Recueil d'observations electro-dynamiques.

Paris: Crochard, 1822 [i.e., 1823].

A beautiful copy bound in contemporary red morocco of the definitive version of this continually evolving collection of important memoirs on electrodynamics by Ampère (1775-1836) and others over the period 1820-1823, beginning with his ‘Premier Mémoire’, the “first great memoir on electrodynamics” (DSB). “Ampère had originally intended the collection to contain all the articles published on his theory of electrodynamics since 1820, but as he prepared copy new articles on the subject continued to appear, so that the fascicles, which apparently began publication in 1821, were in a constant state of revision, with at least five versions of the collection appearing between 1821 and 1823 under different titles” (Norman). Some of the 25 pieces in the collection are published here for the first time, others appeared earlier in journals such as Arago’s Annales de Chimie et de Physique and the Journal de Physique. But even the articles that had appeared earlier are modified for the Receuil, or have additional notes by Ampère, to reflect his progress and changes in viewpoint in the intervening period. Many of the articles that are new to the present work concern Ampère’s reaction to Faraday’s first paper on electromagnetism, ‘On some new electro-magnetical motions, and on the theory of magnetism’, originally published in the 21 October 1821 issue of the Quarterly Journal of Science, which records the first conversion of electrical into mechanical energy and contains the first enunciation of the notion of a line of force. Faraday’s work on electromagnetic rotations would lead him to become the principal opponent of Ampère’s mathematically formulated explanation of electromagnetism as a manifestation of currents of electrical fluids surrounding ‘electrodynamic’ molecules. The Receuil contains the first French translation of Faraday’s paper followed by extended notes by Ampère and his brilliant student Félix Savary (1797-1841). Ampère’s reaction to Faraday’s criticisms are the subject of several of the articles in the second half of the Receuil. The collection also includes Ampère’s important response to a letter from the Dutch physicist Albert van Beek (1787-1856), in which “Ampère argued eloquently for his model, insisting that it could be used to explain not only magnetism but also chemical combination and elective affinity. In short, it was to be considered the foundation of a new theory of matter. This was one of the reasons why Ampère’s theory of electrodynamics was not immediately and universally accepted. To accept it meant to accept as well a theory of the ultimate structure of matter itself” (DSB). The volume concludes with a résumé of a paper read by Savary to the Académie des Sciences on 3 February 1823, and a letter from Ampère to Faraday, dated 18 April 1823 (which does not appear in the Table of Contents), showing that this definitive version of the Receuil was in fact published in 1823. Only three other copies of this work listed by ABPC/RBH.

Provenance: Marcel Gompel (1883-1944) (ex-libris on front paste-down – Répertoire général des ex-libris français: G1896). A Jewish professor at the Collège de France, Gompel worked in the Laboratoire d’Histoire naturelle des corps organisés from 1922 to 1940, under the direction of André Mayer. In World War II he became a hero of the French resistance and was finally tortured and executed on orders from Klaus Barbie, the chief of the Gestapo in Lyon. When Barbie came to trial, the prosecutors used Gompel’s case as a particularly clear and egregious example of his guilt of crimes against humanity. His superb library was stolen by the Nazis.

The collection opens with the ‘Premier Mémoire’ [1] (numbering as in the list of contents, below), first published in Arago’s Annales at the end of 1820. This was Ampère’s “first great memoir on electrodynamics” (DSB), representing his first response to the demonstration on 21 April 1820 by the Danish physicist Hans Christian Oersted (1777-1851) that electric currents create magnetic fields; this had been reported by François Arago (1786-1853) to an astonished Académie des Sciences on 4 September. In this memoir Ampère “demonstrated for the first time that two parallel conductors, carrying currents traveling in the same direction, attract each other; conversely, if the currents are traveling in opposite directions, they repel each other” (Sparrow, Milestones, p. 33).

The first quantitative expression for the force between current carrying conductors appeared in Ampère’s less well-known ‘Note sur les expériences électro-magnétiques’ [2], which originally appeared in the Annales des Mines. Ampère stated, without proof, that, if two infinitely small portions of electric current A and B, with intensities g and h, separated by a distance r, set at angles α and β to AB and in directions which created with AB two planes at an angle γ with each other, the action they exert on each other is

gh (sin α sin β sin γ + k cos α cos β)/r2,

where k is an unknown constant which he stated could ‘conveniently’ be taken to be zero. This last assumption was an error which significantly retarded his progress in the next two years before he stated correctly that k = − 1/2 in his article [13], published for the first time in the Receuil. This article comprised ‘notes’ on a lecture [12] delivered to the Institut in April 1822 in which he surveyed experimental work carried out by himself and others since 1821 (he also published for the first time there the words ‘electro-static’ and ‘electro-dynamic’). The full theoretical and experimental proof of the correct value of k appeared in two articles in Arago’s Annales in 1822, [19] and [20], in an article by Savary [22], and in experiments with de la Rive [17] (see below).

On 20 January 1821 Ampère performed an experiment together with César-Mansuète Despretz (1798-1863) intended to support his own theory of the interaction of electric currents against a rival theory of Jean-Baptiste Biot (1774-1862) and Félix Savart (1791-1841) presented to the Académie on 30 October 1820. This was reported in article [21], the first “experimentally based semi-axiomatic presentation of electrodynamics” (Hofmann, p. 316). A small cylindrical magnet was placed at the same distance from two perpendicular current carrying wires. The Biot-Savart theory predicted that the magnet would experience no net force; Ampère’s theory predicted that the magnet would experience a non-zero torque from the nearby currents. But when Ampère and Despretz performed the experiment the magnet did not move (p. 343). This defeat, together with illness and fatigue, caused Ampère to suspend his electrodynamical researches for several months. What little energy he could muster for electrodynamics was mainly devoted to correspondence.

According to Ampère, magnetic forces were the result of the motion of two electric fluids; permanent magnets contained these currents running in circles concentric to the axis of the magnet and in a plane perpendicular to this axis. By implication, the earth also contained currents which gave rise to its magnetism. It was not long, however, before Auguste Fresnel (1788-1827) pointed out to his friend Ampère that his theory had several difficulties, notably the fact that the supposed currents in magnets should have a heating effect which was not observed. Fresnel suggested that the electric currents circulated around each molecule, rather than around the axis of the magnet. In January 1821 Ampère publicly accepted Fresnel’s idea.

Not everyone was convinced of the identity of electricity and magnetism, however. Humphry Davy (1778-1829) expressed doubts in a letter to Ampère of 20 February 1821 [7]. Ampère’s idea of magnetism created by circulating electric currents was also in direct opposition to a theory put forward by Johann Joseph von Prechtl (1778-1854), and supported by the great Swedish chemist Jöns Jacob Berzelius (1779-1848), according to which electromagnetism was ‘transverse magnetism’ – whereas Ampère eliminated magnetism and showed how all the phenomena could be accounted for by the action of two electric fluids, Prechtl and Berzelius reduced electromagnetism to magnetic action. Berzelius expressed this view in his letter [3]; Ampère responded in a letter to Arago [4].

In April 1821 Ampère wrote to Paul Erman (1764-1851), professor of physics at the University of Berlin and perpetual secretary of Berlin’s Royal Academy, in response to Erman’s Umrisse zu den physischen verhältnissen des von Herrn Professor Oersted entdeckten elektro-chemischen Magnetismus (Berlin, 1821). Ampère declared that his electric theory of magnetism was established “as solidly as a physical theory can be, since, in only admitting it at first as a hypothesis, it serves to predict and make known in advance all the magnetic phenomena formerly known, those which M. Oersted has discovered, and the new properties whose existence in voltaic conductors I have made known. When one finds such an agreement between the facts and the hypothesis from which one started, can one recognize it merely as a simple hypothesis? Is it not, on the contrary, a truth founded on incontestable proofs?” In the same letter Ampère calmly harvested Erman’s experimental discoveries as further confirmatory evidence. “The observations described in the memoir which you have been so good as to send me are all the more new proofs of it. For, if I am not mistaken, they could all be predicted according to the theory in which magnets are considered to be assemblages of what I call electric currents” (Hofmann, pp. 277-8). Erman’s experiments influenced Ampère's investigations of induction in July 1821, in which he very nearly anticipated Faraday's landmark discovery of electromagnetic induction a decade later (see below).

Ampère again stressed the ‘identity’ of electricity and magnetism in a lecture to the Académie on 2 April 1821 [5]. He also expressed his views on the nature of magnetism in a letter to Gaspard de la Rive (1770-1834) [8]. “Perhaps in an attempt to accommodate the positivistic inclinations of some of his Parisian colleagues, or to avoid the adoption of hypotheses, Ampère normally wrote on electricity and magnetism in a phenomenological vein, eschewing noumenal questions. But there were exceptions: [an] example occurred in a letter of 15 May 1821 to the Swiss physicist Gaspard de la Rive, which was published in the recipient's journal Bibliotheque universelle. Adopting the two-fluid theory of electricity then prevalent in France, he spoke, rather in passing, of “the series of decompositions and of recompositions of the fluid formed by the reunion of the two electricities of which one regards electrical currents as composed” (p. 122). Thus at this time Ampère’s aetherian framework was based on electric current regarded as de- and recomposition of fluid(s), and magnetism construed in terms of these currents rotating around each magnetic molecule” (Grattan-Guinness, p. 927).

As far as Ampère was concerned, “The physical theory of electrodynamics was now complete. Given the concepts of the ether and the electromotive force of matter as Ampère had formulated them, all the observed effects could be explained; not only explained, but subjected to mathematical analysis. The combination was a potent one and the accuracy of Ampère’s calculations and the depths of his insight led many to embrace his theory. Ampère, however, was not satisfied with merely creating a model of electrodynamic action. By 1821 he was intoxicated by his vision and convinced that his electrodynamic molecules really existed. They must, then, also explain other areas of physics and chemistry.

“In his ‘Answer to the Letter of M. van Beck’ [i.e., van Beek] [11], published in October 1821, Ampère turned his attention once again to the problem of chemical combination … What determined whether a reaction would take place and if so, with what violence, was the electrical condition of the participating molecules. To explain the mechanism of chemical combination, Ampère had recourse to another analogy; molecules were not only like voltaic piles, but also like Leyden jars. The facts of electrochemistry proved “that the particles of substances are essentially in two opposed electrical states.” In order to preserve its electrical neutrality, each molecule, therefore, decomposed the ambient ether to attract the electricity of the opposite sign. Ampère did not say if this was why each molecule was surrounded by electric currents but his use of the Leyden jar analogy would appear to rule out this possibility. The molecule, presumably, had both an inherent electrical charge and electric currents associated with it. It was the inherent static charge that caused chemical combination; the resultant combination of the two electricities gave rise to heat and light and both the material and energy relations of reactions could be understood in terms of the same mechanism … There can be no doubt that he took his own theory seriously as a general theory of matter. Nor was he alone in this. During the 1820’s Becquerel in Paris and Auguste de la Rive (1801-73) in Geneva used the electrodynamic model in their researches in electrochemistry” (Williams, pp. 150-1).

Late in 1821, however, Ampère’s satisfaction with his theory of magnetism was seriously challenged by Faraday’s discovery of electromagnetic rotation, a development which thrust Faraday immediately into the first rank of European scientists. “In the autumn he had to face a powerful criticism from Faraday, whose paper ‘On some new electro-magnetical motions’ came out in a French translation [9] in Arago’s Annales, soon after its appearance in a London journal. A seminal paper in Faraday’s contributions to the topic, it announced that continuous rotation could occur if a pivoted cylindrical magnet moved around a fixed wire, and also if a pivoted wire moved round a fixed magnet. In October he sent to Ampère and [Jean-Nicolas-Pierre] Hachette (1769-1834) one of his pieces of apparatus, and Ampère demonstrated its working to the Académie in November.

“From the theoretical point of view, the chief challenge to Ampère’s view was Faraday’s conviction that such motions could not be explained by theories based on inter-molecular forces. Faraday’s alternative, drawn from this and other experiments, was to give preference to curved ‘lines of force’; but Ampère was anxious to preserve his own approach. Accordingly, when the translation was prepared, he had a set of appendicial notes [10] made by a new helper, Félix Savary, polytechnicien of the promotion of 1815 and thus one of Ampère’s old students, and in 1821 principally a geographer by profession. Ampère added his name to these notes to indicate his agreement with them. In his second note Savary rejected Faraday’s implicit claim in the paper that the rotatory motion could be taken as a ‘primitive fact’ in electromagnetic phenomena, and in the next note he showed how that motion could be explained in Ampère’s terms” (Grattan-Guinness, p. 928).

“In his original article describing the discovery of a continuous rotation of one extremity of a current-carrying wire around a magnet, as well as the rotation of one extremity of a magnet around a current-carrying wire, Faraday stated the following: “Having succeeded thus far, I endeavoured to make a wire and a magnet revolve on their own axis by preventing the rotation in a circle round them, but have not been able to get the slightest indications that such can be the case; nor does it, on consideration, appear probable.” Ampère, on the other hand, considered that this new kind of motion might be produced in the laboratory. He was also the first to obtain it experimentally. He communicated his discovery to the Academy of Sciences of Paris in 7 January 1822 [14]. In order to obtain continuous rotation of a magnet around its axis, Ampère initially floated it in mercury by the help of a counterweight in its lower extremity. By closing the circuit, a constant current flowed vertically downwards through the upper extremity of the magnet, leaving laterally along its lower portion and going through the mercury. When this constant current was flowing through the magnet, it rotated around its axis relative to the ground” (Assis & Chaib, p. 123). Ampère wrote to Faraday in April 1823 describing these electromagnetic rotation experiments [24].

In the letter to van Beek [11] described earlier, Ampère described an experiment, suggested by Fresnel, to decide whether in a ring of copper macroscopic currents would be induced by a nearby coil or magnet. A first trial in July 1821 produced a negative result which fitted well into Ampère’s theory of molecular currents. When he repeated the experiment with a more powerful magnet in August 1822, however, he indeed obtained an effect, and realized that this was the induction of currents by magnets. But as a consequence of his struggle with Faraday’s rotations, he concentrated on his magnetic theory. Although the positive result of the induction experiment again opened the way for both interpretations of magnetization, it did not provide any positive hint concerning which of them should be preferred. Thus Ampère declared only that the result did not refer to his theory, and decided not to pursue it further. A decade later, when Faraday again discovered electromagnetic induction and gained great publicity, Ampère bitterly complained about his former disregard of the result.

Between 1821 and 1822, Gaspard de la Rive, van Beek and Faraday performed some experiments showing that the poles of a cylindrical magnet are not located exactly at the extremities of the magnet, as was predicted by Ampère’s theory. These experiments forced Ampère to modify his conception of microscopic currents. In a letter addressed to Gaspard de la Rive, dated 12 June 1822 [15], Ampère included [a figure which] presents the equilibrium configuration of the microscopic currents around the particles of the magnet, due to the interaction of all microscopic currents. That is, due to the collective interactions between the small current-carrying loops, the planes of these molecular currents should no longer remain orthogonal to its magnetic axis … This final conception of molecular currents presented by Ampère, with their planes inclined relative to the axis of an uniformly magnetized bar, is accepted in its essence up to the present time” (Assis & Chaib, p. 105).

As described earlier, Ampère had concluded in his article [13] that the constant k in his law for the force between current carting wires should be equal to −1/2. This implied, however, that two collinear and parallel current elements should repel one another when both currents flowed in the same direction towards the same point in space. Sceptical about this prediction, he performed with Auguste de la Rive, in September 1822, in Geneva, an experiment to test it [reported in [17], pp. 284-5] … This experiment has received several names in the literature: “Ampère’s floating wire experiment”, “Ampère’s hairpin experiment” and “Ampère’s bridge experiment.” Ampère himself gave a very clear description: “Two very interesting electro-magnetic experiments have lately been made by M. Ampère, in the laboratory of M. de la Rive at Geneva. M. Ampère had been induced, from his mathematical investigations, to expect a repulsion between two portions of an electrical current passing in the same direction, and in the same right line, or that every part of an electrical current would repel the other parts, a result which may be comprehended by conceiving an endeavour in the current to elongate itself. The experiment which M. Ampère has contrived to illustrate this action of the current consisted of dividing a dish into two parts by a division across the middle, and filling each division with mercury, a piece of wire was then bent into the form of the letter U, but the curved part was bent to one side, so that the two limbs of the wire might lie on the mercury one on each cell, and the bent part pass over the division without touching it. The wire was covered with silk, except a small portion at each extremity, by which the communication was established with the mercury” (Assis & Chaib, p. 145). “Ampere and Auguste de La Rive reported that as soon as a current was sent through the circuit, and regardless of the direction of this current, the originally stationary floating wire was propelled across the mercury pool away from the terminals connected to the power source. Ampere immediately attributed this phenomenon to repulsive forces between collinear pairs of current elements, that is, pairs in which one member is an element of the current in the mercury flowing between the bare end of the wire and the adjacent terminal, and the other is an element of one of the linear segments of the wire. Interpreted in these terms, the experiment represented a striking confirmation of the prediction Ampère had made to the Académie three months earlier. The importance Ampère ascribed to this demonstration was promptly reflected in the way he publicized it. For example, in sharp contrast to his ambiguous and incomplete descriptions of induction, the text he composed for his verbal report to the Académie includes a thorough and accurate account of the floating-wire demonstration” (Hofmann, pp. 317-8).

In his article [22], Savary provided further support for Ampère’s conclusion that k = −1/2 by analyzing an experiment carried out in 1820 by the chemists Joseph Louis Gay-Lussac (1778-1850) and Jean-Joseph Welter (1763-1852). “Initially they utilized an unmagnetized steel ring which did not interact with a compass needle. If this ring was broken into pieces, its pieces also had no influence upon the magnetized needle. They then coiled a toroidal helix around this ring and a constant current flowed through it. The current was then turned off and the helix was removed out of the ring. The ring did not interact with a compass needle placed nearby. However, when the ring was broken into pieces, each piece did now interact with the magnetized needle. Each piece behaved now as a small magnet. That is, each small piece of the ring was magnetically polarized with a North and a South pole, so that it became magnetized” (Assis & Chaib, p. 149). Savary showed that the results of this experiment were possible only if k = 1 or −1/2, and as previous experiments by Ampère had shown that k could not be positive he could conclude that k = −1/2. “Savary’s contribution was well publicized by Ampère. He wrote several complimentary reviews for influential journals and wrote to la Rive that Savary’s presentation of his work to the Académie marked “a kind of epoch in the history of dynamic electricity”” (Hofmann, p. 321).

List of Contents (author is Ampère unless otherwise stated):

  1. Premier Mémoire. De 1’Action exercée sur un courant électrique, par un autre courant, le globe terrestre ou un aimant, pp. 3–68
  2. [AMPÈRE & Gillet de LAUMONT] Additions au mémoire précédent – note sur les expériences électro-magnétiques de MM. Oersted, Ampère, Arago et Biot, pp. 6992
  3. [BERZELIUS] Lettre à M. Berthollet sur l’État magnétique des corps qui transmettent un courant d’électricite, pp. 9399
  1. Lettre de M. Ampère à M. Arago, pp. 99–108
  2. Notice sur les Experiences électro-magnétiques de MM. Ampère et Arago, lue à la séance publique de l’Académie royale des Sciences de Paris, le 2 avril 1821, pp. 109–112
  3. Lettre de M. Ampère à M. Erman, secrétaire de 1'Académie Royale de Berlin, pp. 113–120
  4. [DAVY] Extrait d’une Lettre de Sir H. Davy à Mr. Ampère, pp. 120–121
  5. Extrait d’une Lettre de Mr. Ampère au Prof. De La Rive, pp. 121–124
  6. [FARADAY] Mémoire sur les mouvemens électro-magnétiques et la théorie du magnétisme, pp. 125–158
  7. [AMPÈRE & SAVARY] Notes relatives au Mémoire de M. Faraday, pp. 158–167
  8. Réponse de M. Ampère à la Lettre de M. Van Beck [sic], sur une nouvelle Expérience électro-magnétique, pp. 169–198
  9. Exposé sommaire des nouvelles Expériences électro-magnétiques faites par différens Physiciens, depuis le mois de mars 1821, lu dans la séance publique de l’Académie royale des Sciences, le 8 avril 1822, pp. 199–206
  10. Notes sur cet exposé des nouvelles Expériences relatives aux Phénomènes produits par 1'action électrodynamique, faites depuis le mois de mars 1821, pp. 207–236
  11. Expériences relatives aux nouveaux phénomènes électro-dynamiques que j'ai obtenus au mois de decembre 1821, pp. 237–250
  12. Extrait d'une Lettre de M. Ampère au Prof. De La Rive sur des expériences électro-magnétiques, 22 June 1822, pp. 252–258
  13. De l’Action qu’exerce la Terre sur les conducteurs voltaïques, pp. 259–261
  14. [De la RIVE] Mémoire sur l’Action qu’exerce le globe terrestre sur une portion mobile du circuit voltaïque, pp. 262–286
  15. [Remarks on the preceding memoir], pp. 286–292
  16. Second Mémoire. Sur la Détermination de la formule que représente 1'action mutuelle de deux portions infiniment petites de conducteurs voltaïques, pp. 293–318
  17. Additions au Mémoire précédent. Extrait d’un Mémoire présenté à l’Académie royale des Sciences, dans la séance du 16 septembre 1822, pp. 319–324
  18. Exposé méthodique des phénomènes électrodynamiques et des lois de ces phénomènes, pp. 325–344
  19. [SAVARY] Extrait fait par M. Savary du Mémoire qu’il a lu à l’Académie royale des Sciences, le 3 fevrier 1823, pp. 345–354
  20. [Observations additionelle], pp. 354–364
  21. Extrait d’une Lettre de M. Ampère à M. Faraday (Paris, 18 avril 1823), pp. 365–378

Table, pp. ‘[357]–360’ (errata on p. ‘360’)

Errata, p. 383.

The bibliographical complexity of this work is a direct result of Ampère’s modus operandi: “His work was marked by flashes of insight, and it often happened that he would publish a paper in a journal one week, only to find the next week that he had thought of several new ideas that he felt ought to be incorporated into the paper. Since he could not change the original, he would add the revisions to the separately published reprints of the paper and even modify the revised versions later if he felt it necessary” (Norman). Our version of the Receuil is more extensive than the most complete copy owned by Norman, and is probably that alluded to in the note to item 45 in the Norman catalogue: “Another, probably later version, has been noted with additional pages 361-378, plus an additional page of errata (p. 383) and ten instead of nine plates.” This copy additionally has pp. 223-236, which are missing from the Norman copy and to which the additional plate refers.

Ekelof 819; Norman 44-45 (less complete issues); Ronalds 10; Wheeler Gift 784 (copy with 344 pages only – “The author’s classical investigations in electro-dynamics together with experimental illustrations. Also a paper by De la Rive on the action of the earth on a movable circuit carrying a current”). Assis & Chaib, Ampère’s Electrodynamics, 2015. Grattan-Guinness, Convolutions in French Mathematics, 1800-1840, 1990. Hofmann, André-Marie Ampère, 1995. Williams, Michael Faraday, 1965.



8vo (204 x 126 mm), pp. [ii], [1-3], 4-167, [1, blank], 169-250, 252-258, [1, blank], 259-378, [1], 358-360, 383, with 10 folding engraved plates (plates 1-5 signed by Adam after Girard), one small text woodcut. Contemporary red morocco gilt by Lefebvre, flat spine richly decorated and lettered in gilt, borders of covers gilt-tooled within double rules, inner gilt dentelles, all edges gilt. A very fine copy.

Item #4768

Price: $15,000.00