ATTEMPTS TO MEASURE ANNUAL STELLAR PARALLAX HOOKE TO BESSEL by MARI ELEN WYN WILLIAMS A thesis submitted for the degree of Doctor of Philosophy of the University of London Department of Humanities Imperial College August 19 81 ABSTRACT One of the most obvious consequences of the Copernican theory is that the stars should appear to have shifted slightly when observed from the Earth at opposite ends of its orbit around the Sun. The angle through which the star appears to move is a measure of its annual parallax, a knowledge of which enables astronomers to calculate stellar distances. Parallactic angles, even of the nearest stars, are very small and despite many claims to have measured such an angle, three hundred years elapsed between the publication of Copernicusfs theory in 1543 and the production of acceptable values of annual stellar parallax in the 1830s. In this thesis I trace the development of interest in the problem of the detection of parallax through the period between HookeTs pioneering attempt in the 1660s to design and build an instrument specifically to measure parallax, and the eventual publication by Bessel in 1838 of the widely accepted value for the parallax of the double star 61 Cygni. The production of suitable instruments was essential for the successful measurement of parallax, and I outline the main developments in telescope design. However, good instrumentation alone was not enough; in addition astronomers had to be prepared to devote their time to making very precise observations of carefully selected stars over a period of at least a year. Also, it was vital to develop effective methods of analysing the data once they had been accumulated. As the eighteenth century passed phenomena other than parallax that affect stellar positions were discovered - aberration of starlight, nutation of the EarthTs axis and proper motions - so by 1800 the problems of data reduction were considerably more complex than they had been in 1700. The thesis therefore traces the parallel developments in astronomical theory and instrumentation necessary for the eventual identification of parallax. Finally, I assess the significance.of such a measurement within both the slowly emerging new discipline of sidereal astronomy, ancj the study of astronomy as a whole. 3 A CKNOWLEDGEMENTS I wish to thank most sincerely my supervisor, Dr. M. B. Ha 11, for her unstinting help and guidance, and my gratitude is extended equally to Professor A. R. Hall for his encouragement. While carrying out the research for this thesis I have received endless help from many librarians and my thanks are extended to all. In particular, I wish to acknowledge the assistance of Mile. Alexandre at the Paris Observatory, and that of Dr. D. W. Dewhirst and Miss Jean Sanderson at the library of the Institute of Astronomy, Cambridge. A number of people have read and helpfully criticised various drafts of the thesis and to them also I am most grateful. I wish especially to thank Dr. Albert Van Helden, of the Department of History, Rice University, Houston, for reading much of my work, apd Professor G. J. Whitrow, of the Mathematics Department, Imperial College, for reading the chapter on Bessel. Throughout my research I have been supported financially by a major state studentship from the Department of Education and Science, which I gratefully acknowledge. I also wish to thank London University Central Research Fund Committee for extra financial assistance for a visit to libraries in Paris. Finally, I thank very sincerely my friends and colleagues at Imperial College, both in the former Department of History of Science and Technology, and in the new Department of Humanities, for providing such a pleasant working environment throughout my time at the college. 4 CONTENTS Page Chapter One Introduction 6 Chapter Two Early Ideas, 1660-1720 2.l^Prologue: the sixteenth and early seventeenth centuries " 16 2.2 Ilooke, Wren and J. D. Cassini 21 2.3 Wallis, Flamsteed and Jacques Cassini 27 2.4 Ole Roemer 34 2.5 Jacques Cassini and Edmond Halley 37 2.6 Conclusions 43 Chapter Three James Bradley and the eighteenth century TgapT 3.1 Introduction 47 3.2 Immediate reaction to Bradley's paper 49 3.3 The mid-century: Maskelyne, Lalande and Michell 57 3.4 William Herschel and the method of relative parallax 65 3.5 Piazzi and the end of the gap 72 Chapter Four Problems of precision in the early nineteenth century 4.1 Introduction 83 4.2 Instrumental problems: the introduction of the altazimuth 85 4.3 Problems of data reduction 96 4.4 The controversy between Brinkley and Pond 106 4.5 Conclusions 119 Chapter Five The early role of F. W. Bessel 5.1 Bessel at Lilienthal: 1806-1810 122 5.2 The study of 61 Cygni: 1810-1816 131 5 5.3 The Fundamenta Astronomiae 142 5.4 Conclusions 151 Chapter Six The successful identification of annual stellar parallax 6.1 Introduction: the 1820s 156 6.2 Developments in the study of double stars 166 6.3 Improvements in precision 177 6.4 The three accepted measurements 185 6.5 Why were they believed? 197 Chapter Seven Conclusions 207 Notes 223 Bibliography 261 Appendices 273 / Chapter One INTRODUCTION "If we are to understand the nature of heavenly bodies, it is essential to know how far away they are; only then can we calculate the sizes and energy outputs of planets, stars and galaxies, and ultimately determine the scale of the Universe itself". 1.1) One of the basic measurements upon which all calculations of astronomical distances beyond the solar system rely is that of annual stellar parallax. Today, it is for that reason that the measurement of parallax forms a sometimes tedious, but nevertheless essential, part of routine activity at astronomical observatories 12). By now the distances of the closest stars - those lying within 10 light years of the Sun - are believed to be known from measurements of parallax to an accuracy of 2%. Interest in the measurement of annual stellar parallax has a long history, stretching back at least to 1543> and the larger part of that time tells of astronomers* failure to detect the phenomenon. In the main, it is with that history of failed attempts that this thesis deals. "First it will be necessary to define the word parallax, which will be frequently used in the ensuing discourse and will be a stumbling block to those that do not well understand it". (3) Thus said Edmond Halley in a lecture reputedly delivered at Gresham College sometime later than 1679 14) he then explained that "the word is originally greek and signifies distance". Halley followed this with an astronomical definition of parallax as the angle subtended by the radius of the Earth at the centre of any celestial object; from a knowledge of this angle and the Earth's radius, simple triangulation gives us the distance of the object from the Earth. As the subject of Halley's lecture was lunar parallax this was a sufficient definition. For our purposes, however, we must extend the basic concept; we are still con- cerned with an angle subtended at a celestial object, but by annual stellar parallax we mean the angle subtended at a star by the mean radius of the Earth's annual orbit. Because of the enormous distance of even the nearest stars, although the radius of the Earth clearly must subtend an angle at them, it is immeasureably small, and a larger baseline for the triangulation is required. The only other possibility is to use the radius lor diameter) of the Earth's orbit round the Sun; either of these is in fact just large enough to subtend very small, but noticeable angles at the Sun's nearest stellar neighbours. Originally astronomers sought the angle subtended by the diameter of the orbit, but since the early nineteenth century it has been standard practice to quote parallactic angles as those subtended by the mean radius. A number of accounts of the history of attempts to detect parallax have been written, the earliest of which in fact antedates by 4 years the first successful measurements: in 1834 G. R. Fockens, a student at the University of Leyden, wrote a dissertation reviewing fully the state of the problem of parallax detection, and including a large section devoted to its history (5)» Subsequently, during the 1840s, the German astronomer C. A. F. Peters, then working at the new observatory at Pulkowa, devoted much of his time to the study of stellar parallax. The results of his observations and reductions were published in 1846 by F. G. W. Struve (6), but Peters wished to add a detailed survey of all previous attempts to detect the phenomenon; he finally published this work in 1853 (7)• By the mid-nineteenth century, therefore, two detailed chronologies of attempts to detect parallax 8 existed. Since then the history has been sketched in all the major histories of astronomy (8); but of course these accounts cannot be detailed. If anything, the importance of the problem of parallax is overstated in historians1 attempts to create a larger role for the history of 'stellar* astronomy than is justified. This is particularly true of nineteenth century histories of astronomy. During the second half of the nineteenth century the study of the stars had become as important an undertaking as that of the solar system, and those people engaged in writing the history of the subject deemed it necessary to trace histories of the two branches of astronomy of equal significance. This trend has diminished slightly during the twentieth century, but it is still present, as fnewT histories of astronomy are produced based in form on the old.