Wavelengths – Nineteenth-Century Spectroscopy

and the Birth of Modern Astrophysics (Part II)

Selected Topics From the History of Astronomical Spectroscopy - Theoretical

Development, Instrumentation, and Spectroscopic Discoveries in Astronomy

Andrew Bell [ã2002] - for the Northern California History of Astronomy

Luncheon and Discussion Association (NCHALADA) - March 30, 2002

At our January meeting, we discussed the historical foundations of spectroscopy up to about 1875, together with the subsequent introduction and early development of precision instrumentation by Henry Rowland (concave diffraction gratings) and Albert Michelson (the interferometer). At this month's meeting, we have planned to focus on major astrophysics applications. I will continue working from the same high-level outline that I prepared for the January meeting, but have added some additional reading suggestions and discussion topics. I have identified what I think of as six fundamental application areas for early astrophysics. We'll have time at this meeting to discuss how knowledge was developed and astrophysics was practiced in each application area, starting with the earliest preconceptions, but with particular attention on the period of approximately 1875 to 1925. I see this as the period immediately following the introduction of those new instrumentation tools and early conceptual frameworks which can in hindsight now be recognized as prerequisites for productive analysis in each application area.

[Sections 1-4][See notes from NCHALADA LIX]

5. Fundamental Tools for Early Astrophysics

Astrophysics could not have developed with the spectroscope all by itself, and I have also come to think of some of the earliest results in astrophysics as having preceded the spectroscope (e.g., detection of the local motion of our own solar system). While astrometry [the locations of the stars] and photometry [the measurement of their brightness] might have developed independently of astrophysical concerns, both disciplines provided tools of fundamental importance to the early astrophysicists. And absent the development of photography, astronomical spectroscopy could never have advanced beyond study of the spectrum of the sun and those of the very brightest stars.

a. Astrometry [modern astrometry probably began with Tycho Brahe; by standard of measurements with sufficient accuracy and precision for later analysis of proper motion. By time of the late 18th century, many observers had concluded that minimum parallax must be one arc-second or less (or stars with larger parallax would have already been detected), and identification of the direction of motion of our own solar system had been identified as a reasonable target for analysis. William Herschel correctly identified "solar apex" as towards direction of Hercules in 1783. See Hoskin (1963) for additional detail. Many writers distinguish between practices of "classical" astronomy and astrophysics in later decades of nineteenth century; high precision astrometric measures would have been important to both groups of practitioners.]

b. Photometry [significant efforts at quantitative measures of stellar magnitudes began with "extinction" tests made using stacks of filters; latter part of 18th century] [early photographic efforts were greatly complicated by limited early understanding of photochemistry, including reciprocity assumptions and distinctions between photographic and visual magnitudes; note also that photoelectric experiments began late in the 19th century, with Joel Stebbins having developed and then used a selenium-based photocell to accurately measure the Algol light curve (including first detection of the secondary minimum) at Illinois in 1909] [all of this is from a first quick read of Hearnshaw's "Measurement of Starlight" (1996)]

c. Photography - early photographic applications required extraordinarily long exposure times [e.g., Fath (1911) describes exposures of up to ten hours apiece to obtain early spectrograms of the spiral nebulae using the 60-inch reflector at Mt. Wilson, even while working at a dispersion of just 400 Angstroms to the millimeter, while Pease (1915) describes a single thirty-four hour exposure used to obtain a radial velocity measurement for Andromeda using the same telescope three years later, "an excellent spectrogram ... was obtained on the nights of November 12th, 13th, 14th, 15th and 16th, 1914 (!!)]. Late nineteenth and early twentieth century advances in photochemistry must have been at least as important as were the new large-aperture refracting and then reflecting telescopes, in making a significantly greater breadth of objects available for spectral analysis. Even as late as the early 1950s, "the telescope operators at MacDonald would cringe when they saw us show up with our spectro-photometry apparatus" [______, 199_] [the operators would be in for a long and terribly exacting night].

· early development – the Herschels played an important role here, as well. There is a fair amount of detail about the early development of photography in Chapter 4 of the single modern book-length biography of John Herschel (Buttmann, "Shadow of The Telescope," 1970); see also [Schaaf (1992)]. While Niepce and Daguerre are properly credited with first bringing early photochemical experiments to fruition, their work was preceded by nearly a century of investigations by others (notably Priestley; see Buttmann at pp. 131-132). Herschel and Talbot were able to reproduce Daguerre's results very quickly (within several weeks) after its first general announcement early in 1839. Buttmann describes Herschel as the "only one" among the earliest photographic experimenters to be a "competent chemist with both theoretical knowledge and practical experience" (p. 137), and so ascribes Herschel's rapid success to his earlier training in chemistry.

There is a great deal of detail on the mid-century development of astrophotography in Chapter 4 of Hearnshaw (1996). Later in the nineteenth century, E.E. Barnard (first at Lick, then from Yerkes) produced many of the earliest high-quality astrophotographs [e.g., see his Milky Way star field photographs, reproduced in the second article of the inaugural number of The Astrophysical Journal (1895)]. Keeler directed a great deal of his attention to development of the 36-inch Crossley reflector as a tool for high-quality astrophotography, during the period of his sadly brief tenure as the second director at Lick [Osterbrock (1984), at pp. 297-307].

· spectrographic applications – it took even longer to get good spectroscopic photographs than it did for development of optical astrophotography [early applications (e.g., stellar classification projects at Harvard) were necessarily limited to relatively low-dispersion studies, using hundreds of Angstroms to the millimeter] [just as with photometry, these efforts were further complicated by differing photosensitivity to differing portions of the visual, infrared, and ultraviolet spectrum] [notable

early exception: good photographic maps of the solar spectrum; e.g., Rowland's of 1888].

6. Early Astrophysics Applications

I think of the first approximate three-quarters of a century of astronomical spectroscopy as having laid the foundation for modern astrophysics. By this, I mean to include much of our current understanding of (following Comte), both "the stars' shapes, their distances, their sizes and their movements" and "their chemical composition [and] mineralogical structure," as well as (leaving Comte far behind), the composition, evolution, and physical mechanics of the larger scale universe.

For the purposes of our discussion, I have selected six fundamental application areas that I think of as the most important separate areas of early study: physics and chemistry of the sun, visual and spectroscopic analysis of binary stars, local-scale physical mechanics, analysis of the nebulae, stellar classification and evolution, and large-scale physical mechanics. There will certainly be overlaps between the various categories. For example, improved understandings of the physics and chemistry of our own sun and information obtained from the analysis of binary stars both contributed in key ways to our modern understanding of stellar structure and evolution. Similarly, analysis of the nebulae contributed in important ways to our understanding of both the local-scale and large-scale structure of the universe. Under "local-scale physical mechanics," I mean to include work addressing the physical dynamics of our own galaxy (e.g., subjects such as our own position relative to the galactic center and the placement of our own and neighboring spiral arms, but probably excluding galactic rotation). Under "large-scale physical mechanics," I would refer to the rotation of our own and nearby galaxies, plus subjects such as the relative motion of our own Local Group, mechanics of entire clusters and superclusters of galaxies, and the Hubble constant.

a. Physics and chemistry of the sun – one important part of the nineteenth-century history of solar physics was the struggle to identify an appropriate heat source for the sun. While I have previously described Lockyer's meteoritic precipitation theory as "strange," it did have a long heyday during the period that people were trying to figure out what could make the sun so hot (and stay that way for so long), but before any of the necessary details in nuclear physics had coalesced [c.f. Trimble (1995), or imagine last year's Leonids and think about them as "one of the slow nights"]. The principal competing theory was that of gravitational contraction, first advocated by Helmholtz and subsequently adopted by both Huggins and Kelvin. Powell (2001) describes the evolution of 19th century geological knowledge about the age of the earth and the difficulties in reconciling this knowledge with hypothetical fuel sources for the sun, while also calling attention to an important qualifying note in one of Kelvin's early papers, "unless sources now unknown to us are prepared in the great storehouse of creation."

Earliest conceptions of the sun (those preceding both spectroscopy and modern understandings of heat and light) were quite speculative and seem incredibly primitive today, albeit with two centuries worth of hindsight. William Herschel looked at the sun as most likely habitable, with sunspots interpreted as windows through a very hot cloud layer to the surface below. [Hoskin describes Herschel's primary source as having been "Astronomy Explained Upon Sir Isaac Newton's Principles," by James Ferguson (1756), which was one of the few standard astronomy texts from the mid-18th century.]

Spectral analysis of the sun began with increasingly accurate catalogs of the solar spectrum (e.g., Fraunhofer, Angstrom and Rutherfurd, then Rowland). Early studies could assume the positions of the solar lines as fixed. With increased precision came discovery of and insight into dynamic behavior of the sun (e.g., Hale's sunspot studies and identification of magnetic fields associated with sunspots). See Hufbauer (1991), Chapters 1 through 3, for additional discussion of solar physics up to 1940.

As our understanding the sun progressed, a great deal of work was required to study the spectral expression of previously unimagined extreme temperature and atmospheric conditions. Much of the early analysis in this area was done under Henry Rowland's direction at Johns Hopkins over the course of the 1890s [e.g., Humphreys and Mohler (1896)] [see also, Hentschel (1993)]. Additional work was required to develop a theoretical framework for the relationship between temperature and pressure conditions and ionization, and how that is expressed in the relative intensity of different spectral lines from the same series. And it took even longer, until the later 1920s, to determine how to read the relative abundance of hydrogen and helium in contrast to those of the heavier elements [Payne (1925) and (1930)]. There was still a lot of basic science left to be done at this point, too. While the "revised preliminary" tables issued by St. John et al (1928) catalogued nearly 20,000 solar lines, almost 8,000 of those lines had yet to be assigned to a specific chemical source.

b. Analysis of binary stars – physical relationships between optical binaries were suspected at least as early as last quarter of 18th century; with first definitive reports by William Herschel in 1803 [e.g., for Castor, although Burnham describes measurements by Bradley beginning in 1718, with a change in position angle of approximately 30 degrees noted by 1759]; see Hoskin (1963) for an extended treatment including extensive quotes from Herschel's early papers. John Goodricke (1765-1786) is credited with first quantitative measures of Algol's light curve, together with the accurate hypothesis describing Algol as an eclipsing binary.

Direct visual observation of physical binaries opened the door for the classical astronomers of the early 19th century to calculate orbits. Once adequate parallax measures became available for some of these pairs (and visual binaries with observable orbits pretty much have to be fairly close by), the techniques of classical mechanics could be used to determine scales and calculate combined masses for such systems [e.g., the two components of Castor are a reasonably well-matched pair at a distance of approximately 50 light-years (thus with a total annual parallax of roughly one-seventh of an arc-second), and so with total luminosity about 40 times that of our sun; their mean separation is about 90 AU, or about twice the distance of Pluto from the sun, and the orbital period is now estimated at approximately 400 years; this means that the total mass of the system must be about six times that of our sun]. Slowly but surely, mid-18th century astronomers were able to identify a limited number of (combined) mass and luminosity data points of this type using classical analytic methods alone.

The later identification and study of spectroscopic binaries added to this knowledge considerably. First of all, there will be more of them (can study closer-orbiting pairs, and pairs at greater distances from us). Also, close analysis of the light curves for eclipsing binaries [e.g., Algol] would reveal additional detail as to the physical size of each component. And as relationships between size, temperature, and luminosity became better understood, analysis of some pairs identified unusual dwarf stars of special interest.

Earliest confirmed finding for a spectroscopic binary is described by Hearnshaw as by Vogel at Potsdam (Algol, 1889); second was by Antonia Maury at Harvard (b Aur, also 1889). Other authors give precedent to Pickering for Mizar, again in 1889. For early publications in PASP and ApJ, see Keeler (1890) reporting Vogel's findings for Spica (includes good description of expected tidal forces) and Vogel (1903) for e Aur [see either Burnham, Jr. (1978) or Moore (1987) for both Algol and e Aur].

Just for good measure, both principal components of Castor are themselves well-matched spectroscopic binaries. The two components of Castor A are each thought to be about twice the diameter of the sun, about 12 times its luminosity and perhaps 1.6 times its mass. They rotate about each other in a fairly high-eccentricity orbit, with a period of nine days at a mean distance of just four million miles. The two components of Castor B are a shade smaller and less massive; each is about six times as bright as our sun; their mutual orbit is more nearly circular, with a mean distance of less than three million miles in a period of [less than three days][check]. And then there is Castor C, which is approximately one arc-minute (or about one light-week) nearly due south of the principal components; it too is a spectroscopic binary, and is made up of two red dwarfs in a very tight, eclipsing orbit with a period of just 20 hours (the magnitude range is 9.1 to 9.6) [all of these details are as described in Burnham, at pp. 915-918 of Vol. II].

c. Local-scale physical mechanics – see Hoskin (1963) for extended discussion of Herschel's work in this area long before the advent of spectroscopy. See Vogel (1898, 1900) and Hartmann (1900) for early work done in this area at Potsdam; and Seares (1918) for discussion of star distributions incorporating both radial and proper motion studies. Radial velocity measures were a significant early research area at Lick, culminating in the publication of a major data catalog by J.H. Moore in 1932 [includes data for 6739 stars, 133 "gaseous nebulae," 18 globular clusters, and concludes with what are described as the "apparent" radial velocities of 90 "extra-galactic nebulae" (all of which are attributed to Stromberg and Humason at Mt. Wilson)].

The Hyades cluster in Taurus is an easy binocular object that is of great importance (both historical and modern) to all measures of astronomical distance [e.g., see Rowan-Robinson (1985) at pp. 47-52]. The original work for this object was done by Lewis Boss in 1908, longtime director of the Dudley Observatory in Albany, New York. While Boss is often cited as an exemplar of the "old" astronomy that was going out of fashion at the turn of the 20th century, I see his work on the Hyades as a quite insightful and inventive amalgam of the best of both the old and the new.

At a then-uncertain distance but with a common radial velocity of approximately 40 kilometers per second in recession for the three brightest cluster members [or a little less than one Angstrom of redshift for the sodium D lines], the Hyades are just far enough away to be readily identifiable as a fairly well-defined aggregation [e.g., in contrast to the Ursa Major moving cluster]. Even today, however, they are very nearly at the outside of the range for reliable direct (ground-based) parallax measures – and yet they are close enough by that good measures of their proper motion were already available by the early 1900s for about three dozen individual members of the cluster. Describing his interest in the Hyades as extending back a quarter-century, Boss explained "for many years I have made unsuccessful attempts to identify a possible radiant, or convergent, for this stream." In 1908 he succeeded. By placing greater emphasis on the outlying members of the cluster, Boss was able to identify a "convergent" point for the proper motion of the Hyades, approximately 28.9 degrees to the east of their current location (about 5 degrees NE of Betelgeuse), "we find that at about 65,000,000 years from the present time it may be supposed that the Taurus stream will appear as a globular cluster about 20' in diameter and constituted largely of stars of magnitudes 9 to 12, with a well marked central condensation" [Boss (1908)].

Lewis Boss' careful application of the "old" science of astrometry made it possible for him use the new astronomers' measure of the apparent common recession velocity to determine the cluster's true bearing relative to our own. The entire group is streaming 28.9 degrees to the east. Boss then used simple trigonometry to derive 45.6 km per second (made up of 22.1 km/sec transverse to our line of sight and 39.9 km/sec in recession) for the common motion of the entire cluster. The transverse component would add up to a bit more than 450 AU [about 2.7 light-days] over an entire century and this can be compared with the cluster's common proper motion of approximately 11.1 arc-seconds per century to show this the Hyades to lie at a distance from us of very nearly 42 parsecs (about 140 light-years).

I enjoy showing people the double cluster in Perseus as the single readily-observable "display object" located in an adjacent spiral arm rather than our own [see Burnham, Jr. (1978), Vol. III at pp. 1438-1447].

d. Analysis of the nebulae – the first nebula to be studied using the spectroscope was NGC 6543, by Huggins in 1864, as cited in numerous references including Chambers (1910). It is a small planetary nebula in Draco (just 0.4 arc-minutes in diameter) of unusually high surface brightness (c. magnitude 6), so much so that "an exposure of just a few minutes [through the 36-inch refractor at Lick] burns out all the detail in the central portion" (Curtis, as cited in Burnham Jr. (1978), Vol. II at p. 870). It is located at a distance of about 5,000 light-years, so its true physical size must be about 30,000 A.U., or roughly half a light-year in diameter. Its distinctive color is "the result of the radiation of doubly ionized oxygen, at 5007 and 4959 angstroms, the so-called 'forbidden lines' characteristic of many of the planetary nebulae" (Burnham Jr. again, at p. 872). William and Caroline Herschel had classified NGC 6543 under their category IV (No. 37), meaning they recognized it as a "planetary nebula" [as distinguished from all of the "bright," "faint," and "very faint nebulae" (most of which are now recognized as galaxies) recorded in their Categories I-III].

The "great nebula" in Orion was of course another of the most significant late nineteenth century study objects. See Osterbrock (1984) at pp. 92-94 and 98-101 for discussion of James Keeler's early work on the Orion Nebula at Lick [including Keeler's identification and careful communication of significant new evidence bearing on a long-running dispute between Lockyer and Huggins as to presence of the diffuse lines of magnesium molecules in the nebula (Lockyer) versus the sharper lines of magnesium atoms alone (Huggins), with serious consequent implications for the former's meteoritic hypothesis].

Bright emission lines were observed for large numbers of what we now know as the intra-galactic nebula during the late 19th and early 20th centuries. Those of "the nebulae" that we now know to be extra-galactic were much harder. I read many of those pre-1925 publications which rejected extra-galactic explanations for the external galaxies as assigning too much weight to the emission lines observed for the nebula inside our own galaxy and too little weight to the results that were slowly becoming available for early extra-galactic targets. See Scheiner (1899) and Fath (1909)(1911)(1913) for early discussions of the spectra of the spiral nebulae, and also Slipher (1913) and Pease (1915) for the first noteworthy radial velocity measurements [with Slipher's measures described as also including the internal rotations].

e. stellar classification and evolution – It took much longer to work out enough of the details about absorption-line spectra (which is what astrophysicists need to look at in order to study the composition and evolution of the stars) to make productive use of them than it did for the bright emission-line spectra which are used to study the chemical composition of the intra-galactic nebula. There are some pretty good reasons why it took so much longer (just for starters, emission lines are few and bright, while absorption lines come in bunches, and they are dark).

A number of the earliest investigations were of nova, with the first recorded spectroscopic study being Huggins' of T CorB in 1866. See Burnham, Jr. (1978), Vol. II at pp. 708-714 for T CorB, as well as J.B. Bouton (1869) for a treatment which appeared three years later in Appleton's Journal. T CorB also proved to be the first example of a recurring nova, with its second appearance early in 1946 resulting in a number of new studies published shortly thereafter.

See Jones and Boyd (1971) for history of the Harvard College Observatory from 1839 to 1919, with a special emphasis on the Pickering era and stellar classification at Harvard. David DeVorkin has recently published a major biography of Henry Norris Russell [DeVorkin (2000)], and this is an excellent resource for much of the later work on stellar evolution. I also recommend Haramundanis (ed.) (1996) and the more difficult to find Payne (1925) and (1930) for Cecilia Payne's work in this area.

Many of the theoretical tools and much of the conceptual understanding first developed for the sun would also have contributed to early astrophysical studies of the stars (e.g., spectral expression under extreme temperatures and pressures, the blackbody radiation model, ionization effects, and relative abundance of light versus heavy elements). There are many more stars than one, and the early analysis of visual and spectroscopic binaries was important as astrophysicists began identifying and enumerating distinct stellar populations (e.g., developing our modern understanding of mass-luminosity relationships, determining the range of star densities and so distinguishing between main sequence and giant stars). Oversampling problems (while stars of very high luminosity are quite rare, they are predominant among those that can be catalogued and inspected visually) probably made a lot of the early work quite difficult. Analysis of cluster populations was helpful in one sense, by controlling for the oversampling problem, but introduces a different bias by drawing out larger numbers of relatively young stars. Finally, the study of variable stars plays a critical role in understanding stellar evolution and might well merit additional consideration as a separate sub-category.

f. large-scale physical mechanics - there are a number of recent survey articles that provide good treatments of much of the early 20th century work addressing nature of the spiral nebulae and resolution of the "island universe" hypothesis [see esp., Rubin (1995); and Trimble (1995)]. Berendzen, Hart, and Seeley (1976) and Smith (1982) both provide good book-length treatments of this subject; I find Smith's treatment a bit more thorough and reliable in its references to the entire body of early work. See also Chapter 8 of Hetherington (1988) for a thought-provoking discussion of van Maanen's role.

I find it quite striking just how much was really quite well established long before 1926 and Hubble's first formal publication of the Cepheid discoveries for M33 [see Smith (1982) at pages 27-28 for W.W. Campbell's summation of the evidence that was already available by 1916]. Even where the direct spectroscopic evidence was thought to be inconclusive [e.g., it was perhaps unfortunate that one of Fath's earliest targets was M77 in Cetus, which is an active Seyfert galaxy with an extraordinarily complex spectrum], there were huge discrepancies between radial velocity measures for the spiral nebulae and those of all of the intra-galactic objects (e.g., individual stars and the planetary nebulae). And there were novas popping up all over the place, usually right smack in the middle of one of the spiral nebulae. See Curtis (1915) and (1917) for tentative early lower bounds on the possible distances of the nebulae, together with Lundmark's extensive publications between 1921 and 1926.

The most important contrary evidence seems to have been Adriann van Maanen's photographic (rather than spectroscopic) observations of galactic rotation, which could never have been reconciled with the great distances (and so sizes) implied by extra-galactic nebulae [c.f., Harlow Shapley's oft-cited and sorrowful, "I believed in van Maanen's results ... After all, he was my friend" (as discussed in Haramundanis (ed., 1996); see Virginia Trimble's Note 3 at pp. xx-xxi, as well as pp. 208-210 for Cecilia Payne's version, putting the Shapley quote into its historical context)]. Also, contemporary estimates for the size of our own galaxy had it far too small, and this had extra-galactic interpretations for the spiral nebulae make them seem far "too big" relative to the accepted size of our own.

Suggested Discussion Topics (Part II)

1. The visible spectrum – many discussions of optics and light contrast the fairly narrow range of the ordinary visible spectrum with the relative breadth of the audible one; e.g., we see just "one octave" (or a bit less) out of the entire electromagnetic spectrum, but can perceive sounds across a range of at least eight octaves. There is a good physiological basis for our perception of light, but how did it turn out that we get to enjoy looking through this one particularly interesting band pass, and not some other one?

2. Achievable precision – I have drawn attention to a number of instances where early practitioners assigned unrealistic or overreaching precision to their observations. In some cases this simply produces reported results that understate the implied or stated tolerance, and there are a lot of examples of that in the history of science. In others [notably, van Maanen's internal rotation measures; less frequently cited, William Herschel's "comet" measurements of 1781] [see Hetherington (1988), at pp. 25-26], it leads to conclusions that have no objective basis, because the observer's unrecognized measurement errors exceed the magnitude of what is being measured.

By any standard, I think that Adriann Van Maanen's internal rotations ought to have been set aside as unsupportable, right from the outset. All of his measured displacements corresponded to a half an arc-second or less between each pair of plates, and usually less than a quarter of an arc-second. Ignoring the fact that he claimed a hundredth of an arc-second as his stated measurement precision, I fail to understand how even a quarter of an arc-second could be measured off of photochemical plates from a ground-based telescope, then or now [c.f., Baade (1963) at pp. 46-47, for a discussion of how difficult it was to achieve half arc-second photographic resolution, working in the mid-1930s].

Even among the less controversial applications and better documented standard catalogs, however, I find the usual reported working precision astounding. While the details of the data reduction techniques used are not ordinarily to be found in published journal articles, there are sometimes hints, and the efforts to squeeze an extra decimal point or two out of the available measurements appear to have been stupendous. One such hint, quoting from J.H. Moore's introduction to the Lick Observatory's General Catalogue of the Radial Velocities of Stars, Nebulae and Clusters (1932): "It is well known that systematic differences of appreciable size exist between the values of the radial velocities determined at different observatories. Their removal and the reduction of the results to a homogenous system become matters for careful consideration in preparing a catalogue of radial velocity determinations."

How "appreciable" might such differences have been? As an example of the measurement precision required, Moore explains that the catalog includes "the velocity determinations by Redman at Victoria of 225 Class K stars of visual magnitude 7.0 to 7.5 [measured] with a dispersion of about 90 A per mm [so the two sodium D lines would be separated by one-fifteenth of a millimeter]. With few exceptions the velocity for each star rests upon a single observation whose standard error is estimated to be of the order of 3 to 5 km/sec." What does it take to measure radial velocity to within three to five kilometers per second? It calls for a tolerance of one-tenth of an Angstrom or less, meaning relative line placements would have to have been measured to at least the micron level (one-thousandth of a millimeter, or one twenty-five-thousandth of an inch. While the data shown for brighter stars in the Lick catalog do include larger numbers of observations, and observations made from larger numbers of observatories. (Where multiple observatories had submitted measures for the same star, these are shown separately in the catalog.) The highest reported dispersions used are still limited to the 10 Angstroms per mm level, however, and the estimated precision generally corresponds to micron-level measurement tolerances.

3. Early applications in modern astrophysics – I have organized my discussion around six fundamental application areas: physics and chemistry of the sun, visual and spectroscopic analysis of binary stars, local-scale physical mechanics, analysis of the nebulae, stellar classification and evolution, and large-scale physical mechanics. This classification system is my own, and I would like to know whether it strikes others as useful. I have tried to arrange these categories more or less in chronological order, according to the approximate historical periods during which it first became possible to study each one productively, and I would like to hear any thoughts as to whether the order seems about right.

There are at least two major categories that I have left out altogether, at least for the time being. One is the origins and evolution of our own solar system. There is certainly a bit of astrophysics in that subject area, but I'm not sure how much of it is spectroscopy-based, and it also isn't clear to me whether this field had developed significant traction during the time period that I am considering. The other missing category is the analysis of variable stars – here, I would identify the study of different classes of variable stars as having contributed in different ways to at least three of my named study areas (local mechanics, stellar classification and evolution, and large-scale mechanics), but am uncertain about describing the study of variable stars as a separate category all by itself.

4. Early astrophysical observatories and observing programs – I have tried to name a "short list" of some of the most important of the early observatories for modern astrophysics; I came up with Allegheny, Harvard, Potsdam, Mt. Hamilton, Yerkes, and Mt. Wilson. And once again, I have tried to arrange this list more or less chronologically; not necessarily in order of when they were founded, but in order of the periods during which I have identified the most significant astrophysical observing programs as having been carried out. Were I to add two more U.S. observatories to this list, they would probably be the Dudley Observatory in Albany, New York and the Lowell Observatory in Flagstaff.

I wanted to be sure that I included at least one observatory that was located outside of the U.S. I chose Potsdam because that's the one from which I have already looked at the largest number of early research papers (e.g. Vogel, Hartmann, Scheiner). While I'm sure that those who are fond of France or England could suggest others, any discussion of the European programs during this time period would probably have to start with a consideration of the ill-fated Carte du Ciel program [see Hearnshaw (1996), pp. 136-142 for discussion of the Carte du Ciel].

5. Social grace and scientific practice – this is a theme that might be used to link our discussion of a number of different episodes from the first half-century of modern astrophysics. Osterbrock (1984) tells the elegant story of James Keeler's early work on the Orion Nebula, together with his careful negotiation of the great divide between Huggins and Lockyer. Smith (1982) and Hetherington (1988) paint a quite different picture of Harlow Shapley's role in the island universe debate and the van Maanen controversy. I think that Heber Curtis' and Knut Lundmark's earliest papers both merit discussion from the same perspective, as examples of relatively junior researchers communicating unexpected results with great caution, so much so as to perhaps delay wider recognition of their important conclusions. And the same could be said of Cecilia Payne's first characterizations of her work on hydrogen and helium abundance. Finally, wide recognition of Vesto Slipher's early work on both radial velocities and spectroscopic rotation rates for the spiral nebulae may have been hampered by the relative isolation and difficult reputation of the institution at which he worked, as well as that of its unusual director, Percival Lowell.

6. Popular coverage of astronomical research – I have included a number of articles from the 19th century popular literature among my reading suggestions [e.g., from publications such as Appleton's Journal, The Overland Monthly, and even Catholic World (for an early report from Father Secchi)]. All of these articles are readily accessible via the Making of America web sites at Michigan and Cornell; both sites include good search engines that make it possible to find many additional examples. While I am struck by just how much such coverage seems to have appeared in the 19th century popular press, it could be interesting to turn the clock forward by a century or so, and consider these examples in comparison to modern scientific journalism. Astronomy has perhaps always been the most "photogenic" of the modern physical sciences, and in this light the 19th century coverage may seem less surprising.

Wavelengths – Reading Suggestions (Part II)

Once again, I will list a number of original publications that are readily accessible on-line via the Astrophysics Data System (ADS). If you have not used this resource before, please refer to the Internet Resources section of the notes for Part I, also for discussion of access to publications available via the Making of America web sites at Michigan and Cornell.

A. Biographical Studies [for biographical sketches of a number of early practitioners, a standard reference is Macpherson's "Makers of Astronomy" (1935); via the Internet, I recommend the broadly defined MacTutor History of Mathematics archive hosted by the School of Mathematics and Computer Science at St. Andrew's University: http://www.history.mcs.st-andrews.ac.uk/history]

· M.A. Hoskin - "William Herschel and the Construction of the Heavens" (1963). Oldbourne Library.

· Bennett, J.A. (1976), On the Power of Penetrating Into Space - The Telescopes of William Herschel, Journal for the History of Astronomy 7, pp. 75-108.

· Wood and Oldham - "Thomas Young: Natural Philosopher" (1954). Cambridge.

· M.W. Jackson - "Spectrum of Belief - Joseph von Fraunhofer and Craft of Precision Optics (2000). MIT.

· G. Buttmann - "The Shadow of the Telescope - Biography of John Herschel (1970). Scribners.

· Barbara Becker (2001), "Huggins and the Origins of Astrophysics" (J. Hist. Ast. 32, pp. 43-62)

· D.E. Osterbrock - "James E. Keeler: Pioneer American Astrophysicst" (1984). Cambridge.

· David DeVorkin - "Henry Norris Russell - Dean of American Astronomers" (2000). Princeton.

· Haramundanis (ed.), "Cecilia Payne-Gaposchkin" (1996, 2nd ed.) Cambridge University Press.

· D.E. Osterbrock - "Walter Baade, A Life in Astrophysics" (2001). Princeton University Press.

B. Basic Astronomy Texts (Historical and Modern)

· John Herschel - "A Treatise on Astronomy" (1834). Carey, Lea, and Blanchard. Philadelphia.

· John Herschel - "Outlines of Astronomy" [1849 (1st), 1858 (5th), 1867 (9th), 1869 (10th)].

· W.M. Williams - "The Fuel of the Sun" (1870). Simpkin, Marshall, and Co. London.

· Simon Newcomb - "Popular Astronomy" (1878). Harper & Brothers. New York.

· Charles A. Young - "The Sun" [1881 (1st), 1883 (2nd), 1895 (Rev.)]. Appleton.

· Agnes Clerke - "A Popular History of Astronomy During the Nineteenth Century" (1887, 2nd).

· Agnes Clerke - "Problems in Astrophysics" (1903). Adam and Charles Black. London.

· Robert H. Baker - "Astronomy" (1938). D. Van Norstrand Co.

· J.B. Sidgwick - "The Heavens Above - A Rationale of Astronomy" (1950). Oxford.

· Walter Baade - "Evolution of Stars and Galaxies" (1963). The MIT Press.

· Albrecht Unsold - "The New Cosmos" (1967, in English: 1969). Springer-Verlag.

· Michael Hoskin - "Stellar Astronomy: Historical Studies" (1982). Science History Pubs.

· Patrick Moore - "Astronomers' Stars" (1987). Routledge & Kegan Paul. London.

· Michael Rowan-Robinson - "The Cosmological Distance Ladder" (1985). Freeman and Co.

· N. Hetherington - "Science and Objectivity: Episodes in the History of Astronomy" (1988). Iowa State.

· E. Karkoschka - "The Observer's Sky Atlas" (1990; 2nd ed., 1999). Springer.

· K. Hufbauer - "Exploring the Sun - Solar Science since Galileo" (1991). Johns Hopkins.

· Spark and Gallagher - "Galaxies in the Universe" (2000). Cambridge University Press.

C. Historical Physics References

· Henry Crew - "The Rise of Modern Physics" (1927). Williams & Wilkins. Baltimore (2nd ed., 1935).

· Shamos (ed.) - "Great Experiments in Physics" (1959). Holt, Rinehart and Winston. New York.

· Weart (ed.) - "Selected Papers of Great American Physicists" (1976). American Physical Society. The complete contents of this book are now available on-line; see http://www.aip.org/gap.

D. Astrometry, Photometry, and Photography

· L.J. Schaff - "Out of the Shadows: Herschel, Talbot, and the Invention of Photography" (1992). Yale.

· Hearnshaw - "The Measurement of Starlight - Two Centuries of Astronomical Photometry" (1996). Camb.

Historical Publications

· Holden (1886) - "Photography, Servant of Astronomy" (Overland Monthly 8, 459:470) [Mich. MOA]

· Schaeberle (1893) - "Terrestrial Atmospheric Absorption ... Photographic Rays" (Cont. Lick Obs. 3)

· Barnard (1898) - "The Great Nebula in Andromeda" (ApJ 8, pp. 226-228)

· Seares (1914) - "The Color of the Faint Stars" (ApJ 39, pp. 361-369)

· Seares (1915) - "Color-Indices in the Cluster NGC 1647" (ApJ 42, pp. 120-132) [OpClust in Taurus]

· Shapley (1915) - "Studies Based on the Colors and Magnitudes in Stellar Clusters"

(ApJ 51, pp. 49-61) [object studied is M68, a globular cluster in Hydra]

E. Physics and Chemistry of the Sun

· K. Hufbauer - "Exploring the Sun - Solar Science since Galileo" (1991). Johns Hopkins

· J. L. Powell - "Mysteries of Terra Firma: the Age and Evolution of the Earth" (2001). Free Press.

· High Altitude Observatory (NCAR) [www.hao.ucar.edu] - historical treatment of solar physics

· Klaus Hentschel (1993), "The discovery of the redshift of solar Fraunhaufer lines by Rowland and Jewell in Baltimore around 1890" (HSPS 23, 219:277) [see esp.: pp. 222-244]

Historical Publications

· Angelo Secchi (1867) - "The Sun" (Catholic World 7, 524:542) [available via Michigan MOA]

· W. M. Williams - "The Fuel of the Sun" (1870). Simpkin, Marshall, and Co. London.

· Simon Newcomb - "Popular Astronomy" (1878). Harper & Brothers. New York.

· Charles A. Young - "The Sun" [1881 (1st), 1883 (2nd), 1895 (Rev.)]. Appleton.

· A. L. Cortie (1891) - "Some Recent Studies on the Solar Spectrum," appeared at pp. 45-53 of general interest publication "Littell's Living Age" [available via Cornell MOA web site]

· St. John, et al - "Revision of Rowland's Preliminary Table of Solar Spectrum Wave-Lengths With an Extension to the Present Limit of the Infra-Red" (1928). [Carnegie Inst. Of Washington – Pub. of Mt. Wilson Observatory; approximately 15 pages of explanatory text followed by 200 pages of tables]

F. Analysis of Binary Stars

· M.A. Hoskin - "William Herschel and the Construction of the Heavens" (1963). Oldbourne Library.

· Robert Burnham, Jr. - "Burnham's Celestial Handbook" (1978). Dover Publications (three volumes).

· Patrick Moore - "Astronomers' Stars" (1987). Routledge & Kegan Paul. London.

· David DeVorkin - "Henry Norris Russell - Dean of American Astronomers" (2000). Princeton.

Historical Publications

· J. E. Keeler (1890) - "Spectrographic Observations of Spica at Potsdam" (PASP 3, 46:48)

· H. C. Vogel (1903) - "e Aurigae – A Spectroscopic Binary" (ApJ 17, 243:244)

G. Local-Scale Physical Mechanics

· Michael Hoskin - "William Herschel and the Construction of the Heavens" (1963). Oldbourne.

· Michael Hoskin - "Stellar Astronomy: Historical Studies" (1982). Science History Publications.

· Michael Rowan-Robinson - "The Cosmological Distance Ladder" (1985). Freeman and Co.

Historical Publications

· Keeler (1890) - "On the Motions of Planetary Nebulae in the Line of Sight" (PASP 2, 265:280)

· Vogel (1898) - "Sources of Error in Investigations ... In the Line of Sight" (ApJ 7, 249:254)

· Campbell (1898) - "Some Stars With Great Velocities in the Line of Sight" (ApJ 8, 157:158)

· Vogel (1900) - "On the Progress Made ... Stellar Motions in the Line of Sight" (ApJ 11, 373:392)

· Hartmann (1901) - "The Motion of Polaris in the Line of Sight" (ApJ 14, 52:65)

· Adams and Van Maanen (1913) - "A Group of Stars ... H and X Persei Clusters" (AstJ 27, 187:188)

· Perrine (1917) - "Preliminary Examination ... Planetary Nebulae ... Preferential Motion" (ApJ 46, 175:178)

· Seares (1918) - "The Brightness of the Stars - Their Distribution, Colors, and Motions" (ApJ 30, 99:133)

IAU Radial Velocity Program

· Frost (1902) - "Cooperation in Observing Radial Velocities of Selected Stars" (ApJ 16, 169:___)

· Slipher (1905) - "Observations of Standard Velocity Stars ... Lowell Spectrograph" (ApJ 22, 318:340)

· Moore - "A General Catalogue of the Radial Velocities of Stars, Nebuale and Clusters" (1932). Publications of the Lick Observatory, Volume 18.

The Hyades Cluster

· Boss (1908) - "Convergent of a Moving Cluster in Taurus" (AJ 26, 31:36) [distance of the Hyades]

· Smart (1939) - "The Moving Cluster in Taurus" (MNRAS 99, 168:180)

· Eggen (1969) - "The Hyades Red Dwarfs and the Distance of the Cluster" (ApJ 158, 1109:1113)

· Upton (1970) - "Calibration of the Hyades-Praesepe ... Stellar Motions (AJ 75, 1097:1115)

· Cooke and Eichhorn (1997) - "A New and Comprehensive ... the Hyades" (MNRAS 288, 319:332)

· Narayanan and Gould (1999) - "A Precision Test of Hipparcos ... the Hyades" (ApJ 515, 256:264)

H. Spectral Analysis of "the Nebulae"

· Michael Hoskin - "Stellar Astronomy: Historical Studies" (1982). Science History Pubs.

· Robert Smith - "The Expanding Universe - Astronomy's 'Great Debate' (1900-1931)" (1982).

Historical Publications

· Scheiner (1899) - "On the Spectrum of the Great Nebula in Andromeda" (ApJ 9, 149:150)

· Hartmann (1902) - "Spectographic Measures of ... the Gaseous Nebulae" (ApJ 15, 287:295)

· Fath (1909) - "The Spectra of Some Spiral Nebulae and Globular Clusters" (PASP 21, 138:143)

· Fath (1911) - "The Spectra of Spiral Nebulae and Globular Clusters (Second)" (ApJ 33, 58:63)

· Fath (1913) - "The Spectra of Spiral Nebulae and Globular Clusters (Third)" (ApJ 37, 198:203)

· Pease (1915) - "Radial Velocity of the Andromeda Nebula" (PASP 27, pp. 134-135)

· Campbell and Moore (1916) - "Note on ... NGC 7293" (PASP 28, p. 286) [Helix Nebula, in Aqr]

· Campbell and Paddock (1918) - "The Spectrum ... NGC 4151 (PASP 30, 68:69) [Spiral in CVn]

· Slipher (1918) - "Unusual Nebular Spectra" (PASP 30, 346:347) [Irr. Gal. in CVn, NGC 4449]

· Reynolds (1923) - "Note on ... Constitution of the Spiral Nebulae" (MNRAS 83, 382:385)

· Reynolds (1924) - "The Condensations in the Spiral Nebulae" (MNRAS 85, 142:147)

· Reynolds (1925) - "The Forms and Development ... Spiral and Allied Nebulae" (MNRAS 85, 1014:1020)

· Walborn and Liller (1977) - "The Earliest Spectroscopic Observations of Eta Carinae and its Interaction With the Carina Nebula (ApJ 211, pp. 181-183)

The Orion Nebula

· D.E. Osterbrock - "James E. Keeler: Pioneer American Astrophysicst" (1984). Cambridge.

· Hoskin (2002) - "The Leviathan of Parsonstown: Ambitions and Achievements" (JHA 33, 57:63)

· Huggins (1889) - "The Photographic Spectrum of the Orion Nebula" (MNRAS 49, 403:405)

· Keeler (1890) - "On the Wave-Length of the Second Line ... the Nebulae" (PASP 2, 281:285)

I. Stellar Classification and Evolution

·