History of nonlinear oscillations theory in France (1880--1940) (Q2406121)

This interesting book complements our knowledge about nonlinear oscillations with a thorough investigation of the important contribution of French investigators to the subject which had grown into the modern Theory of Dynamical Systems. Evolution of many physical systems is described by nonlinear ordinary differential equations which, in general, cannot be solved analytically. Practical interest in nonlinear oscillations required the development of qualitative methods for their analysis which started with the pioneering work of the Frenchman Henry Poincaré (1854--1912), one of the greatest mathematicians and mathematical physicists at the end of 19th century. For a long time, significant advances in the development of the Theory of Nonlinear Oscillations from the end of the nineteenth century until the middle of the 1920s were associated with the two names, the Dutchman Balthasar van der Pol (1889--1959) and the Russian Aleksandr Andronov (1901--1952). There has been much less focus on the legacy of Poincaré and the French mathematical school; this standpoint had not been questioned until recently. The author argues that ``the numerous researches conducted in the domain of oscillations in France and around the world during this time period have never been the subject of an in-depth study. Until now the historiography has primarily been focused on \textit{B. van der Pool}'s contribution entitled: ``On ``relaxation-oscillations'''' [Philos. Mag., VII. Ser. 2, 978--992 (1926; JFM 52.0450.05)].''\ The author acknowledges the importance of the paper published originally in Dutch on several levels: ``Firstly, the discovery of the relaxation oscillations produced by a triode; secondly, the equation of the phenomenon; and, lastly, the theoretical formalization linked to Poincaré's work (1881--1886) most significantly: the ``limit cycles theory.'' \ However, this historical representation does not accurately reflect reality and provides yet another illustration of the `` Matthew effect'': By focusing almost exclusively on van der Pol and a few of his publications dealing with the free oscillations of a triode, this historical reconstruction resulted in the partial, or even complete, overshadowing of previous studies about sustained oscillations, as well as an overestimation of van der Pol's contribution regarding the forced oscillations of a triode. Moreover, the crystallization occurring around van der Pol's article [loc. cit.] has caused a misunderstanding about his own results. The term ``relaxation oscillation''\ was indeed not introduced and defined in 1926, but rather in an article published in Dutch the previous year [\textit{B. Morin} and \textit{J.-P. Petit}, C. R. Acad. Sci., Paris, Sér. A 287, 791--794 (1978; Zbl 0407.53033)]. The analysis of \textit{B. van der Pool}'s works [``Over ``Relaxatietrillingen''. II'', Tijdschr. Ned. Radiogenoot 3, 94--105 (1927); ``Über ``Relaxationsschwingungen''. II'', Jahrbuch der drahtlosen Telegraphic und Telephonie 29, 114--118 (1927); ``Oscillations sinusoïdales et de relaxation'', Onde Électrique 9, 245--256, 293--312 (1930)] shows that the relaxation oscillation phenomenon was first observed in 1880 by Gérard-Lescuyer in a series dynamo machine, then in 1905 by Blondel in the singing arc. It will later be shown that the equation modeling for the triode oscillations was not carried out by van der Pol in 1926, or even in 1922, as Giorgio Israel suggests.'' The author contends that ``in spite of an undeniable wish to develop a mathematical theory on the relaxation oscillation as early as 1926, van der Pol did not manage to set its founding principles. By establishing a link between Poincaré's work (1881--1886, 1892) and this type of oscillation, \textit{A. Andronow} [C. R. Acad. Sci., Paris 189, 559--561 (1929; JFM 55.0882.02)] incorporated it into a broader viewpoint: self-oscillation theory or self-sustained oscillation theory.'' He further argues that van der Pole's contribution ``did not consist in the modeling of the equations for free oscillations of a triode'' but rather ``in the conceptualization of an oscillatory phenomenon possessing two distinctive time scales when given specific parameter values or, in other words, two types of evolution: one slow, one fast.'' Author's reflections on the link self-oscillations and Poicaré's limit cycles pointed by Andronov led to the study of the relevant research conducted prior to the publication of Andronov's note. A careful analysis of the contribution of French researchers Alfred Liénard (1869--1958), Élie Cartan (1869--1951), Henri Cartan (1904--2008), Paul Janet (1863--1927), and André Blondel (1863--1938) brought into focus numerous studies of another device, older than the triode. This device, known as the singing arc, played an important role in the development of the wireless telegraphy at the beginning of twentieth century. And then another important discovery was made. During the period approximately between 1900 and 1914, Poincaré published several fundamental studies on wireless telegraphy. In 1908, twenty years before the publication of Andronov's note, Poincaré lectured on the subject at the École Supérieure des Postes et Télégraphes establishing the connection between his work on the limit cycles and the differential equation describing the self-oscillations in the singing arc. This new type of oscillatory phenomenon was termed at that time as sustained oscillations, before van der Pol called it relaxation oscillation and before Andronov and Witt incorporated it in the self-oscillation theory. The author critically examined the literature sources confronting traditionally accepted views on the development of the Theory of Nonlinear Oscillations. The materials are collected in the book under review which is organized into three parts: From sustained oscillations to relaxation oscillations, From relaxation oscillations to self-oscillations, and From self-oscillations to quasi-periodic oscillations. The first part deals with the development of the foundations of the Theory of Nonlinear Oscillations during the period between 1880 and 1928, including the first triode designs and van der Pol's prototype equation. Describing how a new type of oscillatory phenomenon called sustained (maintained) oscillations was observed in three devices -- the series-dynamo machine, the singing arc, and the triode, Professor Ginoux concludes: ``while historiography credits van der Pol, it appears that André Blondel was the one who -- in 1919, a year before the Dutch scientist -- first came up with a modeling, leading him de facto to the equation for the oscillations sustained by a triode. The same year, Paul Janet showed that the series-dynamo machine, the singing arc, and the triode, were analogous devices governed by the same second-order differential equation, although it was only partially determined.'' In the second part, the first demonstrations of the emerging concepts of relaxation oscillations and self-oscillations in France at the end of the 1920s and the response to van der Pol's and Andronov's work by the French scientific community are described. Analyzing the lectures given in France by van der Pol, Professor Ginoux distinguishes three steps which led the Dutchman to the conceptualization of the relaxation oscillation phenomenon and points out a different path along which Andronov's work proceeded. To the author's regret, after the failure of French scientific community to consolidate own research around a common theme at The conférence internationale sur les oscillations non linéaires held in Paris in 1933 at the Henri Poincaré Institute, ``French mathematicians and engineers carried on in a sporadic and disordered fashion their research on nonlinear oscillations, focusing essentially on van der Pol's concept of relaxation oscillations, which slowly turned into a sort of ``paradigm''''\ discussed in detail in Chapter 8. The third part deals with the work of several mathematical schools on the development of the mathematical methods for the analysis of oscillations arising in radio engineering. The discussion includes the Poincaré-Lindstedt method, van der Pol's method, Krylov-Bogoliubov method, and the `` Van der Pol-Poincaré''\ method. The intensive work on nonlinear oscillations in applied problems spined off the subsequent research on the properties of quasi-periodic functions. The last chapter describes the efforts made by Jacques Hadamard to transmit to a new generation of mathematicians a heritage that could be lost in ``analysis of scientific memoirs,''\ a part of his courses at the Collège de France taught in 1913--1914 and 1920--1921. For this monograph, the author meticulously collected, recorded and studied author biographies, personal correspondence, research on nonlinear oscillations and dynamical systems conducted in France over the period of forty years, between 1880 and 1920. The text contains many interesting details describing experiments and mysterious devices used by researchers. The explanations are generously accompanied with diagrams and illustrations, the texts contains numerous quotations and publication extracts. The book concludes with an extensive list of references and two indices. It is a good reading for anyone interested in mathematics and science, their history and philosophical underpinnings.