisms.

It has been known since the Hippocratic corpus of the 5th century BC that the brain influences muscles. It is recorded there that an injury to one side of the head produces spasms of the opposite side of the body. That the influence of the brain on muscles is mediated by nerves was demonstrated by Erasistratus in Alexandria in the 3rd century BC, who in his dramatic public vivisections would silence the squealing of a pig by pinching the recurrent laryngeal nerves. In the 2nd century AD, Galen did the same and extended his observation to experiments in which he sectioned parts of the spinal cord, showing that motor and sensory function below the lesion were lost, the pattern depending on the size and location of the cut. He concluded that the function of the nerves was mediated by what he called the animal spirits, formed in the brain by a complex process that need not concern us here. They passed through the nerves to the ‘feeling and moving parts’, but how they did so he could not decide.

There matters rested for more than 1000 years, during which the Barbarians destroyed the Roman Empire. By the 5th century, the ability to read Greek was lost in the Latin West. In the centuries that followed, the Greek masters were known only through increasingly and—inevitably, given the difficulties of translation—corrupted Arabic versions of their writings. Ironically, it was a new threat to Graeco-Roman civilization which led to new understanding, as a result of a radically changed approach to knowledge.

In the 15th century, John VII Palaeologus, Emperor of Byzantium, could see that the only hope for survival of his Empire was to gain the support of the Western princes in resisting the Turks. But this was impossible so long as the Church was in schism, the Western princes of course being Catholic and Byzantium Orthodox. In an attempt to secure re-union, the Emperor attended a council of the church held at Ferrara and Florence, Italy, in 1438. He was accompanied by the great scholar John Bessarion, Archbishop of Nicea. Union was agreed, but many Greeks would not accept it. The Turks came and within 15 years Constantinople had fallen and the Byazantine Empire destroyed. Bessarion saw this coming and decided to stay in the West. He brought with him more than 600 Greek manuscripts. These he gave to the Senate of Venice, known for its liberal approach to scholarship. Here they formed the nucleus of the Marciana library and became available to the scholars of the University of Padua. For the first time in many centuries, scholars had access to uncorrupted Greek manuscripts. The result was a rebirth of classical learning, which led in Padua to the scientific renaissance which, over the next century, saw a profound change in the method of inquiry in philosophy and natural science, including medicine. Authority was questioned, and careful observation of the phenomena of nature including disease became central to the new endeavour. Experimentation in the manner of Galen was re-introduced.

Huge advances followed in the next century. The achievements of Vesalius, Falloppius (the spellings are legion and without authority) and Fabricius quickly became widely known and students came from all over Europe to study at Padua. The new approach culminated in Harvey's discovery of the circulation of the blood which in its turn was profoundly influential. The great German physiologist Albrecht von Haller commented in 1754 that ‘The publication of Dr Harvey's great discovery to the world, soon excited a spirit of emulation and empowered all the European professors of anatomy to trace the steps thereof, both in living and dead subjects…the consequences of which were very considerable anatomical discoveries…’. Amongst those discoveries was the knowledge that eventually illuminated the understanding of nerve function. Vesalius in 1543 reiterated Galen's view that nerves function by means of the animal spirits contained within them, asserting that the animal spirits were distributed by the nerves and ‘…may be regarded as the busy attendants and messengers of the brain’.

Thomas Willis, over a century later, took the same view, though unlike Vesalius he thought that the spirits were transmitted in the nerves both downwards and upwards, thereby providing the basis of movement and sensation. But there was a problem about transport. Neither Vesalius nor Willis could see evidence that the nerves were hollow. Willis thought that they were ‘like an Indian [sugar] cane’. How the spirits might be transmitted remained a mystery for them.

Isaac Newton in the early 18th century, invoking his principle of a universally distributed ether, thought that sensation was mediated by vibrations in the ether within the nerves. It was a popular view for a time and received support from many, including Richard Mead, but by the mid-18th century powerful arguments were being brought against it, notably by Haller on the grounds that nerves are soft and—since there was no tension in them—vibrations could not be transmitted effectively. David Hartley's solution was to suppose that the contents of the nerve fibres could behave like a fluid and also transmit vibrations by virtue of the presence of minute particles within them which ‘were subject to the powers of attraction and repulsion’.

Fifty years earlier, a theory combining mechanics and tubular flow had been developed by Descartes to account for reflex action. He supposed that fine fibrils passing up tubules within nerve fibres opened trapdoors in the ventricles allowing animal spirits to flow out and impinge on the pineal where perception occurred. The mobile pineal then redirected the spirits back down the specific tubules to the muscles of the stimulated part. The demonstration by Steno in 1669 that the pineal was fixed was a considerable blow to the Cartesians.

Von Haller considered carefully the nature of the agent of nerve action. He termed it non-committally the vis nervosa. Having rejected Newton's vibratory conjecture, he considered the possibility, already discussed (by Stephen Hales and Alexander Monroe II), that it might be electrical. Although, as already mentioned, he concluded that it was not, Wrisberg (in an annotation in Haller's First Lines of Physiology) concluded that it was. There was much interest in electrical phenomena in the 18th century. They had been known since the time of the ancient Greeks who named them using the Greek word for amber, the rubbing of which produced them. Haller rejected the view that the agent of nerve action was electrical, partly because of his observation that a ligature tied round a nerve gave rise to paralysis. If the ‘nervous fluid’ were electrical, it would, he thought, by-pass the ligature and cause the muscle to contract. He concluded that it was ‘watery, of a lymphatic or albuminous nature’ and that it was transmitted in tubes within the nerve fibres which he had not seen but deduced must be there. Anthony van Leeuwenhoek had claimed to have seen them, though much changed by the method of preparation; Haller seems not to have been aware of this.

A phenomenon which particularly attracted the attention of natural philosophers in the mid-18th century was animal electricity. Electric fish were known to the ancient Greeks (who recommended their use as a treatment for headache). John Hunter had demonstrated the structure of the electric organ of Torpedo, concluding that it was designed to ‘[form] manage and store the electric fluid’. By the late 18th century, the stage was set for the definitive demonstration of the electrical nature of the agent of nerve conduction and muscle contraction. Luigi Galvani's De viribus electricitatis in motu musculari commentarius (translated as Commentary on the effect of electricity on muscular motion) caused a considerable stir when it was published in 1791. His Commentary provides a delightful and charmingly domestic account of his experiments and the circumstances in which they were conducted. Most were performed on frogs. He made a convincing case for the existence of intrinsic animal electricity and showed clearly that it was identical to atmospheric electricity (as in storms) and ‘artificial’ electricity produced by rubbing amber. His answer to another of Haller's objections to electricity being the nature of the agent of nerve action—namely that electrical fluid would diffuse away through the good conducting medium provided by the tissues—was that the oily covering of the nerves visible under the microscope would ‘…prevent the effusion and dissipation of the electric fluid’. Galvani also showed that electric phenomena existed in the nerves and muscles of sheep. His nephew, Giovanni Aldini, went further and, on a visit to London in 1803, concluded that it was also present in man—as shown by the title of his book An account of the Galvanic experiments performed on the body of a malefactor executed at Newgate—January 17th 1803. An expanded version of the book including macabre illustrations of experiments on guillotined criminals was published the following year in Paris (Figure 1).

Fig. 1

Illustrations of Galvanism in a human, and in a dog. From Aldini, J. Essai: Théorique et expérimental sur le Galvanisme, Paris (1804).

Fig. 2

Compression of the spinal cord. From Charcot, J-M (1881).

But there was argument. One of Galvani's experiments involved hanging dissected frogs with a bronze hook through the spinal cord on iron railings outside his house. When the hook touched the railing the frog twitched. Alesandro Volta who had initially accepted Galvani's views about intrinsic animal electricity, later rejected them, concluding that the essential fact was the dissimilarity of the metals in this and other experiments of Galvani. Interest in Galvani's work then declined. But in the 1830s, Carlo Matteucci (a worker characterized by E. G. T. Liddell as exhibiting more industry than insight) provided a convincing demonstration of the intrinsic nature of animal electricity when he showed that it was possible to stimulate a muscle to contract when its nerve was laid on another actively contracting muscle.

An important step forward was taken in Florence in 1827 when Leopold Nobili showed that, using an astatic galvanometer, it was possible to detect a flow of current up the body of a flayed frog from muscles towards the spinal cord. It was therefore disappointing when Matteucci and Longet in 1844—using this instrument—were unable to detect a current in nerve. It soon became clear, however, that the instrument was simply not sensitive enough. The problem was rectified by du Bois-Reymond who, with a much improved instrument, detected an electrical change (which he termed the negative variation) accompanying activity.

An aspect of nervous action, which had perplexed investigators from the middle of the 18th century, was its speed. Haller guessed that it was 9000 feet in a minute, though others thought it faster. It was Hermann von Helmholtz who settled the matter in 1850 by showing convincingly that the velocity of conduction in frog nerve was 35–40 metres per second. A little later, he and Baxt found the velocity in motor nerves of man to be 35 metres per second. It was at once clear, as Liddell has pointed out, that the electricity in nerves was unlike that in a wire. New concept were needed to account for it. They evolved in the late 19th century and the first half of the 20th century, culminating in the ionic hypothesis of Hodgkin and Huxley published in 1952.

So much for peripheral nerve, but what of the CNS? It was implicit in much of the early writing that what happened in the periphery was also likely to happen in the CNS, though on this there was still no agreement until towards the end of the century. In 1877, Richard Caton reported in the British Medical Journal that using a galvanometer, it was possible in rabbits and monkeys to detect ‘feeble currents of varying direction … when the electrodes are placed on two points of the external surface, or one electrode on the grey matter and one on the surface of the skull’. This, in fact, was the first recording of an EEG. Conduction in the spinal cord was demonstrated in 1891 by Gotch and Horsley.

So far we have been considering what one might term good impulses—the impulses which mediate the functions of the normal nervous system. What of bad impulses—those which determine the clinical manifestations of disease in the nervous system?

Galvani, in the first sentence of the Commentary, says that his reason for publishing his observations on nerves and muscles was that ‘… we might be able more surely to heal their diseases’. In his last chapter, he reached the conclusion that paralysis is due to perturbations of animal electricity, though his interpretation of the mechanism by which they do so still depended on the humoral theory of Galen.

By the mid-19th century, with the convincing demonstration that the agent of nerve action was electrical, Galvani's conjecture seemed increasingly likely. But acceptance was not universal. Vulpian (1866), Professor of Medicine in the University of Paris, after paying tribute to Helmholtz in his Lectures on the general and comparative physiology on the nervous system, concluded that—unlike many physiologists—he was still not persuaded that the electrical phenomena of nerves and muscles and the manifestations of their activity were identical. In particular, he was concerned about the implications of the neurotropic functions of nerves which were by then fairly well known. However, Vulpian was persuaded at least that spinal cord tracts were excitable.

These uncertainties did not exist for Vulpian's younger colleague Jean-Martin Charcot, who in his classical account of disseminated sclerosis published just 2 years later says of the characteristic tremor ‘…the long persistence of the axis cylinders deprived of medullary sheathing, in the midst of the foci of sclerosis, probably play an important part here. Transmission of voluntary impulses would still proceed by means of the denuded axis cylinders, but it would be carried on irregularly, in a broken or jerky manner and would thus produce the oscillations which disturb the due execution of the voluntary movements’. Here Charcot is clearly evoking bad impulses as an explanation for bad movements. In a later lecture on compression of the spinal cord (in which he describes and illustrates demyelination at the compressed sites), he attributes the loss of function to the changes he observed at post-mortem: degeneration and demyelination. Early in the 20th century, Gordon Holmes, when pathologist at the National Hospital Queen Square, made similar observations and addressed explicitly their physiological implications. He concluded that demyelination must produce conduction block. Denny-Brown, investigating peripheral nerve injury experimentally during the Second World War, reached the same conclusion. But still in the 1950s there had been no direct recording of impulses at the site of damage, no demonstration of the way in which the impulses are bad.

Archie McIntyre, Professor of Physiology at the University of Otago at that time, appreciated the relevance that such information would have both for physiology and for understanding the pathophysiology of multiple sclerosis. He proposed a series of experiments on the PNS which revealed that normal conduction changed abruptly at the site of a focal demyelinating lesion: the prodromal positivity of the compound action potential was greatly enhanced as the negative component disappeared, signalling complete conduction block. Surviving impulses were conducted at a reduced velocity. The same was soon found to be true of demyelination in the spinal cord where it was also found that the ability of demyelinated fibres to transmit trains of impulses faithfully was impaired. In the 1980s, it was shown the remyelination in the spinal cord (as in peripheral nerve) restored good impulses that could follow each other at normal frequencies. Bostock and Sears found that persistently demyelinated peripheral fibres were capable of acquiring the ability to conduct. Indirect evidence was later adduced to show that the same is probably true in the human CNS, and that this is one of the mechanisms contributing to recovery.

Sidney Ochs provides a richly referenced background to the evolution of our knowledge of the nature of normal conduction in the PNS and muscle. He is not much concerned with the CNS (which after all behaves in general like the PNS, as was for so long assumed). But he does provide a detailed account of the history and background of his own important work on the nature and function of transport mechanisms in nerve. He also discusses the neural events underlying learning and memory and brings the book to an end with a reconsideration of the philosophical problems of sensation and perception which so concerned early scholars. The book shows signs of haste in revision and in production: Vesalius made his great contribution at Padua, not Bologna (though he did demonstrate his methods there); and it was Helmholtz not Du Bois-Reymond who made the definitive measurement of conduction velocity. There are more typographical and syntactical errors than usual. For all that, the book is an enjoyable and instructive read.

References