by Rene Roy
better came along.
16 Third Earl of Rosse, On the Construction of Specula of Six-Feet Aperture: and a Selection from the Observations of Nebulae Made with Them, Philosophical Transactions of the Royal Society of London, 1861, Vol. 151, pp. 681–745.
17 Fourth Earl of Rosse, Observations of Nebulae and Clusters of Stars made with the Six-foot and Three-foot Reflectors at Birr Castle, from the year 1848 up to about the year 1878, The Scientific Transactions of the Royal Dublin Society, 1879, Vol. 11.
18 D. W. Dewhirst and M. Hoskin, The Rosse Spirals, Journal for the History of Astronomy, 1991, Vol. XXII, pp. 257–266.
19 Ronald Buta did an analysis of the Birr Castle observations in 2010 and concluded that the Birr observers saw spiral structure in 75 bright galaxies.
20 S. Alexander, On the Origin of the Forms and the Present Condition of Some of the Clusters of Stars, and Several of the Nebulae, The Astronomical Journal, 1852, Vol. 2, pp. 95–160.
21 I. Roberts, A Selection of Photographs of Stars, Star-clusters and Nebulae, Vol. I, London: The Universal Press, 1893; I. Roberts, A Selection of Photographs of Stars, Star-clusters and Nebulae, Volume II, London: “Knowledge” Office, 1899.
14:11:09, subject to the Cambridge Core
.011
194
Part III – Organizing the World of Galaxies
Then, being in possession of photographs of nearly 500 spirals, Heber Doust Curtis
presented the first major compendium of photographic galaxy images in 1918, with a special
focus on spirals.22,23 The Lick astronomer even thought that all non-galactic “nebulae”
were spirals, predicting – wrongly – that ellipticals would be seen as spirals with higher
angular resolution. Curtis illustrated the shape of the spirals with attention to the dust, and
correctly inferred that it affected the appearance of the objects. Plates of dozens of spirals
photographed with the 36-inch Crossley telescope were presented as a mini atlas at the
end of the long article. The images were displayed as sequences of shapes to illustrate the
relative importance and spatial distribution of internal dust.
A Contentious Path to Galaxy Classification
John Henry Reynolds (1874–1949) was a highly successful British industrialist – another
astronomy dilettante businessman and gentlemanly scientist – who was passionate about
astronomy.24 He had constructed his own 28-inch telescope and he served as the president
of the Royal Astronomical Society. Unusually, Reynolds played in the same field as pro-
fessionals. He published his own classification of spiral galaxies (calling them “spindle”
nebulae) in 1920, well ahead of Hubble, who corresponded with him.25 On the basis of the
degree of central condensation and granularity of the “outer whorls” or structure of the cir-
cular pattern, Reynolds had created seven classes of spirals some of which turned out to be
identical to Hubble’s. Hubble referred much later and only in passing to Reynolds’ works,
and not to the original work that must certainly have inspired him.26
It is well known that Hubble was extremely sensitive regarding the priority of discov-
ery and of the proposal of concepts. He did not easily accept nor recognize competition.
This is clearly highlighted by the ensuing Hubble–Lundmark controversy that showed a
trait of Hubble’s character.27 This confrontation was also a sign that galaxy classification
had become an intense battlefield of clashing concepts – and personalities. Hubble came
up with his famous “tuning fork” diagram in the mid 1920s (Plate 9.1).28 However, as
early as 1920, the Swedish astronomer Knut Lundmark (1889–1958) had proposed a clas-
sification system of “anagalactic” nebulae that was quite analogous to Hubble’s scheme
(Fig. 9.2).29 Lundmark’s scheme contained some slight but important differences,
22 H. D. Curtis, Descriptions of 762 Nebulae and Clusters Photographed with the Crossley Reflector, Publications of the Lick Observatory, 1918, Vol. XIII, Part I, pp. 9–42.
23 H. D. Curtis, A Study of Occulting Effects in the Spirals, Publications of the Lick Observatory, 1918, Vol. XIII, Part II, pp. 43–55.
24 A fine review and assessment of John H. Reynolds’ work and contributions are presented in D. L. Block and K. Freeman, Shrouds of the Night, Masks of the Milky Way and Our Awesome New View of Galaxies, New York: Springer, 2008, pp. 183–213.
25 John H. Reynolds, Photometric Measure of the Nuclei of some Typical Spiral Nebulae, Monthly Notices of the Royal Astronomical Society, 1920, Vol. 80, pp. 746–753.
26 D. L. Block and K. Freeman, op. cit., 2008, pp. 200–203.
27 This episode has been particularly well discussed in Part III, Chapter 6 of R. Berendzen, R. Hart and D. Seely, Man Discovers the Galaxies, New York: Science History Publications, 1976.
28 E. P. Hubble, Extra-Galactic Nebulae, Contributions from the Mount Wilson Observatory, Carnegie Institution of Washington, 1926, Vol. 324, pp. 1–49; Extra-galactic Nebulae, The Astrophysical Journal, 1927, Vol. 64, pp. 321–369.
29 Anagalactic was a word equivalent to non-galactic used mainly by European astronomers. Galaxies was the word used almost immediately after Hubble’s death. Hubble stuck to the end to “extra-galactic nebulae.”
14:11:09, subject to the Cambridge Core
.011
9. The Galaxy Classification Play-Off
195
Fig. 9.2 Knut Lundmark as a student in 1908. Credit: unknown photographer at H. Tegström & Co.
especially in emphasizing the degree of light concentration towards the center. In 1926,
a furious Hubble wrote to Lundmark, accusing him of plagiarism and of having stolen
his classification scheme following the 1925 Cambridge Astronomical International Union
(IAU) meeting, where Hubble had presented his proposal.
As shown by the Finnish astronomer Pekka Teerikorpi, Hubble had incorrectly and
unfairly accused Lundmark.30 Teerikorpi demonstrates that Lundmark had worked on
galaxy classification some years before his 1926 paper; the concepts of his system and the
sequence of the main classes had appeared in print before the 1925 Cambridge IAU meet-
ing. Finally, “Lundmark’s and Hubble’s systems were in fact so different that they hardly
could have served as models for each other’s.” Allan Sandage also thought Hubble had
overreacted and treated Lundmark unfairly. Sandage wrote that the “exchanges poisoned
the association between Hubble and Lundmark for years. It would not be mended until the
30 P. Teerikorpi, Lundmark’s Unpublished 1922 Nebula Classification, Journal for the History of Astronomy, 1989, Vol. 20, pp. 165–170.
14:11:09, subject to the Cambridge Core
.011
196
Part III – Organizing the World of Galaxies
1950s, when Hubble invited Lundmark to inspect the archival plate file of galaxies that had
been photographed by the new 200-inch Palomar reflector.”31
This clash certainly illustrates how possessive scientists can be of their ideas, but it high-
lights the importance galaxies were now taking among professionals as a rapidly developing
research field. Also, the Lundmark–Hubble skirmish shows that a good concept may bur-
geon simultaneously in several minds which are trying to tackle the same problem. For
example, James Jeans had come up with a “Y” classification to emphasize that the “great
nebulae” could not all be placed in a continuous sequence.32 Some say that the Hubble tun-
ing fork could be viewed as just an extended Y scheme. “In popular astronomical textbook
s,
we only read of the Hubble classification classes, the Hubble tuning fork and the Hubble
luminosity profile for elliptical galaxies. Behind stage loomed the giant Mr. Reynolds, a
man whose name is almost unknown to students of astronomy.”33 There goes the whiggish
history of science. Mark Twain had already denounced Stigler’s law of eponymy in harsher
words: “It’s just the way, in this world. One person does the thing, and the other one gets
the monument.”
The Hubble Sequence
Let us come back to the Hubble system of galaxy classification, often called the “Hubble
sequence” (Plate 9.1). Although Hubble himself would have liked it to be so, the “sequence”
does not represent an evolutionary trend. Hubble divided galaxies into three broad classes,
following their appearance at visible wavelengths: ellipticals, lenticulars and spirals. He
added a fourth class, the irregulars (or Lundmark’s “Magellanic”), characterized by a lack
of symmetry. Indeed, many galaxies could not be classified; several of these unclassifiable
cases gave rise to debate and controversy and were later recognized as “peculiar.”
On the Hubble sequence, spirals divide into two categories, normal (e.g. Messier 81,
Plate 7.1) and barred (e.g. NGC 7424, Plate 6.4), giving two branches, hence, “tuning fork.”
Hubble’s tuning-fork diagram runs from the left with the ellipticals, with the lenticulars at
the center, and to the right, the two branches of the fork of the normal spirals (top branch)
and barred spirals (bottom branch). The irregulars (e.g. NGC 4449, Plate 6.2) are hanging
at the right for apparent lack of morphological connection with the rest. Hubble nicely
presented his perspectives and the purposes of his scheme in a fine little book, The Realm
of Nebulae. The 1936 book is a classic and remains a most useful source in the context of
historical development.34
Let me add a few technical details on the Hubble classification that will be important
for the discussion of atlases of galaxies in the next chapter. The symbol E, for elliptical,
31 A. R. Sandage, Centennial History of the Carnegie Institution of Washington,Volume 1: The Mount Wilson Observatory, Cambridge: Cambridge University Press, 2004, p. 488.
32 J. Jeans, Astronomy and Cosmogony, Cambridge: Cambridge University Press, 1928, Figure 53.
33 D. L. Block and K. Freeman, Shrouds of the Night, Masks of the Milky Way and Our Awesome New View of Galaxies, New York: Springer, 2008, p. 212, citing Mark Twain from Tom Sawyer’s Conspiracy (published posthumously).
34 E. P. Hubble, The Realm of Nebulae, New Haven: Yale University Press, 1936.
14:11:09, subject to the Cambridge Core
.011
9. The Galaxy Classification Play-Off
197
is followed by an integer (0 for spheroidal to 7 for the most flattened ellipsoid), which
indicates the degree of ellipticity multiplied by 10; it is defined the normal way for an
ellipse as e = 1 – ( b/ a), a and b being the semi-major and semi-minor axis; E is 10 × e.
Ellipticals are truly ellipsoidal structures in three-dimensional space. At the center of the
fork are the lenticulars denoted by the symbol S0. They have a bright, large, central bulge,
very similar to elliptical systems, and are surrounded by a disk-like structure as in spirals.
Lenticulars, although recognized as a separate class, were not understood in Hubble’s time,
and many successive researchers scratched their heads over them until they fell into place
quite recently (Chapter 11). Lenticulars look like the smooth spindle-shaped galaxies seen
at the top and bottom edge of the field on Plate 6.3.
Spirals display the most spectacular shapes. They are multi-component systems, dom-
inated by a flattened disk, with gas and star-forming regions concentrating generally in
a two-armed spiral structure. Spirals also have a bright central region, the bulge. Harlow
Shapley and John S. Paraskevopoulos subdivided the Sc end of the spiral sequence into Sc
and Sd to describe systems with greater arm openness and a lack of central condensation.
Thus, spirals are denoted by the symbol S followed by a, b, c or d as suffixes.35 Those
with bigger bulges and smoother arms are the Sa spirals, while Sd spirals show loose frag-
mentary arms and have no bulge. The suffixes apply to barred spirals as well, and are then
denoted as SBa, . . . , SBd.
Because of their lack of bulge or of structural symmetry, irregular galaxies do not fit
into the Hubble sequence. Hubble had a hard time with these intractable entities. Gérard
de Vaucouleurs, who later refined Hubble’s scheme, managed to integrate them into his
global scheme in a less arbitrary fashion. De Vaucouleurs divided irregulars into those with
some faint spiral structure as Irr I (e.g. the Large Magellanic Cloud, Plate 6.5) and irreg-
ulars with smooth features as Irr II; those with no obvious structure were described as Im
(e.g. the Small Magellanic Cloud). Hence, de Vaucouleurs recognized clearly the category
invented by Lundmark in 1922. Another Swedish astronomer, Erik Holmberg (1908–2000),
introduced finer divisions.36 All this fine-tuning of viewing galaxy shapes led to the entire
classes for spirals used today: Sa, Sab, Sb, Sbc, Sc, Scd, Sd, Sdm, Sm and Im, the latter
two corresponding to Lundmark’s Magellanic types. Finally, ellipticals and lenticulars are
sometimes referred to as “early type” while spirals and irregulars are “late type.” This unfor-
tunate nomenclature has no meaning in terms of temporal evolution. It is very confusing
and should be avoided.
The modern classification system of galaxies is not without parallel to the binomial
nomenclature for naming living species introduced by Carl Linnaeus in 1753.37 Linnaeus
used a binomial name, where the first part identifies the genus of the species and the second
part identifies the species within the genus. Galaxy shapes are related in much simpler ways
35 H. Shapley and J. S. Paraskevopoulos, Galactic and Extragalactic Studies, III: Photographs of Thirty Southern Nebulae and Clusters, Proceedings of the National Academy of Sciences of the United States, 1940, Vol. 26, pp. 31–36.
36 E. Holmberg, A Photometric Study of Nearby Galaxies, Meddelanden fran Lunds Astronomiska Observatorium, 1950, Ser. II, 128, 1.
37 C. Linnaeus, Species Plantarum, Stockholm: Holmiae, Impensis Laurentii Salvii, 1753.
14:11:09, subject to the Cambridge Core
.011
198
Part III – Organizing the World of Galaxies
than living species, but the galaxy classes Sa, SBc, S0 or E1, have an analogous binomial
structure that is a powerful visual descriptor. Praising Hubble’s scheme, Allan Sandage
has summarized it well: “The great merit of Hubble’s system is the bin size of the clas-
sification boxes.”38 Big enough to be significantly inclusive and narrow enough that even
the neophyte can apply the scheme successfully. However, as we will see later, Gérard de
Vaucouleurs found the bin sizes of Hubble’s original system too broad, and applied a finer
division.
Defining a Systematic Sample
A useful atlas depends on a well-defined and agreed sample of objects. And the sample must
be large enough. The Harvard astronomers Shapley and Ames published “A Survey of the
External Galaxies Brighter than the Thirteenth Magnitude” in 1932.39 This catalogue
of
1,249 bright galaxies was based on a photographic survey conducted from Arequipa, Peru,
using two tiny telescopes, the 2-inch Ross–Tessar and 2-inch Zeiss–Tessar. Small lenses
indeed, but they served two clear purposes: the small scale of the photographs gave the
advantage that most nebular objects looked sufficiently stellar for reliable brightness inter-
comparisons and precise apparent magnitude determinations.
With their photometric catalogue, Shapley and Ames aimed “to provide a systematic
census of the inner parts of the metagalactic system where heretofore no comprehensive
photometry has been available.” Although it had no images, the Shapley–Ames survey may
be considered as the seed for most galaxy atlases that were to be produced in the following
80 years. It provided the basic sample for the study of “the uniformity of distribution of the
galaxies, the clustering of the nearer systems, the relation of the apparent distribution to the
obscuring clouds in low galactic latitude, and similar problems.”
The Shapley–Ames catalogue included a discussion on the sizes and shapes of galaxies
“having Hubble as the authority for the description.” The catalogue also gave an overview
of several early schemes of galaxy classification (those of Wolf, Curtis, Reynolds, Rein-
muth and Hubble) and of their cross-correspondence. Karl Reinmuth had adopted Wolf’s
scheme (Fig. 4.4).40 Catalogues are essential to maintaining the usage of a classification
system. They provide systematic lists of objects as complete sets or representative ones.
This certainly worked most fruitfully for the Shapley–Ames catalogue.
The Shapley–Ames catalogue sample was indeed the basis of several later works, in
particular the long-term program conducted by Allan Sandage and collaborators at the
Carnegie Observatories. The later revisions of the catalogue ( The Revised Shapley–Ames
Catalog of Bright Galaxies, or RSA) by Sandage and Tammann (1981, 1987) were solidly
38 A. R. Sandage, Centennial History of the Carnegie Institution of Washington,Volume 1: The Mount Wilson Observatory, Cambridge: Cambridge University Press, 2004, p. 489.
39 H. Shapley and A. Ames, A Survey of the External Galaxies Brighter than the Thirteenth Magnitude, Annals of the Astronomical Observatory of Harvard College, 1932, Vol. 88, pp. 41–76.