by Rene Roy
40 K. Reinmuth, Die Herschel-Nebel nach Aufnahmen der Künigstuhl-Sternwarte, Veroeffentlichungen der Badischen Sternwarte zu Heidelberg, Berlin and Leipzig: Walter de Gruyter, 1926, Vol. VI. Reinmuth used Wolf types, and distinguished the directions of spiral arms.
14:11:09, subject to the Cambridge Core
.011
9. The Galaxy Classification Play-Off
199
anchored to the work of Harlow Shapley and Adelaide Ames. The most spectacular outcome
was the massive The Carnegie Atlas of Galaxies by Sandage and Bedke in 1994 that will be
discussed in Chapter 10.41 Allan Sandage wrote repeatedly about galaxy classification, its
development and the role of the RSA based on the Shapley–Ames catalogue.42 Several other
galaxy surveys used the Shapley–Ames catalogue to define and select samples of galaxies.
Gérard de Vaucouleurs “Improvements”
In parallel to Sandage’s work on galaxy classification, Gérard de Vaucouleurs published a
pioneering work on the classification and morphology of galaxies in 1959.43 Entering the
stage from the side, de Vaucouleurs thoroughly reconsidered the history of the morpholog-
ical classification of galaxies.
De Vaucouleurs had spent the period from 1951 to 1957 at Mount Stromlo Observatory
outside Canberra, Australia (see Fig. 10.1), then named the Commonwealth Observatory.
While there, he carried out a survey of the bright galaxies listed in the Shapley–Ames cat-
alogue south of the declination –30º. He had been using the 30-inch Reynolds reflector,
which had been donated to the Australian institution by John Reynolds, the same individual
whom we saw proposing the classification scheme that inspired Hubble. It was the second
largest in the southern hemisphere until the 1950s.44 In 1957, de Vaucouleurs published his
survey in a finely illustrated article that was a vanguard of the future atlases of galaxies.45
Plate I of the paper is an initial two-dimensional representation of de Vaucouleurs’ modi-
fied scheme (Fig. 9.3). Later, de Vaucouleurs changed this evocative visual representation
into a three-dimensional one. He also introduced a more detailed notation system to better
distinguish shapes of galaxies and their finer structural types. The improved classification
scheme was a complete and most elegant classification system based on morphology, but
was perceived as more complicated than Hubble’s.
While de Vaucouleurs’ articles were not atlases, the numerous plates at the end of his
articles were a carefully assembled set of images that illustrated most extensively his pro-
posed classification scheme. As such, they were sorts of mini-atlases. Like Hubble and
Sandage, de Vaucouleurs sorted galaxies along a sequence, with ellipticals starting on the
left, and going to the right as shapes flattened; then the spirals and disk systems taper, with
the irregulars at the extreme right, in a systematized way. His most original contribution
was to add a third dimension to Hubble’s diagram (Fig. 9.4). This shrunk the broad divi-
sions of Hubble and allowed finer bins for sorting objects. Set orthogonal to the sequence,
41 A. R. Sandage and J. Bedke, The Carnegie Atlas of Galaxies, Washington DC: Carnegie Institution of Washington Publications, 1994.
42 See for example A. Sandage, Classification and Stellar Content of Galaxies Obtained from Direct Photography, in Galaxies and the Universe, A. Sandage, M. Sandage and J. Kristian (editors), Chicago: University of Chicago Press, pp. 1–55.
43 G. de Vaucouleurs, Classification and Morphology of External Galaxies, Handbuch der Physik, 1959, Vol. 59, pp. 275–310.
44 The Reynolds telescope was destroyed during the 2003 Australian firestorm. The “Melbourne Telescope” was the largest telescope in the south from 1869, though the design of its tube and mounting prevented it from being a truly useful instrument until its extensive modification in the early 1960s. It, too, was destroyed in the 2003 firestorm.
45 G. de Vaucouleurs, Survey of Bright Galaxies south of –35° Declination, with the 30-inch Reynolds Reflector (1952–1955), Memoirs of the Commonwealth Observatory, 1956, Vol. III, No. 13, + 8 plates.
14:11:09, subject to the Cambridge Core
.011
200
Part III – Organizing the World of Galaxies
Fig. 9.3 “Mark I” version of de Vaucouleurs’ classification system of galaxies proposed in 1957.
From G. de Vaucouleurs (1957), Memoirs of the Commonwealth Observatory.
planes appear like the faces of a clock and convey secondary morphological features such
as bars, rings and lenses, with a gradation like on a clock (Fig. 9.5).46 Visually, it is a most
elegant arrangement.
46 This effort by de Vaucouleurs led to another massive catalogue of bright galaxies ( RC3) that has been used extensively by researchers: G. de Vaucouleurs, et al., Third Reference Catalogue of Bright Galaxies, Three volumes, New York: Springer-Verlag, 1991. It contains more than 23,000 objects with extensive size, photometric and kinematic data.
14:11:09, subject to the Cambridge Core
.011
9. The Galaxy Classification Play-Off
201
Fig. 9.4 The volume classification system of galaxies by Gérard de Vaucouleurs. Credit: G. de
Vaucouleurs (1959), Handbuch der Physik.
Gérard de Vaucouleurs’ main contributions to galaxy classification were:
r Introducing the SA for intermediate bar (SAB) classifications between the pure spiral S
and the strongly barred SB.
r Recognizing “outer pseudo-rings” as an important aspect of morphology.
r Strongly recognizing the apparent continuity in galaxy structure, to the point that de Vau-
couleurs felt he could place any given galaxy at a very specific point within his classifi-
cation volume.
r Defining a consistent way of classifying lenticulars, or S0 galaxies, which makes the
category a reasonable stage and not a “garbage bin.”47
Variation on a Theme, a Spectral Classification for Galaxies
Although convergence on the classification criterion of morphology occurred relatively
quickly, there were attempts to use other properties to distinguish and classify galaxies.
The most serious endeavor was that of American astrophysicist William Wilson Morgan
(1906–1994), a leader in stellar spectroscopy. With fellow spectroscopists Philip Childs
Keenan (1908–2000) and Edith Kellman (1911–2007), Morgan had previously devised the
very successful and widely used MKK system for classifying stellar spectra, an autonomous
scheme “without having to appeal to any theoretical picture.”48 As much as stars could be
47 I am indebted to R. Buta for highlighting these contributions (e-mail to author 22 February, 2015).
48 W. W. Morgan, A Morphological Life, Annual Review of Astronomy and Astrophysics, 1988, p. 4. The standard MKK system was based on a strict comparison of the intensities of well-picked spectral lines. The main physical determinant of line intensity is temperature, with secondary influences of abundance and gas pressure or gravity.
14:11:09, subject to the Cambridge Core
.011
202
Part III – Organizing the World of Galaxies
Fig. 9.5 Cross-section of the de Vaucouleurs system near class Sb. From Buta et al. (2007), The de
Vaucouleurs Atlas of Galaxies. Courtesy of Cambridge University Press.
easily classified, Morgan felt that the fantastic diversity of galaxies noted by Reynolds was
truly challenging. “The classification of the optical forms of galax
ies is a very different
problem from that of the classification of stellar spectra. The ordered appearance of the
spectrograms contrasts sharply with the apparent disorder of the forms of many galaxies,
even in the case of galaxies possessing a degree of order, we find them in fantastic variety.”49
This did not stop Morgan from giving it a try.
49 W. W. Morgan, A Morphological Life, Annual Review of Astronomy and Astrophysics, 1988, p. 6.
14:11:09, subject to the Cambridge Core
.011
9. The Galaxy Classification Play-Off
203
Fig. 9.6 The Morgan–Mayall classification system emphasizes the types of stars dominating the light
of the galaxies. Credit: From Morgan (1962) , The Astrophysical Journal. C
AAS. Reproduced with
permission.
The essence of the venture was to distinguish in galaxy spectra the types of stars that
were most effective in contributing to the integrated light. Working later with the American
astronomer Nicholas Ulrich Mayall (1906–1993), Morgan proposed, in 1957, a spectral
classification of giant galaxies using quantifiable properties of galaxies as opposed to visual
estimates employed for morphological screening.50 He and his colleagues used the spectra
of galaxies as a whole but isolated the spectrum of the nuclear region in cases where this
region was dominant; there could be noticeable spectral differences between the nucleus and
the surrounding regions. Using the characteristics of the galaxy spectrum in the blue spectral
region and the degree of central concentration of light, Morgan and Mayall’s spectroscopic
indicators produced seven groups (Fig. 9.6).
Because the relative size of the amorphous central region played a role, the system mim-
icked one of Reynolds’ and Hubble’s main criteria, the central concentration of brightness.
Despite its relative rigour, the Morgan–Mayall system was complicated and did not catch
on with the community. Some of its elements remain in use, for example the notation of “D
systems” for giant ellipticals (often associated with a strong radio source) surrounded by a
huge external envelope of stars. When located close to the centers of clusters of galaxies,
these giant ellipticals are called cD galaxies. Beyond this, the Morgan–Mayall spectro-
scopic scheme has seen little usage, probably because the integrated spectra of a galaxy
50 W. W. Morgan, Some Characteristics of Galaxies, The Astrophysical Journal, 1962, Vol. 135, pp. 1–10.
14:11:09, subject to the Cambridge Core
.011
204
Part III – Organizing the World of Galaxies
can be so diverse and gives little indication of the nature of the object, especially for the
neophyte or the less familiar researcher.
What does Classification Miss Out?
As will be discussed in the next chapter, many researchers raised caveats regarding classi-
fication schemes for galaxies, especially Hubble’s one. The Hubble classification uses the
appearance of galaxies on images. But these objects are three-dimensional. Orientation in
space (and dust obscuration) creates viewing aspects or silhouettes that are not necessarily
related to intrinsic physical properties. Furthermore, the appearance of a galaxy depends on
the wavelength of the light that can pass through the filters used (Chapter 4). Notwithstand-
ing these limitations, and whether the Hubble or the de Vaucouleurs scheme is used, the
morphological types correlate with several key physical properties: luminosities, masses,
colours, gas content and star-formation rate. Consequently, the morphological classification
schemes have been useful and continue to be, despite the somewhat subjective and arbitrary
steps of assigning galaxies to classes.
Another problem hampering a simple classification model is that there are many galaxies
in unusual states. As de Vaucouleurs warned, “It cannot be too strongly emphasized that
the regular, beautiful spiral patterns of such objects as M 31 [Messier 31], M 33, M 51,
M 81, M 101, etc. represent exceptions rather than the rule.”51 As will be seen in Chapter
10, several of the peculiar galaxies appear to be “normal” galaxies (spirals, ellipticals or
even irregulars) in the process of interacting or merging. Their shapes are distorted through
the formation of tidal tails and their appearances are modified by enhanced star formation
and dust lanes. No standard classification scheme can be applied to the “pathological” or
“freak” cases, terms used by Harlow Shapley and Walter Baade (see, e.g., Plates 6.6 and
11.1). However, with the aid of computer simulations, observers and theoreticians have
been able to identify the various stages of the merging process (Chapter 6).
Despite these caveats, it is generally accepted that galaxy morphology holds answers to
several fundamental questions. Ronald Buta reminds us of the main questions that hinge on
an understanding of galaxy formation and evolution: “Why is the Hubble sequence a con-
tinuous sequence, and what physical parameters underlie this continuity? What is the role
of angular momentum, dissipation, and galaxy–galaxy interactions? How do the various
patterns originate, and how do they change with time?”52 And a more difficult but exciting
question is how dark matter shapes galaxies.
Let us now review the atlases of galaxies published in the last 80 years, and the role they
played in framing the collective empiricism of students of galaxies.
51 G. de Vaucouleurs, Classifying Galaxies, Astronomical Society of the Pacific Leaflets, 1957, No. 341, p. 332.
52 R. Buta, Galaxy Morphology and Classification, in The World of Galaxies, H. G. Corwin Jr. and L. Bottinelli (editors), New York: Springer-Verlag, 1989, pp. 29–44. See also R. Buta, Galaxy Morphology, in Planets, Stars, and Stellar Systems, Volume 6, W. C. Keel (editor), Dordrecht: Springer, 2013, pp. 1–89; R. Buta, Galaxy Morphology, in Secular Evolution of Galaxies, J. Falcón-Barroso and J. H. Knapen (editors), Cambridge: Cambridge University Press, 2013, pp. 155–258.
14:11:09, subject to the Cambridge Core
.011
10
Atlases of Galaxies, Picturing “Island-Universes”
The human understanding, from its peculiar nature, easily supposes a
greater degree of order and equality in things than it really finds.
Francis Bacon 1
In fact, when looked at closely enough, every galaxy is peculiar.
Halton Arp 2
Why Are There So Many Galaxy Atlases?
I was 16 or 17 years old when I received my first major scientific book as a gift from
my mother; it was The Hubble Atlas of Galaxies by Allan Sandage. I could hardly read
English then, but the black-and-white halftone images of the atlas were stunning; they have
stayed imprinted in my mind. This galaxy atlas became for me, as for many other young
people of the 1960s, the beginning of an amazing journey, a career in astronomy. Reading
a few elementary astronomy books, and some slightly outdated texts, had already whet
my appetite. It was Larousse Encyclopedia of Astronomy by Lucien Rudaux and Gérard
de Vaucouleurs that got me hooked, a massive work that the young French astronomer de
Vaucouleurs had updated and refreshed in a most professional way following the death of
lead author Lucien Rudaux (1874–1947).3
Gérard de Vaucouleurs, who, as
we saw in the previous chapter, proposed a more detailed
classification scheme than Hubble’s one, arrived into astronomy as a keen and passionate
amateur. Young de Vaucouleurs was at first interested in the planets and asteroids of the
solar system. He met his wife and astronomy co-worker while both were studying at La
Sorbonne in Paris (Fig. 10.1). Antoinette (Piétra) de Vaucouleurs (1921–1987) became a
key collaborator in everything that Gérard was involved with.4 The time spent by the de
1 F. Bacon, Aphorism 45, Novum Organum, 1620. See Hanover Historical Texts Project on-line.
2 H. C. Arp, The Atlas of Peculiar Galaxies, Pasadena: California Institute of Technology, 1966.
3 The English version was a translation and revision from the French of Astronomie, les astres, l’univers, Paris: Librairie Larousse, 1948.
4 M. Capaccioli and H. Corwin Jr. (editors), Gérard and Antoinette de Vaucouleurs: A Life for Astronomy, Singapore: World Scientific, 1989.
205
14:11:09, subject to the Cambridge Core
.012
206
Part III – Organizing the World of Galaxies
Fig. 10.1 Gérard and Antoinette de Vaucouleurs in Paris, 1962. Image provided via Ken Freeman,
photographer unknown.
Vaucouleurs at the Commonwealth Observatory on Mount Stromlo in Australia, from 1951
to 1957, was most productive. Throughout his career, Gérard was passionate about and most
proficient in photography.5 He wrote several books and articles about it. The couple’s early
works established them as key contributors and leaders of galaxy astronomy. In addition to
exploring the morphology of galaxies and of the Magellanic Clouds, two important pieces
of research came out from their time in Australia: (1) the relationship between mass and
luminosity in elliptical galaxies, and (2) the evidence for the organization of galaxy clusters
into larger structures or superclusters.
In 1958, the de Vaucouleurs went to the United States to work first at the Lowell Observa-
tory, then Harvard College Observatory. They also visited the Mount Wilson Observatory
headquarters in Pasadena, California, as guests of Allan Sandage. Sandage gave Gérard