Eclecticism, Opportunism, and the Evolution
of a New Research Agenda:

William and Margaret Huggins and the
Origins of Astrophysics


Barbara J. Becker

A Dissertation submitted to The Johns Hopkins University
in conformity with the requirements for the degree of
Doctor of Philosophy
Baltimore, Maryland

Copyright ©1993 by Barbara J. Becker
All rights reserved




"A Sudden Impulse..."

In "The New Astronomy" of 1897, William Huggins described his first encounter with prismatic analysis in this way:

It was just at this time [in 1862] that I happened to meet at a soirée of the Pharmaceutical Society, where spectroscopes were shown, my friend and neighbour, Dr. W. Allen Miller, Professor of Chemistry at King's College, who had already worked much on chemical spectroscopy.52  A sudden impulse seized me to suggest to him that we should return home together.  On our way home I told him of what was in my mind, and asked him to join me in the attempt I was about to make, to apply Kirchhoff's methods to the stars.  At first ... he hesitated as to the probability of our success.  Finally he agreed to come to my observatory on the first fine evening, for some preliminary experiments as to what we might expect to do upon the stars.53

This sounds somewhat like the story of Paul's conversion on the road to Damascus. It is well to keep in mind that Huggins was reconstructing events which had taken place some thirty-five years earlier.  The only corroborating witness was no longer alive.54

Substantiating Huggins' tale is problematic for at least two reasons:  His notebook entries are not sufficiently complete to provide independent confirmation or refutation of any published account of Huggins' investigative activities between 1860-1865; and, the subject of Kirchhoff's theory on Fraunhofer's lines -- including the suggestion that it be extended to the interpretation of stellar spectra -- was being widely discussed by January 1862.55

In this section, I shall present a brief biographical sketch of the chemist, William Allen Miller (1817-1870), followed by a discussion of opportunities Huggins may have had to become acquainted with the method of spectrum analysis. I shall argue that it is unlikely Huggins heard of spectrum analysis for the first time in January 1862.  To claim otherwise would require Huggins to have been isolated from issues of contemporary concern in the wider scientific community.  Instead, I shall suggest that Huggins' alleged January 1862 encounter with Miller would have been more catalytic than inspirational in precipitating their cooperative working relationship and stimulating the gradual development of a plan to subject starlight to prismatic analysis.

William Allen Miller (see Figure 10) was Huggins' neighbor, living across the street from him at 103 Tulse Hill Road.  According to Huggins, the experienced spectroscopist and analytic chemist agreed to come to work in his observatory "on the first fine evening" following the Pharmaceutical Society soirée.56  If so, they would have set to work sometime in January or February 1862.

Figure 10. William Allen Miller (Photograph from the Royal Astronomical Society Library, Add MS 94, No. 1).

We saw in the first chapter that Huggins recorded occasional visits by neighbors and friends to his observatory.  However, I have uncovered only a single brief reference to Miller in Huggins' notebook entries for 1862 and that is embedded in Huggins' observations of Saturn:  On 3 June 1862, Miller and his family visited Huggins' observatory in order to catch a glimpse of a shadow, possibly that of Titan, on the surface of Saturn but were disappointed.57  There is no mention of any collaborative spectroscopic work with Miller in early 1862 in any of Huggins' observatory notebooks.

Clues to when Huggins' first observations with Miller were likely made can be deduced from their first jointly published paper on stellar spectra.  The particular stars and planets Huggins and Miller described as having observed can be compared with the positions these objects would have occupied in the early evening London sky between the date of the Pharmaceutical Society's soirée (15 January 1862) and the date Huggins and Miller submitted their first report on their observations to the Royal Society (19 February 1863).58  Jupiter, for example, would have been in a good position for observation beginning in late-February of 1862 and throughout the spring and summer.  During the summer, the constellation Cygnus, and the stars in the handle of Ursa Major would have been easily visible after sunset.59  Late fall would have been a good time to examine Mars and the constellations Lyra, Andromeda, Pegasus, Cassiopeia and Triangulum.

The brighter stars like Betelgeuse in Orion, Aldebaran in Taurus, and Sirius in Canis Major, whose spectra Huggins and Miller were able to observe well enough to sketch schematically, as well as Procyon in Canis Minor, Capella in Auriga, and Castor in Gemini, whose spectra were measured only approximately, are all well-placed in the northern hemisphere during January and February.  These stars could have been observed in both 1862 and 1863.  Thus, they could have been the first and/or the last objects observed.60

But in early 1862, Miller was preoccupied with his own research problems. In his continuing efforts to photograph metallic spark spectra, Miller had found, much to his dismay, that the sensitivity of photographic plates was greater in the more refrangible, or blue-violet end of the spectrum while the prisms he was using to disperse the arclight absorbed these rays.61  He was busy searching for materials with sufficient dispersive power which were also transparent to shorter wavelengths of light.62  This was a daunting task complicated by the discovery that he was in danger of being partially forestalled on this work by George Stokes.63  Pressed in early 1862 to ready the results of his own research for publication, Miller was more likely to have been free to participate with Huggins in a spectroscopic examination of these stars at the conclusion of the period in question.

At the time of the Pharmaceutical Society's soirée in January 1862, William Allen Miller was Chair of Chemistry and a Dean at King's College, London.  He was a founding member of the Chemical Society and had recently been elected Treasurer and Vice President of the Royal Society, a position he would hold until his untimely death in 1870.  By 1862, Miller was an experienced spectroscopist and photographer, having begun his first experiments in prismatic analysis some twenty years earlier in a lumber-room located below the chemical lecture theatre at King's which he had converted into a make-shift laboratory.64  Miller had hoped that his research would reveal some relationship between the chemical properties of the substances he examined and the pattern of colorful lines observed in their flame spectra.65  But spectroscopy is a very sensitive analytic tool, far more sensitive than Miller or his colleagues were aware at the time.  Consequently, Miller's efforts were confounded by the ubiquity of sodium and its seemingly omnipresent spectral signature.66

Back in 1845, Miller had read a report at the annual meeting of the British Association describing the results of his examination of the dark lines observed when light passed through colored vapors.67  At that time, he hoped his study would aid in determining the cause of the dark lines in the solar spectrum, a problem of "much difficulty and obscurity."  Unlike his colleagues in the physical sciences who wished to resolve the theoretical problem that selective absorption of light by gases presented to then-current views on the nature of light, Miller, a chemist, wanted to use the spectral signatures observed to expedite the positive identification of chemical elements.

In his teaching, Miller emphasized methods of qualitative analysis to prepare his students in the theory and practice of modern chemistry.  One of Miller's students kept a diary of his medical school years (1860-1864) and often referred to his classes with Miller.  While the study of chemistry did not seem to be this young man's strong suit, he commented favorably on Miller's presentation of the polarization of light in October of his first term of study, marvelled at electrical demonstrations in November, and in December -- just months after Bunsen and Kirchhoff's paper on chemical spectrum analysis appeared in the Philosophical Magazine68 -- he wrote, "Professor Miller regaled us with some beautiful experiments, showing the different colours of the various metals during combustion."69  Miller's demonstration indicates nothing concerning his knowledge or acceptance of Bunsen's and Kirchhoff's spectral work, but it does show that as an empirical method of chemical analysis, flame testing had become a matter of routine at the introductory level in the universities by 1860, even with no physical theory on which to found it.

In his influential textbook, Elements of Chemistry which appeared in a second edition in 1860, Miller addressed with interest the problem of the fixed lines in the solar spectrum.  These lines, Miller noted, "are independent of the nature of the refracting medium, and occur always in the same colour, and at corresponding points of the spectrum."  While admitting their utility for opticians in determining accurately a substance's index of refraction, Miller added, "No satisfactory explanation has yet been found for the cause of this phenomenon."70  His long-standing interest in the Fraunhofer lines made Miller more receptive than most to the news of Kirchhoff's theory regarding their cause.

Miller's early spectral work resurfaced after nearly two decades of obscurity when, as part of the thinly-veiled attempt, referred to earlier, to establish Britain's priority in spectroscopic investigation, William Crookes reprinted the text of Miller's 1845 British Association report in Chemical News in May 1861.71  Crookes' action was important for two reasons.  First, it demonstrated the depth of feeling among British scientists that their fellow countrymen had made significant contributions to the development of spectrum analysis which had been unjustly ignored.  Second, by casting Miller as a representative of that aggrieved community, Crookes elevated Miller's past efforts in prismatic analysis to pioneering status and ensured him an attentive audience for his current spectroscopic work.

Crookes prefaced Miller's text with two brief introductory paragraphs in which he remarked that Miller's earlier spectral work surpassed that being done more recently, a clear reference to the work of Bunsen and Kirchhoff.  In fact, Crookes declared Miller's old spectral maps of the elements to be more accurate than Kirchhoff's thus setting off a minor priority dispute which kept Miller in the public eye.72  Henry Roscoe complained to George Stokes:

I see that Crookes the Editor of the "Chemical News" wishes to claim for W. Allen Miller the discovery of the power of luminous gases to absorb light of the same kind as they emit, but I cannot believe (from the paper as published) that Miller could think of claiming this point. His sentence on this subject is not only vague, but as I read it, positively incorrect.73

Anxious to set the record straight, Kirchhoff later submitted a paper to the Philosophical Magazine in which he purported to chronicle the history of spectrum analysis.  In it he soundly criticized Crookes' statement and called into question the value of Miller's early spectroscopic work:

I only have to remark that, by way of experiment, I have laid Professor Miller's diagrams before several persons conversant with the special spectra, requesting them to point out the drawing intended to represent the spectrum of strontium, barium, and calcium respectively, and that in no instance have the right ones been selected.74

Kirchhoff angrily chastised Crookes for wrongly stating that Miller's early spectroscopic investigations anticipated his own.75

In 1861, the British Association met in Manchester.  Miller took the opportunity afforded him as President of the Chemistry Section to recount the progress in the science of chemistry during the past year, giving special notice to Bunsen's discovery of two new elements, cesium and rubidium, by means of spectrum analysis.76  He also delivered an evening lecture on "The New Method of Spectrum Analysis" which Agnes Clerke later described as being one of "considerable historical value."77  The illustrated lecture was crowded with individuals who had come to hear about Kirchhoff's method and theory.  Instead "they were not a little surprised to find Kirchhoff occupying the end of a long series of illustrious names, from Newton in 1701 to Wollaston in 1802 and Fraunhofer in 1815."78

Perhaps prompted by this lecture's success, Miller repeated it before the Pharmaceutical Society of London at its soirée on the evening of 15 January 1862, the very soirée described by William Huggins in his retrospective.  It is interesting that in Huggins' 1897 reminiscent account there is no reference to Miller's lecture that evening.  Instead, Huggins recalled only that there had been a display of spectroscopes.79

The full text of Miller's January 1862 presentation was published both in the Pharmaceutical Journal and in the Chemical News.80  In his introductory remarks, Miller referred briefly to Fraunhofer's early examination of the spectra of several bright celestial objects.  The text of the speech provides a footnote informing readers that these celestial bodies included Venus, the moon, Sirius, Castor, Procyon, Capella and Betelgeuse.81  We cannot know whether any mention was made of these particular objects in the actual presentation.  The body of Miller's talk focussed on the spectra observed in colored flames, the power of the method for chemical analysis and disclosure of new elements, and Kirchhoff's recently completed map of the solar spectrum.  Miller concluded his talk on a note of caution:

It seems to be not unfitting at the present moment to put some ardent minds upon their guard; for it appears by some to have been hastily assumed, in a spirit of self-confidence, that we already have the key to everything upon this subject.82

He acknowledged it was possible that Kirchhoff's interpretation of spectral patterns would provide a "further glimpse into the machinery of the universe," but he believed that pursuit of these researches would ultimately teach investigators a "lesson of reverent humility."83  Miller's practical bent led him to emphasize the utility of the new method for qualitative analysis.84

Miller was not the only one giving lectures on spectrum analysis.  I have already discussed Henry Roscoe's paper given before the Chemical Society in June 1861.  Roscoe was twice invited to give popular lectures on spectrum analysis at the Royal Institution.  The first consisted of just one lecture presented on 1 March 1861, and introduced Bunsen and Kirchhoff's recently announced observations of chemical spectra with a special emphasis on the newly discovered element, cesium.85 The second was presented as a series of three lectures during March and April 186286 and would have held great interest for Huggins and Miller as the two men shaped the form and structure of their own research program.  In them, Roscoe entertained his audiences with dramatic demonstrations and enthusiastic allusions to Bunsen and Kirchhoff's contributions.  Even though he stressed a need to proceed slowly, it was clear that Roscoe's confidence in the method's promise was unshakable.

The first two lectures introduced Fraunhofer's early work on stellar spectra and described how the examination of spectra could be employed in chemical analysis.  In the third lecture, Roscoe turned to the application of spectrum analysis to the luminous bodies in the heavens.  He proposed an analogy concerning the possibilities for extending human understanding in this area of research with the marvels that could be imagined to follow the discovery of the philosopher's stone. "We have, in short," he said, "to prove that chemical analysis has now stretched forth her hand beyond the limits of our earth, and that she now affords us certain information respecting the chemical constitution of the sun and the far distant fixed stars."  He then proceeded to explicate Kirchhoff's work on unravelling the mysteries of the sun.  In his conclusion, he talked about founding a "new stellar chemistry":

As regards the chemical composition of the fixed stars, all we know is that they contain substances not found in the solar atmosphere, because their spectra exhibit totally distinct dark lines ... it appears that each fixed star possesses a spectrum peculiar to itself.87  We are, however, only on the threshold of these subjects; the possibility of obtaining an insight into the composition of stellar matter has indeed been shown, but it remains for the astronomer and the physicist to obtain for us that knowledge which we so much desire."88

Miller's talk at the Pharmaceutical Society soirée may have given Huggins the incentive or, perhaps, the opportune moment, to discuss the possibility of applying Miller's spectroscopic expertise to Huggins' astronomical interests.  It is quite easy to imagine that their subsequent collaboration grew out of a conversation the two men shared during their alleged carriage ride home that evening.  But if Huggins' first introduction to the new possibilities for analytic chemical research using spectra was through a public lecture being given by Miller in 1862, we must conclude that he did not know of Miller's address to the British Association at the previous year's meeting, neither was he aware of the discussions which had taken place during the previous year at the Chemical Society, nor had he read the papers written on the subject published in the Philosophical Magazine or the publications of the Royal Society.

Huggins' Notebook Entries:  1860 - 1865

In his 1897 retrospective essay, Huggins described the dramatic shift this new interest in spectroscopy imposed on his old ways of doing things:

Then it was that an astronomical observatory began, for the first time, to take on the appearance of a laboratory.  [See Figure 11.]  Primary batteries, giving forth noxious gases, were arranged outside one of the windows; a large induction coil stood mounted on a stand on wheels so as to follow the positions of the eye-end of the telescope, together with a battery of several Leyden jars; shelves with Bunsen burners, vacuum tubes, and bottles of chemicals, especially of specimens of pure metals, lined its walls.

The observatory became a meeting place where terrestrial chemistry was brought into direct touch with celestial chemistry....  This time was, indeed, one of strained expectation and of scientific exaltation for the astronomer, almost without parallel; for nearly every observation revealed a new fact, and almost every night's work was red-lettered by some discovery.89

Figure 11.  Interior of William Huggins' observatory (from Huggins and Huggins, Atlas, 4).

This reminiscent exuberance stands in stark contrast to Huggins' notebook account. There is nothing recorded in his first observatory notebook between 18 July 1860 and 31 March 1862.  There are no clues in his correspondence, nor in any biographical narrative to explain this long hiatus in his notebook record.  It is perplexing as well as disappointing.  Missing are any references to his first spectroscopic research with William Allen Miller.  Even if it is assumed that Huggins preserved this particular notebook for records of telescopic or visual observations, it is difficult to understand why he made no note in it of the brilliant daytime comet of 1861 (Tebbutt's) which sported a coma as large and bright as the moon and a tail between 100-120° in length.90  Neither did Huggins record the transit of Mercury on 11 November 1861.91

Huggins' notebook entries for 1862 begin in March, not with spectroscopic observations, but with descriptions and drawings of Saturn as though there had been no interruption in his observations of that planet recorded two years earlier. Between 31 March and 2 June, he recorded twelve different observations of Saturn, eight of which took place between 12 May and 21 May.

This concentrated attention on Saturn is easy to understand.  In the spring of 1862, Saturn was at opposition and its rings were seen edge-on.  The high reflectivity of Saturn's rings makes it difficult to search for objects orbiting near the outer limits of the ring system when the rings are tilted toward our view.  The axial tilt of Saturn causes the ring plane to appear edge-on as seen from earth approximately once every 14 years.  Because the rings are so thin, they virtually vanish from sight when in this orientation.  At such times, Saturn's large moon Titan can be observed to transit the planet casting its shadow on Saturn's surface. Observing and timing such events was a challenge for amateur observers, and Huggins' notebook entries show that he participated in this activity.92  But amateurs like Huggins observing the 1862 edge-on apparition of Saturn had more than transits of Titan on their minds.

During the previous edge-on apparition of Saturn in the fall of 1848, two observers, William Lassell in England, using an excellent 24-inch reflector of his own construction, and George Phillips Bond in Cambridge, Massachusetts, using a Merz 15-inch refractor, independently announced having detected the presence of what appeared to be a new moon orbiting Saturn.93  Lassell, a brewer by profession, but a serious amateur astronomer by avocation, was a veteran satellite discoverer. Just seventeen days after the first confirmed observation of Neptune in September 1846, Lassell announced the probable existence of a satellite, a claim which was verified the following year.94  In 1851, Lassell announced he had observed two small moons of Uranus -- Ariel and Umbriel.95  In just five years, one planet and four moons had been added to the sun's family, and the four moons had all been found by an amateur observer.96  In 1862, there was no doubt some hope that yet another one among them would be rewarded for his keen eyes and diligent observation with the discovery of a new moon.

That satellite discovery was a prime objective of serious amateurs during this period can be inferred from the attention given in published accounts to the appearance of bright points in the ring plane.97  William Huggins shared this interest.  His correspondence with his friend William Dawes provided him with predictions of the locations of Saturn's known satellites which permitted him readily to distinguish any new or unexpected objects that might appear.  Unfortunately, no new moons were confirmed during this favorable opposition.  Huggins' note to the RAS on the subject of his own observations of Saturn is lengthy and detailed reporting color, relative brightness, times, and other salient features of the planet's appearance.98

All of Huggins' recorded observations from August 1862 until July 1865 occupy just two sides of a single page in the notebook.  Only two observations are noted in 1863.  The first, in April, includes drawings of Saturn and Jupiter.  The second, in October, describes fluctuations observed in a barometer during the course of a storm, the only such observation ever recorded in the notebooks.99

There are a few pages of notes in another notebook, referred to as Notebook 4 in the Wellesley collection, which deserve mention at this point.  These entries were originally thought to have been records of experiments performed in 1871. But it is clear from the data in them that these were notes taken during Huggins' work with William Allen Miller on the chemical spectra of elements completed in the fall of 1863.100  The existence of these few notes relating to spectroscopic study raises the provocative question of where the others might be.  I would argue that Huggins' exclusion of these records from his log of visual and telescopic observations indicates that he viewed his spectroscopic work at that time as part of a different enterprise, a separate research agenda.101

The single entry in Notebook 1 for 1864 provides a brief verbal and pictorial description of sunspots.102  Although there is no record of it in the surviving notebooks, we know from other sources that Huggins was actively involved at that time in the on-going debate in the RAS on the physical interpretation of the mottled appearance of the sun's visible surface.  Three years earlier, James Nasmyth, a retired engineer known for his invention of the steam hammer and avid amateur astronomer, claimed he had observed a filamentary structure in the bright surface of the sun which he likened to willow leaves in appearance.  Nasmyth described these long and slender shapes as layered in a helter-skelter fashion over the entire solar surface but more clearly organized near the edges of sunspots.  The initial enthusiasm with which this claim was greeted turned to controversy when it was challenged in late 1863 by William Rutter Dawes, whom we have met in connection with Huggins' early career.  The course and nature of the controversy is well-described elsewhere.  Suffice it to say that Huggins has been credited with bringing the controversy to a close by pointing out a middle ground couched in terms upon which both sides could find some agreement.103

Huggins' few 1865 entries are dominated by calibrations of a new micrometer.104  Once again, there is justifiable concern that other records may have been made, but no longer survive.  In a letter written fourteen years later, Huggins referred to a notebook entry dated 17 March 1864 concerning a spectroscopic observation of Sirius.  The excerpt which Huggins quotes in his letter bears a vague family resemblance to several recorded in March 1868.105  Huggins occasionally confused dates and it may be that this is just another instance, but I have been unable to find an entry elsewhere which is sufficiently similar in wording to the extract he provided in his letter to identify it with certainty as the note in question.

Based on information in his own publications and passing references to his work in remarks published by others, Huggins was clearly much more active than this scant record would suggest.  Looking first at Huggins' contributions to the Monthly Notices, we find that he submitted two notes on his 1862 planetary observations,106 a report from his observatory,107 and a paper describing his observations with Miller of changes in the spectrum of a star as it was occulted by the moon.108

Huggins' contemporary published accounts of his spectroscopic research first appeared in the Proceedings of the Royal Society and the Philosophical Transactions. Some of the papers which bore his name were co-authored with Miller, who was at that time a high ranking officer in the Royal Society.  Other papers Huggins wrote himself, but were necessarily communicated to the Society by Miller as Huggins was not yet a Fellow.  All of these papers dealt with the spectral study of three principal subjects:  stars,109 terrestrial chemicals,110 and nebulae.111  The papers will be discussed in turn in a later section as we examine the probable course of Huggins' spectroscopic work from 1862 to 1865.

The earliest reference to Huggins' spectral work in any published form is to be found in the first number of the Astronomical Register.  It was simply noted that at the Society meeting held on 14 November 1862, "Mr. Huggins read a paper on the 'Stellar Spectrum.' ... The Meeting was very numerously attended, the rooms in fact being inconveniently crowded." 112  The fact that the meeting was crowded may have exposed more people to the news of this pioneering study.  On the other hand, it may simply have made it harder for everyone to hear what had been said. Unfortunately, no details of this paper are given, nor is any discussion recorded.

"to ascertain, if possible, the constituent elements of the different stars"

Priority concerns soon pressed Huggins and Miller into print.  This can be inferred from the fact that a translation of a paper by Giovanni Battista Donati on his study of Fraunhofer lines observed in stellar spectra appeared in the January 1863 number of the Monthly Notices (see Figure 12).113 

Figure 12.  Schematic drawing of the spectra of 15 stars by Giovanni Battista Donati [from Mon. Not. Roy. Astr. Soc., 23 (1863):  103].

Huggins and Miller submitted a brief report of their own work on stellar spectra in the Proceedings the very next month.114  It told of their observations of multitudes of absorption lines in the spectra of a number of selected fixed stars (see Figure 13). 

Figure 13.  Schematic drawing of the spectra of (from top to bottom) Aldebaran, Sirius, Betelgeuse, and the sun, by William Huggins and William Allen Miller [from Proc. Roy. Soc. 12 (1863):  444].

While the details concerning their apparatus were reserved for publication later in the Philosophical Transactions,115 this preliminary report contained schematic drawings showing the placement of absorption lines in the spectra of three stars -- Sirius, Betelgeuse and Aldebaran -- in comparison with those of the Sun.  Additionally, because Huggins and Miller had heard of similar work being done in America by Lewis M. Rutherfurd (see Figure 14),116 they wished it to be known that their work was not confined to the study of these three stars alone, but that they had, in fact, spent the past twelve months examining the spectra of thirty to forty other stars with this apparatus as well as those of Mars, Jupiter and the moon.117  Miller and Huggins did not refer to Donati's work in any published account of their spectral observations until their joint paper on stellar spectra appeared in the Philosophical Transactions.

Figure 14.  Schematic drawing of the spectra of 18 stars, the sun, moon, and planets by Lewis M. Rutherfurd [from Am. Jnl. Sci. 35 (1863):  74].

Figure 15.  Schematic drawing of the spectra of 19 stars and the sun produced at the Royal Observatory, Greenwich [from Mon. Not. Roy. Astr. Soc., 23 (1863):  190].

Huggins and Miller arranged a "form of apparatus" capable of bringing into view the stronger lines in the brighter stars and permitting them an opportunity to verify the earlier work of Fraunhofer and Donati.  This may have been some jerry-built arrangement which provided them with preliminary qualitative impressions. But the goal of their research was far more ambitious.  They aimed "to ascertain, if possible, the constituent elements of the different stars."  Huggins and Miller, therefore, expended considerable time and energy perfecting an apparatus capable of attachment to the observing end of a telescope with sufficient dispersion to resolve the sodium D lines and the magnesium b band to permit rigorous comparison of stellar spectra to be made against that of the sun.118  Miller's influence can be inferred in this if we recall his allusion during his Pharmaceutical Society lecture to Fraunhofer's having detected these very lines in the spectra of a few bright stars. As Huggins and Miller stated in the Proceedings paper, "After devoting considerable time to the construction of apparatus suitable to this delicate branch of inquiry, [we] have at length succeeded in contriving an arrangement which has enabled [us] to view the lines in the stellar spectra in much greater detail than has been figured or described by any previous observer."119  (See Figure 16a)

Figure 16a.  William Huggins' early stellar spectroscope [from Heinrich Schellen, Spectrum Analysis in its Application to Terrestrial Substances (Longmans, Green & Co.: London, 1872), 466].

Figure 16b.  Spectroscopic apparatus used by James Carpenter of the Royal Observatory, Greenwich [from Mon. Not. Roy. Astr. Soc. 23 (1863):  188].

A poignant description of the difficulties Huggins and Miller faced in trying to construct this apparatus occupies two full pages in Huggins' 1897 retrospective. Here he pointed to their need to find alternatives to liquid-filled prisms so that the apparatus could move freely in any plane without fear of leakage, the need for breadth in the spectrum produced by point-like stars to allow the eye to examine the dark absorption lines more easily, the need for careful alignment of both the telescopic and comparison spectra, and the need to make drastic alterations to the traditional organization of observatory space in response to the demands made by the intrusion of chemical laboratory instruments.120  Resolving these problems was a slow and gradual process.  In fact, Huggins and Miller claimed it was only complete just "before the end of the year 1862" allowing them time to make careful observations of Aldebaran, Betelgeuse and Sirius before they submitted their initial report to the Royal Society in February 1863.121  When they finished their later, more complete paper for the Philosophical Transactions, they had extended their survey of stellar spectra to include some fifty stars, but only the spectra of Aldebaran and Betelgeuse had been mapped to a level of completeness which satisfied Huggins and Miller (see Figure 17).  Sirius' being located so close to the horizon much of the time made it a more difficult object to observe.

Figure 17.  Drawing of spectra of Aldebaran (top) and Betelgeuse by Huggins and Miller [from William Huggins, On the Results of Spectrum Analysis Applied to the Heavenly Bodies (W. Ladd:  London, 1866), plate facing title page].

The most unfortunate aspect of not having a notebook record to accompany these finished papers is the loss of direct access to the active process in the development of this apparatus.  The Royal Society publications abstract the difficulties faced into euphemistic terms which mask the nitty-gritty of the work involved.  Thus Huggins and Miller "devote considerable time" to constructing their apparatus and "at length" they "succeed."  Huggins' Nineteenth Century retrospective essay is all the more striking for its attention to the messy details of a small sampling of the difficulties Huggins remembered from his and Miller's struggle to make their instruments match demands made by their observational goals.  But even here, the problems which are mentioned are the ones whose solutions really mattered in the end.  Huggins could recognize those problems in 1897.  By then, the deadends they may have pursued in the spring of 1862 were forgettable wastes of time.  Still, the difference in style between the published scientific papers Huggins co-authored with Miller in the early 1860s and the popular essay he wrote in 1897 show that he was already learning the art of converting actual experience into the carefully contrived scientific jargon which the elites of the community would accept and the members-at-large would expect.  He had had some practice doing this in his short contributions to the Monthly Notices, but now he was being tutored by a master.

[click on footnote number to return to text]

52. Although the implication in the quote as taken from context is that this is coincident with his first hearing of Bunsen and Kirchhoff's work at some time shortly after its original announcement, the soirée to which he refers took place on 15 January 1862.

53. Huggins, "The New Astronomy," 911-2.

54. A similar account crediting Huggins with initiating the collaboration was given in the obituary for Miller which appeared in the Proceedings.  This, of course, was also written after Miller's death and no source for the information is given.  It is possible that Miller, who had a personal interest in astronomy, was the one who initiated the collaboration.  [Coutts Trotter], "William Allen Miller," Proceedings of the Royal Society 19 (1870):  xix-xxvi. 

55. It should be recalled here that Giovanni Battista Donati of Florence and Lewis M. Rutherfurd of New York were already observing stellar spectra by January 1862.  Huggins would have become aware of Donati's work in January 1863, when a translation of a paper describing observations of stellar spectra done in August 1860 was published in the Monthly Notices.  Rutherfurd's work on stellar spectra began sometime after December 1861, but was not announced until January 1863, in the American Journal of Science.  Huggins later claimed to have first learned of Rutherfurd's work in February 1863.

56. Huggins, "The New Astronomy," 912.

57. William Huggins, 3 June 1862, Notebook 1.  The entry merely states "Very fine.  Dr. Miller & family.  The shadow not visible."  What shadow is meant here is unclear.  A transit of Titan had been predicted for the previous evening, but Huggins had had problems with cloud cover.  If Huggins expected Titan's shadow to cross Saturn once again on 3 June, he was far more naive and inexperienced than appears evident from his notebook entries and publications.  It is possible he was referring to the shadow of the rings themselves on Saturn's surface.  Why that would not have been visible, however, given Saturn's edge-on orientation, is inexplicable.

58. J. Norman Lockyer, "Note on Communication of a Note on the 'Lines in the Spectra of some of the Fixed Stars,'" Monthly Notices of the Royal Astronomical Society 23 (1863):  179-80; William Huggins and William Allen Miller, "Note on the Lines in the Spectra of some of the Fixed Stars," Proceedings of the Royal Society 12 (1863):  444-5.

59. The paucity of summertime objects is undoubtedly due to the late hour at which the sky becomes dark at London's latitude.

60. As note is made at this time of an attempt to capture the spectra of Sirius and Capella on a photographic plate during the course of these observations, it is likely that visual work was done during the first apparition of these stars while the photographic work was tried only after considerable experience at these difficult observations had been acquired.

61. The physical process that causes light to change direction when it passes through a prism is called "refraction."  Each wavelength (or color) of light is refracted to a different degree causing white light (which is composed of a wide range of wavelengths) to be dispersed into a colorful band with red on one end and blue/violet on the other.  Because refraction was the principal means of dispersing white light in the nineteenth century and earlier, the degree of bending experienced by an individual wavelength of light was termed its "refrangibility."  Red light, which is bent less than other wavelengths, was referred to as "less refrangible" by Huggins and his contemporaries while blue or violet light was described as "more refrangible."  Today, refraction is no longer the sole means of producing a spectrum and it is more common to refer simply to the red end or the blue end of the spectrum.  Nevertheless, in keeping with the usage of Huggins's day, I shall describe the extremes of the spectrum as being more or less refrangible.

62. See, William Allen Miller, "On the Photographic Transparency of various Bodies, and on the Photographic Effects of Metallic and other Spectra obtained by means of the Electric Spark," Proceedings of the Royal Society 12 (1862):  159-66. This appeared in its full form in Philosophical Transactions 152 (1862):  861-87.

63. Stokes had developed a method of observing the spectrum of an electric spark using a fluorescent material in lieu of photography.  See George G. Stokes, "On the Long Spectrum of Electric Light," Proceedings of the Royal Society 12 (1862):  166-8.  That he was working on this in complete ignorance of Miller's investigation is clear from a letter to Henry Roscoe, 7 February 1862, Add MS.7656.R795, Stokes Papers, Cambridge University Library.  After finding out about their overlapping efforts, the two men exchanged letters in which they discussed the nature of their work.  See William Allen Miller to George Stokes, 25 April 1862, Add MS. 7656.M533 and 29 April 1862, Add MS. 7656.M535, Stokes Papers, Cambridge University Library.  It is interesting that in these letters, Miller makes no reference to any concurrent work on stellar spectra.

63. Although this is reported in his obituary in the Proceedings of the Royal Society [(1879) 19:  xix-xxvi] and repeated in Agnes Clerke's essay on Miller in the 1894 edition of the Dictionary of National Biography ( 37: 429-30), there is no record at King's College as to where this room might have been located or how it might have been altered to make such research possible.  On a visit to the College, I was fortunate to have been provided with a tour of the depths of what comprised King's old building by Professor D. I. Davies who pointed out rooms which he conjectures may have been those appropriated by Miller for his spectroscopic experiments.

65. See, for example, William Allen Miller, "Experiments and observations on some cases of lines in the prismatic spectrum produced by the passage of light through coloured vapours and gases, and from certain coloured flames," Philosophical Magazine 27 (1845):  81-91; idem., "On action of gases on the prismatic spectrum," The Chemist 6 (1845):  404.

66. For a discussion of the interpretive problems posed by the abundance of sodium in samples used by early spectroscopists, see McGucken, 24-6.

67. William Allen Miller, "On the Action of Gases on the Prismatic Spectrum," [abst.] Report of the British Association (1845, Cambridge):  28-9; For the full text of Miller's talk, see  "Experiments and observations on some Cases of Lines in the Prismatic Spectrum Produced by the Passage of Light through Coloured Vapours and Gases, and from Certain Coloured Flames," Philosophical Magazine, Third Series, 27 (1845):  81-91.

68. Gustav Kirchhoff and Robert Bunsen, "Chemical Analysis by Spectrum-observations," Philosophical Magazine, Fourth Series, 20 (1860):  89-109.

69. See diary entries for 25 October 1860); 22, 23 and 28 November 1860; 1 December 1860 in Shephard T. Taylor, The Diary of a Medical Student during the Mid-Victorian Period 1860-1864 (Jarrold and Sons, Limited:  Norwich, 1927).  I am indebted to Professor D. I. Davies at King's College, London, for bringing this book to my attention.

70. William Allen Miller, Elements of Chemistry:  Theoretical and Practical, Part I. Chemical Physics, 2nd ed. (John W. Parker and Son:  London, 1860):  146.

71. W. A. Miller, "Experiments and Observations on some Cases of Lines in the Prismatic Spectrum Produced by the Passage of Light through Coloured Vapours and Gases, and from Certain Coloured Flames," Chemical News 3 (1861):  304-7. This paper originally appeared in 1845 in the Philosophical Magazine.  For the full citation, see note 67.

72. William Crookes, Chemical News  3 (1861):  303-4.  Portions of Crookes' comments were later cited by Kirchhoff in the Philosophical Magazine [ 25 (1863): 250-62] and subsequently included in Henry Roscoe's Spectrum Analysis, 120.

73. Henry Roscoe to George Stokes, 6 February 1862, Stokes Papers, Add MS 7656.R794, University of Cambridge Library.

74. Gustav Kirchhoff, "Contributions towards the History of Spectrum Analysis and of the Analysis of the Solar Atmosphere," Philosophical Magazine, Fourth Series, 25 (1863):  250-62; 255.

75. Ibid., 261.

76. William Allen Miller, "Address to the Chemistry Section," Report of the British Association (1861, Manchester):  75-6.  A lengthy summary of this talk was given in Chemical News 4 (1861):  159-61.

77. [Agnes M. Clerke] "William Allen Miller," Dictionary of National Biography 37, Sidney Lee, ed. (Macmillan & Co.:  New York, 1894):  429-30; 429.

78. [Trotter], "Miller," xxii.

79. Huggins, "The New Astronomy," 911.

80. William Allen Miller, "On Spectrum Analysis," Pharmaceutical Journal, Second Series, 3 (1862):  399-412; idem., Chemical News 5 (1862):  201-3; 214-8.

81. Miller, "On Spectrum Analysis," 400.

82. Ibid., 412.

83. Ibid.

84. Report on Presidential Address by William Allen Miller to the Chemical Section of the British Association, in Chemical News 4 (1861):  159-61; 159.

85. Henry Roscoe, "On Bunsen and Kirchhoff's Spectrum Observations," lecture presented to the Royal Institution on March 1, 1861, published in Chemical News 3 (1861):  153-5 and 170-2.

86. Henry Roscoe, "A Course of Three Lectures on Spectrum Analysis," published in Chemical News 5 (1862):  218-22, 261-5, and 287-93.

87. In this, Roscoe was stressing the differences noted by Fraunhofer between the sun and the individual stars he had observed.  Miller, in his lecture to the Pharmaceutical Society, emphasized the similarities mentioned by Fraunhofer, in particular that Fraunhofer had seen the sodium D lines in Procyon's spectrum, while in the spectra of Capella and Betelgeuse, he had seen the magnesium b band as well as the D lines.  The difference in presentation of Fraunhofer's observations is subtle, yet interesting in terms of what the two men seem to have believed further investigation of stellar spectra might reveal.  See J. Fraunhofer, "A short account of the results of recent experiments upon the laws of light, and its theory," Edinburgh Journal of Science 7 (1827):  101-13, 251-62 and 8 (1828):  7-10; this appeared originally as "Kurzer Bericht von den Resultaten neuerer Versuche über die Gesetze des Lichtes, und die Theorie derselben," Gilbert's Annalen 74 (1823):  337-78.

88. Henry Roscoe, Lecture III in "A Course of Three Lectures on Spectrum Analysis," (12 April 1862) in Chemical News 5 (1862):  287-93; 292.

89. Huggins, "The New Astronomy," 913.

90. For an account of what became known as Tebbutt's Comet, which appeared in the summer of 1861, see Reverend R. Main, "Observations of Comet II, 1861," Monthly Notices of the Royal Astronomical Society 22 (1862):  55-7.  This comet was reportedly so bright between 29 June and 1 July that it cast shadows at night, making it possibly one of the brightest comets of the century.  See Roberta J. M. Olson, Fire and Ice:  History of Comets in Art (Walker and Company:  New York, 1985):  97.  For drawings of Comet 1861 II by Secchi and De la Rue, see Carl Sagan and Ann Druyan, Comet (Random House:  New York, 1985):  175, 179, and 180.

91. Observations on the transit were noted by others in England:  see the reports of J. W. Jeans and Joseph Baxendell in Monthly Notices of the Royal Astronomical Society 22 (1862):  42-3.

92. Dawes had published in the Monthly Notices a list of dates and times when the transit of Titan could be observed.  One scheduled for 17 May was observed and recorded by Huggins.  A second one scheduled for 2 June, was unfortunately clouded out.

93. William Lassell, "Discovery of a new Satellite of Saturn," Monthly Notices of the Royal Astronomical Society 8 (1848):  195-7.  W. C. Bond, "Discovery of a new Satellite of Saturn," Monthly Notices of the Royal Astronomical Society 9 (1848):  1-2.  [Note:  the actual sighting was by G. P. Bond, but as W. C. Bond was a Fellow of the RAS, the note was communicated by him.  This has occasionally led to some confusion.]

94. Lassell's original announcement concerning the alleged satellite of Neptune was couched in tentative terms.  The observation on which this claim was based was made when Neptune was positioned too near the sun to permit a second search. Lassell stated "One, or perhaps two, luminous points have been seen, which may be satellites; but this will require further scrutiny." [ Monthly Notices of the Royal Astronomical Society 7 (1848):  157.]  In this same announcement he claimed to see what appeared to be a ring crossing Neptune's disk, a feature which he later reported others had seen as well.  The following summer, when Neptune had returned to a more favorable position, Lassell confirmed his sighting of the satellite, as did Otto Struve and G. P. Bond.  With such a questionable first observation, however, it is not clear that these subsequent sightings can reasonably be called confirmations.  See also, William Lassell, "On a satellite of Neptune," Monthly Notices of the Royal Astronomical Society 11 (1850):  61-2; Robert W. Smith, "William Lassell and the Discovery of Neptune," Journal for the History of Astronomy 14 (1983):  30-32; Robert W. Smith and Richard Baum, "William Lassell and the Ring of Neptune:  A Case-study in Instrumental Failure," Journal for the History of Astronomy 15 (1984):  1-17.

95. William Lassell, "In a letter dated...," Monthly Notices of the Royal Astronomical Society 11 (1851):  201.

96. In the case of the satellite of Saturn, Hyperion, the professional astronomer, Bond actually saw the moon first, on 16 September 1848, while Lassell, the amateur, did not spot it until the 18th.  However, both confirmed their first observation on 19 September, making them co-discoverers.

97. W. R. Birt, "On the Appearance of Saturn's Ring, 1862," Monthly Notices of the Royal Astronomical Society 22 (1862):  295;  Rev. W. R. Dawes, "Saturnian Phenomena," Monthly Notices of the Royal Astronomical Society 22 (1862):  297-9.

98. William Huggins, "On some Phenomena attending the Disappearance of Saturn's Ring, May 19th, 1862," Monthly Notices of the Royal Astronomical Society 22 (1862): 295-6.

99. William Huggins, 13 April and 30 October 1863, Notebook 1.

100. For a discussion of the contents of the individual notebooks which is essentially correct except on this one point, see Julie Morgan, "The Huggins Archives at Wellesley College," Journal for the History of Astronomy 11 (1980): 147.

101. It is also possible that Miller kept the records of these early experiments.  I shall raise this suggestion again in chapter 4, in connection with Huggins' and Miller's photographic efforts.

102. William Huggins, 23 March 1864, Notebook 1.

103. For a record of the discussion at the RAS on the question of the features on the sun's surface, see Astronomical Register 2 (1864):  4-6; 98-101.  For an historical analysis of the debate, see C. F. Bartholomew, "The Discovery of the Solar Granulation," Quarterly Journal of the Royal Astronomical Society 17 (1976): 263-89.

104. William Huggins, 3 July 1865 and 6 July 1865, Notebook 1.

105. William Huggins to J. Norman Lockyer, 6 December 1878, Lockyer Papers, Exeter University Library.

106. William Huggins, "On the Periodical Changes in the Belts and Surface of Jupiter," Monthly Notices of the Royal Astronomical Society 22 (1862):  294; and "On some Phenomena attending the Disappearance of Saturn's Ring, May 19th, 1862," Monthly Notices of the Royal Astronomical Society, 295-6.

106. "Mr. Huggins' Observatory," Monthly Notices of the Royal Astronomical Society 25 (1866):  107-9.

108. William Huggins, "On the Disappearance of the Spectrum of E Piscium at its Occultation of January 4th, 1865," Monthly Notices of the Royal Astronomical Society 25 (1865):  60-2.

109. Huggins and Miller, "Note on the Lines in the Spectra of some of the Fixed Stars," Proceedings of the Royal Society 12 (1863):  444-5; idem., "On the Spectra of some of the Fixed Stars," Proceedings of the Royal Society 13 (1864):  242-4. The first of these articles is simply a preliminary note presumably to establish priority and announce their intention for continued investigation.  The second article is a brief version of a paper with the same title submitted to the Philosophical Transactions [ 154 (1864):  413-35].

110. William Huggins, "On the Spectra of some of the Chemical Elements," Proceedings of the Royal Society 13 (1863):  43-4.  This article is an abstract of a paper submitted to the Philosophical Transactions [ 154 (1863):  139-60].  In its abstracted form it also appeared in the Philosophical Magazine, Fourth Series, 27 (1863):  541-2.

111. William Huggins, "On the Spectra of some of the Nebulae," a supplement to the paper "On the Spectra of some of the Fixed Stars," Proceedings of the Royal Society 13 (1864):  492-3.  This was an abstract of a paper of the same name submitted to the Philosophical Transactions [ 154 (1864):  437-44].  In its expanded form it also appeared in the Philosophical Magazine, Fourth Series, 31 (1864): 523-32.  Another paper authored by Huggins and entitled "On the Spectrum of the Great Nebula in the Sword-handle of Orion," appeared in the Proceedings of the Royal Society [ 14 (1865):  39-42] as a follow-up of his earlier paper on the spectra of nebulae.

112. Astronomical Register 1 (1863):  2.

113. Prof. Donati, "Memorie Astronomische," Monthly Notices of the Royal Astronomical Society 23 (1863):  100-7; originally published in the Annals of the Museum at Florence, 1862.

114. This paper was submitted for publication on 19 February 1863.  Huggins and Miller, "Note on the Lines in the Spectra of some of the Fixed Stars," Proceedings of the Royal Society 12 (1863):  444-5.

115. Huggins and Miller, "On the Spectra of Some of the Fixed Stars," Philosophical Transactions 154 (1864):  413-35.

116. Lewis M. Rutherfurd, "Astronomical Observations with the Spectroscope," American Journal of Science 35 (1863):  71-7; Deborah Jean Warner, "Lewis M. Rutherfurd:  Pioneer Astronomical Photographer and Spectroscopist," Technology and Culture 12 (1971):  190-216.  Huggins later claimed to have heard of Rutherfurd's work just as he arrived at the Royal Society to deliver his own paper on the same subject.  See, William Huggins to H. H. Turner, 15 February 1893, Correspondence of the Society, Royal Astronomical Society Library.

117. Certainly the length of study and the extent of Miller's and Huggins' work impressed Lockyer in his report on the paper in the Monthly Notices of the Royal Astronomical Society.

118. Huggins and Miller, "On the Spectra of Some of the Fixed Stars," Philosophical Transactions 154 (1864):  413-35; 414-9.

119. Huggins and Miller, "Note on the Lines in the Spectra of some of the Fixed Stars," Proceedings of the Royal Society 12 (1863):  444-5; 444.

120. Huggins, "The New Astronomy," 912-3.

121. Huggins and Miller, "On the Spectra of Some of the Fixed Stars," Philosophical Transactions 154 (1864):  413-35; 414.


William Huggins' Early Astronomical Career

  • Chapter 2

Unlocking the "Unknown Mystery of the True Nature of the Heavenly Bodies"

The Astronomical Agenda:  1830-1870

    • Part 2

"A sudden impulse..."

Reception of Spectrum Analysis Applied to the Stars

  • Chapter 3—

Moving in the Inner Circle

Cultivating Advantageous Alliances; Opportunism and Eclecticism

Opportunism and Eclecticism (continued)

Achieving "A mark of approval and confidence"

  • Chapter 4—

Margaret Huggins: The myth of the "Able Assistant"

The Solitary Observer

Celestial Photography

Diversity and Controversy: Defining the Boundaries of Acceptable Research

  • Chapter 6—

Solar Observations at Tulse Hill

The Red Flames

The Eclipse Expedition to Oran

Photographing the Corona Without an Eclipse

The Bakerian Lecture