Secondary Logo

Journal Logo


History of anaesthesia: Who discovered the neuromuscular junction? The opposing views of Claude Bernard and Alfred Vulpian

Cousin, Marie-Thérèse

Author Information
European Journal of Anaesthesiology: January 2013 - Volume 30 - Issue 1 - p 1-4
doi: 10.1097/EJA.0b013e328357657f

History of anaesthesia

It is widely believed that, as a result of his studies on curare, Claude Bernard (1813–1878) postulated the idea of the neuromuscular junction as a specific structure on which the poison acts. Thus, Thomas, in the book Curare: its history and use,1 wrote in 1964: ‘they [C. Bernard and M. Pelouse] showed that both nerve and muscle retained their original excitability after curarisation; hence, the point of action lay in the junction between the two’. This notion was repeated in France in 1979.2 The role of Claude Bernard in identifying the neuromuscular junction as the key site of action of curare has remained largely unquestioned to the present day. This article considers the important role of one of Bernard's pupils, Alfred Vulpian (1826–1887), in the discovery of the true site of action of curare (Figs 1 and 2).


Claude Bernard's theories

Faced with the accumulation of experimental evidence from his contemporaries and an increasingly precise description of the neuromuscular junction, Claude Bernard proposed to the Académie des Sciences in 1856 that the nerve was affected next to the muscle during its final ‘ramification’.3 However, in his ‘Leçons sur les substances toxiques et médicamenteuses’, published in 1857, in the same text in which he described his watch glass experiments (Fig. 3), he continued to maintain that the nerve was paralysed and that curare ‘uncoupled it’ from the spinal cord. He based his view upon only one observation that, in a curarised frog, stimulation of the spinal roots leading to the motor nerves was without effect. In his ‘Leçons sur la physiologie et la pathologie du système nerveux’, published in 1858, he wrote again that ‘curare completely destroys the motor nerves’.4 Moreover, at a conference in 1864, he said again that ‘the motor nerve was destroyed by the poison’5 while recognising its reversibility: ‘the paralysis of this element [the nerve] which lasts as long as the curare stays in the blood and in contact with it but which ceases when the poison is eliminated’.


All his life, Claude Bernard was interested in curare and at the end of his career he was still considering new experiments. In 1872, he wrote ‘one could do a neat experiment with curare and strychnine with the aim of showing that the motor nerve becomes separated from the cord’.6 He illustrated his ideas with experimental setups, diagrams and the following headings: ‘on a curarised dog, the nerves are still excitable on muscle while the spinal cord no longer stimulates them. Thus there is spinal and not muscular discontinuity’. And further that ‘basing the experiment on the action of curare causing the death of the motor nerve ...: showing that the nerves remain excitable and the cord no longer has any action on them; thus the paralysis comes from a central nervous interruption and not from the muscle’. However, although, in 1872 in his ‘Physiologie Générale’, he represented the end of the nerve as the point of action of curare, he still wrote a few pages later: ‘it is therefore a central and not a peripheral interruption of the motor nerve that explains the paralysis of voluntary movement after curare poisoning’.7 It can be seen at this point that Claude Bernard was trapped in a maze of contradictions.

Vulpian's own researches

It was for Felix Edmé Alfred Vulpian (1826–1887) to refute Claude Bernard's hypothesis of a direct action of curare on the motor nerve. At that time, Vulpian was not the senior professor he later became, but rather a young researcher at the Flourens laboratory of the Museum d’Histoire Naturelle, who had followed assiduously the lectures given by Claude Bernard at the College de France. Vulpian gradually arrived at a different conclusion.

He carefully repeated the experiments and did not differ in his first conclusions from Claude Bernard.8,9 However, by 1861, doubtless after having repeated the watch glass experiment, he had somewhat changed his point of view: ‘some physiologists (to whose opinions I subscribe completely) think that curare does not act on the motor nerves, at least not for the greater part of their length and that the real site of action (of curare) ... is the muscle itself or one of its components’.10 He would soon be certain that the site of action was not in the muscle itself.

It was during his lectures on the physiology of the nervous system in 1864, and published in 1866, that he distanced himself from the accepted teaching and declared with authority that ‘curare interrupts the communication between the motor nerve fibres; ... it blocks the action of the motor nerve fibres but leaves their physiological properties intact .... Thus curare appears to act upon the peripheral extremity of motor nerves’.11 However, some lines afterwards he asked: ‘does curare really act upon these extremities? ... other researchers, myself included, think that curare does not act at all on the motor nerve fibres themselves. ... We believe that curare acts by interrupting the communication between the nerve and muscle fibres’. Curare could act on the motor end-plate on condition that the action did not involve a sort of expansion of the nerve fibre, as described at that time, but by acting on a different structure. Furthermore, in an article describing a new experiment, he spoke of ‘peripheral modifications beyond the nerve’.12

In 1875, in his work Leçons sur les médicaments et substances toxiques,13 published in 1877 and again in 1882, he led his students step by step towards the concept of the neuromuscular junction, starting from the frog model of Claude Bernard up to the watch glass experiments. With the same assuredness that he had already shown in his treatise on the nervous system, he noted in the title page of the fifth lesson (pp. 220): ‘the action of curare does not take place in reality on the ends of the nerves themselves but rather on the intimate points of contact of the terminals and the fascicules of the muscle’. If this were the case, what could be the nature of the connection? He had written in the fifth page of the foreword of the book: ‘what is the nature of this intimate connection between the terminal fibres of the motor end-plate and the substance of the muscle fibres? Is there a connection through an intermediary substance (substance unissante). Is there a simple contact? Is there another possibility which has so far not been considered?’

Further on, in that fifth lesson, making reference to the anatomists, he noted (pp. 240): ‘should not we consider that curare is acting in a special way on the terminal motor end-plate? ... Actually, we know nothing. What is certain is that these end-plates, in a curarised animal, do not show any histological change that can be detected with our available research methods’. Vulpian, being sure of his hypothesis, went so far as to say (pp. 265) that ‘but this negative result is of little value: modifications, whether molecular or chemical, may exist and could be sufficient to impede the function of one or other anatomical structures which neither the microscope or any other available research method at the present time can show. ... The action being produced is a physiological interruption’.

The controversy between Vulpian and Bernard

It is evident that Vulpian did not invent the concept of the motor end-plate, the morphology of which had been described in 1862 by several German and French histologists including Willy Kühne and Charles Rouget. Both Claude Bernard and Vulpian knew of these works, particularly that of Kühne, which had been performed in Bernard's laboratory, and it had been reported to the Académie des Sciences in 1862 by none other than Bernard himself.14 Charles Rouget, a French histophysiologist, had presented several communications on the subject between 1862 and 1866 and had even shown the academicians photographic plates (no doubt the first microscopic photographs in the history of the neuromuscular junction).

Elsewhere, Bernard had written that ‘curare acts on the peripheral extremity of the nerve, perhaps on the motor end-plate’ (pp. 230).8 However, in his own works, Bernard had not realised the implications of his claim. Further, in his final works, he distanced himself completely from the concept of the neuromuscular junction and refuted the hypothesis of Vulpian: ‘Vulpian does not believe that curare paralyses the motor nerve more than the sensory one. His opinion is that the poison acts on the muscle or some ‘intermediary structure’ (pp. 237).8 Claude Bernard based his view upon two observations: that with small doses of curare only reflex movements were abolished, whereas galvanic stimulation directly on the nerve still caused contraction of the muscles; also, as already said, that the stimulation of the spinal roots leading to the motor nerve had been, on one occasion, without effect. He thus concluded, as we have seen, that the reflex arc was interrupted at the level of the spinal cord and that this was the point of action of curare. The nerve was ‘uncoupled’ from the cord – he said ‘unhooked’ (pp. 193, 199, 218, 323–324).7

Of the ‘unhooking’ concept, so dear to the heart of Claude Bernard, Vulpian had responded that ‘if there is no mistake, it is an exceptional fact (pp. 248)’.13 One cannot however accept the position of M. Bernard that the motor nerve fibres (in the animal model) are, as a result, unhooked physiologically from the spinal cord...’ (pp. 255).13 Taking up this position, opposed as it was to that of Claude Bernard, required a certain amount of courage on the part of Vulpian, given the authority of the master. In 1870, he claimed it was his ‘duty’ to support the work ‘which, it seems to me, must lead to rejection of the explanation proposed by Claude Bernard’, but as a precaution he put forward his ideas after having described the details of the experiment on which he was basing his position.

It was in 1875 that he actually wrote ‘one cannot conclude with Claude Bernard ...’. In 1880, 2 years after the death of the great physiologist, Vulpian stated in the foreword of the later edition of his Leçons: ‘curare is (according to M. Bernard) a nerve poison, that is to say that it actually kills nerves. This position is more than just inaccurate; it should be clearly said to be wrong’ (foreword pp. IV).13


The reputation of Claude Bernard was sufficiently great not to be tarnished by this reminder of his wrong position at the end of his life. If he had lived longer and had the time to make the planned experiments, perhaps he would have realised his errors. It is now our duty to give the credit to Vulpian who, although less illustrious, was all the more rigorous in uniting the knowledge of his period in both histology and biology and in first suggesting the motor end-plate as a special site of the action of curare. This was a brilliant deduction. Gradually the structure of the end-plate would become clearer. At the beginning of the 20th century, acetylcholine was recognised as the chemical neurotransmitter in autonomic nervous system synapses and in 1936, Dale et al.15 showed that acetylcholine is also present in the neuromuscular junction. In the middle of the century, Katz and Miledi16 demonstrated the distinct membrane potentials of nerve, end-plate and muscle and the latency between their activations17 which can be explained by the cleft between the structures. The electron microscope had shown the end-plate and the synaptic cleft where the acetylcholine receptors lie. Finally, the receptor was chemically isolated in 1970.18 The coupled transmitter–receptor is just this intermediary substance, this substance ‘unissante’ (joining), evoked by Vulpian.


Assistance with the Editorial: thanks to David Baker for translation and to J. Pierre Haberer for reviewing the manuscript.

Financial support and sponsorship: none declared.

Conflict of interest: none declared.

Comment from the Editor: this Editorial was checked and accepted by the Editors, but was not sent for peer review.


1. Thomas KB. Curare: its history and usage. London:Pitman Medical Publishing Co.; 1964.
2. Baumann J, Desmonts JM. L’anesthésie. Collection ‘Que sais-je ?’ Paris, France: PUF ; 1979.
3. Bernard C. Physiological analysis of the muscle and nerves properties by means of curare. CR Acad Sci 1856; 43:825–829.
4. Bernard C. Leçons sur la physiologie et la pathologie du système nerveux. 1858; Paris, France:Baillière, 199.
5. Bernard C. Le curare. La science expérimentale. 1878; Paris, France:Baillière, 236–315.
6. Bernard C. Notes détachées. 1937; Paris, France:Baillière, 193.
7. Bernard C. De la Physiologie générale. Paris, France: Hachette, 1872. pp. 230, note 3.
8. Vulpian A. Some experiences with curare. CR Soc Biol 1853; 2:73–75.
9. Vulpian A. Physiological experiences on animals poisened with curare and ventilated. CR Soc Biol 1858; 5:33–36.
10. Vulpian A. About the duration of nerves and spinal cord properties. Gaz Hebd Méd-Chi 1861; 8:350–354.
11. Vulpian A. Leçons sur la physiologie générale et comparée du système nerveux. Paris, France: Doin; 1866. pp. 207–210.
12. Vulpian A. Experience about the physiological action of curare. Archives de Physiologie Normale et Pathologique 1870; 1:171–172.
13. Vulpian A. Leçons sur l’action physiologique des substances médicamenteuses et toxiques. Paris, France:Doin; 1882.
14. Kühne W. About the ending of the motor nerves in muscles of some higher animals and in man (note presented by Claude Bernard). Compte-rendus de l’Académie des Sciences (Paris) 1864; 58:1025–1027.
15. Dale HH, Feldberg W, Vogt M. Release of acetylcholine at voluntary motor nerve endings. J Physiol 1936; 86:353–380.
16. Katz B, Miledi R. Propagation of electric activity in motor end terminals. Proc R Soc Lond B Biol Sci 1965; 161:453–482.
17. Katz B, Miledi R. The measurement of synaptic delay and the time course of acetylcholine release at the neuromuscular junction. Proc R Soc Lond B Biol Sci 1965; 161:483–495.
18. Changeux JP, Kasai M, Lee CY. The use of a snake venom toxin to characterize the cholinergic receptor protein. Proc Natl Acad Sci U S A 1970; 67:1241–1247.
© 2013 European Society of Anaesthesiology