The Role of Images in Astronomical Discovery
Page 29
train the practitioner how to see scientific objects in unison. “The schooling of the senses
is probably not qualitatively different from that undergone by the fledgling musician, cook,
or weaver – as Aristotle noted, the paths to skill, on the one hand, and to understanding,
on the other, pass through the same stations of perception, memory, and experience. But
the scientific path is greatly straightened by the demands of collective empiricism, which
require a degree of coordination seldom achieved (or desired) in the traditional arts and
crafts.”5
Starting with Flemish cartographer and mathematician Gerardus Mercator (1512–1594),
the word “atlas” came to refer to large sized paper. “But it was in the nineteenth century
that the oversized sheets lent their name to a much wider species of literature, the genre
of systematic presentation of scientific pictures. There were atlases of every sort. From the
pathology of the human body to African plants, from atlases of fossils to ones of micro-
graphs and X rays.”6 Among books and tools of learning, scientific atlases are indeed piv-
otal; they are specialized products that must meet the demands of insistent communities.
There is even a market for atlases, and some have even talked about their “economy.”
Atlases are different from guidebooks and encyclopedia. A guidebook is a companion
tool to recognize natural objects while an encyclopedia treats the whole knowledge of a field
in a comprehensive manner, giving broader coverage of topics. The atlas is more systematic
than a field guide and does not aim to comprehensively cover the whole discipline. To
prepare for an in-depth discussion of how atlases of galaxies were put together (Chapter
10), I want to examine first how the classification of galaxies was addressed.
The Purpose of Atlases: Shaping Galaxies
The current classification scheme of galaxy shapes emerged in the first three decades of
the twentieth century. To see objects “in unison” requires some organizational scheme.
The classification of objects is one of the driving forces for creating scientific atlases; in
reverse, the atlas images highlight and illustrate the classification scheme. A small number
of shared properties are used to sort the objects into categories. As emphasized by Steven J.
Dick, a “class of objects” in astronomy is different from a “species” in biology, and is much
less distinctive than elements in chemistry or elementary particles in high-energy physics.
While there are no transitions between elements and particles, there are intermediate states
in objects of biology and astronomy. A main-sequence star, like the Sun, will transition
to be a red giant, later a planetary nebula, and finally a white dwarf. An electron will not
become a proton, nor will hafnium transform into tungsten; it is that or something else.
“Thus, astronomical classes retain some of the ambiguities of biological species, but not
5 L. Daston, On Scientific Observations, Isis, 2008, Vol. 99, No. 1, p. 107.
6 P. Galison, Image & Logic, A Material Culture of Microphysics, Chicago: University of Chicago Press, 1997, p. 131.
14:11:09, subject to the Cambridge Core
.011
190
Part III – Organizing the World of Galaxies
their evolutionary mechanisms, introducing both practical and theoretical problems with
classification.”7
Classification is a sort of epistemological art. As beautifully stated by Gérard de Vau-
couleurs, “one of the first tasks which confronts the student of almost any category of
objects – atoms, molecules, plants, animals, stars, galaxies, etc. – is to arrange them in
some classification scheme bringing in order and logic where nature offers only a bewil-
dering variety of individuals.”8 American astronomer Allan Sandage (1926–2010) put for-
ward some useful considerations. To have a “good” classification system, “some type of
significant ordering is necessary; ideally it will disclose the underlying true connections
among characteristics of the manifold objects.”9 Sandage then questioned what we mean
by “significant” and by “true,” then goes on to devote a page and a half to how the “nearly
un-surmountable challenge” can be met. For galaxies, the shared property at the root of the
classification scheme, that has been so far most favored, has been and continues to be mor-
phology or the three-dimensional shape of a galaxy. For Sandage, de Vaucouleurs and other
classifiers, morphology reveals “true connections.” More recently, supplementary classifi-
cation criteria using dynamic properties (e.g. counter-rotation) have been proposed.
As we saw in Chapter 5, the main shapes of non-galactic “nebulae” were recognized rela-
tively early: first, spirals by William Parsons in 1845, then, ellipticals by Stephen Alexander
(1806–1883) in 1852,10 and much later, Magellanic or irregular systems by Knut Lundmark
in 1922. The speculative work of Alexander had a striking dimension. A supporter of the
Nebular Hypothesis, Alexander viewed most nebulae as systems in disintegration, which
followed, rather than preceded, stellar formation (see Fig. 2.11). Attempting a grand order-
ing scheme in 1919, Heber Curtis threw galactic and non-galactic “nebulae” in the same
bag, not unlike the scheme proposed by Max Wolf in 1908 (Chapter 4; Fig. 4.4). Curtis
divided all nebulae into three groups: planetaries, diffuse nebulae in the galactic plane, and
spirals in the “extragalactic” region. The distinction between “in the plane” and “out of
the plane” (of the Milky Way) stuck for some time. When using the term “extragalactic
nebulae,” Hubble carefully made this distinction, without at first committing fully to the
notion that spirals were external. However, Curtis clearly considered spirals to be a special
group, systems external to the Milky Way, and the most abundantly observed, in the order
of millions. Until the adoption of Hubble’s orderly scheme, classification exercises were
more like shots in the dark.
Along Came Edwin Hubble
The merit of Hubble resides not in his invention of the fundamental categories but the fact
that he saw order enough to be able to sequence galaxy shapes in “significant” ways. Hubble
7 S. J. Dick, Discovery and Classification in Astronomy, Controversy and Consensus, Cambridge: Cambridge University Press, 2013, p. 239.
8 G. de Vaucouleurs, Classifying Galaxies, Astronomical Society of the Pacific Leaflets, 1957, No. 341, p. 329.
9 A. R. Sandage, Centennial History of the Carnegie Institution of Washington, Volume 1: The Mount Wilson Observatory, Cambridge: Cambridge University Press, 2004, p. 230.
10 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. The work appeared in a series of articles running through several issues.
14:11:09, subject to the Cambridge Core
.011
9. The Galaxy Classification Play-Off
191
brilliantly picked out and highlighted meaningful variations among the non-galactic “nebu-
lae.” In this way, it can be said that Hubble’s “tuning fork” diagram for galaxy classification
clarified and merged earlier schemes (see Plate 9.1). More systematic and thorough than its
predecessors, Hubble’s
scheme has remained the paradigm of galaxy classification, based
on morphology. It has survived important modifications and refinements later brought to it,
most notably by the French astronomer Gérard de Vaucouleurs (1918–1995).
Hubble’s dichotomous diagram was a powerful tool for distinguishing the categories
of galaxies. Disk galaxies differ from ellipticals by their distinct three-dimensional shape
and overall dynamics. For both classes, the degree of flattening and light concentration has
a physical meaning and possibly relates to events that happened at the time of formation
(Chapter 6). Among disk systems, bar spirals are distinguishable from normal spirals by
features that indicate differences in large-scale stellar and gas kinematics. Hubble envis-
aged – and he was right – morphology as a tracer of fundamental processes underlying
the formation, assembly and evolution of galaxies. Putting aside his usual circumspection,
Hubble went even further. Influenced by the ideas of the British astrophysicist James Jeans
(1877–1946) on nebular evolution, he thought that his “sequence” of shapes was evolution-
ary: in his view, galaxy systems started as ellipticals; with time, they became flattened disks.
However, he wisely refrained from making this shaky (and incorrect) theoretical connection
when he proposed his classification scheme.11
The great merit of Hubble’s scheme was its relative simplicity and the ease with which
it can be illustrated with spectacular images. For Hubble, as for the German biologist and
naturalist Ernst Haeckel, “what appeared to be so complicated needed to be made plain –
morphology, form were the tools.”12 This inspired the taxonomists of galaxies and led to the
production of practical and splendid atlases. The Hubble Atlas of Galaxies, first produced
by Allan Sandage in 1961, was a masterpiece that set the standard and tone for followers
(Chapter 10).
In summary, driving forces for galaxy atlases were to reflect contemporary knowledge
about the world of galaxies and to educate students and researchers new to the field. Fur-
thermore, by adopting Hubble’s relatively simple scheme, authors of galaxy atlases could
serve a wide set of purposes as we will see in Chapter 11. Atlases also provide working
objects for research programs. But they do more than that: they help to build communities
of practitioners. Daston discussed extensively a great classic of functional scientific atlases:
the International Cloud Atlas – the 1896 trilingual version with text in English, French and
German – produced for meteorologists of the world to see clouds “in unison.”13 The first
edition featured several printed colour plates, instead of hand-coloured plates, and even
included an unusual colour photograph of a cirrus cloud. This century-old scholarly work
is famous for its many successive editions and for being issued in several languages. Like
this cloud atlas and atlases of other scientific disciplines, galaxy atlases aimed to assemble
“standardized objects” for researchers and students. Carl Størmer’s Photographic Atlas of
11 See discussion in R. Smith, The Expanding Universe, Astronomy’s ‘Great Debate’ 1900–1931, Cambridge: Cambridge University Press, 1982, pp. 147–151.
12 O. Breidbach, Visions of Nature: The Art and Science of Ernst Haeckel, Munich: Prestel, 2006, p. 274.
13 L. Daston, On Scientific Observations, Isis, 2008, Vol. 99, No. 1, pp. 97–100.
14:11:09, subject to the Cambridge Core
.011
192
Part III – Organizing the World of Galaxies
Auroral Forms attempted the same.14 Atlases are like sorts of bibles that bring practitioners
to a common “creed.” On the practical side, they are the handbooks that guide day-to-day
work on galaxies.
Through its production and usage, a scientific atlas represents the epitome of scientific
observation (see Introduction), building communities by observing together. Lorraine Das-
ton and Elizabeth Lunbeck have brilliantly described this process as collective empiricism.
It is “the collection, transmission, and distillation of the experience of many inquirers into
weather proverbs, astronomical tables, medical regimens, botanical descriptions, socioeco-
nomic statistics, and a myriad of other findings.”15 The atlas became the tool to “calibrate
the eyes and hands” of the dispersed community of galaxy observers and students (more on
this in the next two chapters).
Classification for a Purpose
All atlases use and highlight a classification system of the scientific objects of interest. They
also do more. They try to convey lots of information: for galaxies, images can be supported
by quantitative photometric measurements in order to disentangle the mix of structures. The
photometric maps pinpoint the different kinematic or evolutionary components of galaxies.
Taken at different orientation angles, radial light profiles help to separate the central bright
nuclei and amorphous bulges from the flatter and more extended exponential disks; and
they provide more specific details, such as the spatial enhancements of star concentration, or
bars/ovals/rings that highlight dynamic instabilities. Finally, the strength of dark lanes gives
a sense of the amount and distribution of dust in a galaxy. To the perspicuous eye and trained
user, atlas images illustrate the effect of internal perturbations (bars), secular evolution,
gas accretion and nuclear activity. By organizing all of this, a classification scheme helps
to interpret morphology as it relates to environmental density and the merging history of
galaxies.
As the morphology provides clues on the formation and evolution of the gas and stars of a
galaxy, products of computer simulations can be related to the observed shapes of archetypal
objects. Hence, atlases, assembling sequences of observed configurations, become useful
collections of working objects for theoreticians and simulators to test their models on.
Brief Historical Overview of a Meandering Process
Hubble was not the first to tackle the challenge of classifying nebulae. To grasp the success
of his scheme, it is useful to recall a bit of history. The first attempts at classifying “nebulae”
can be seen in the 1811 drawings of William Herschel (Fig. 2.4) and, more spectacularly, in
the publications of works conducted later under the leadership of William Parsons, the Third
14 C. Størmer, Photographic Atlas of Auroral Forms, Oslo: International Geodetic and Geophysical Union, 1930.
15 L. Daston and E. Lunbeck, Observing Together: Communities, in Histories of Scientific Observation, Chicago: University of Chicago Press, 2011, p. 369.
14:11:09, subject to the Cambridge Core
.011
9. The Galaxy Classification Play-Off
193
Earl of Rosse, at Birr Castle, Ireland, where a spiral structure was found in several “nebulae”
(Fig. 0.2). As we saw in Chapter 2, Parsons encouraged the faithful visual rendition of
“nebulae” through drawings made by careful observers using the large Birr Castle reflectors
over three decades (1848–1878).16,17 The Birr Castle observers found spirality in dozens of
other “nebulae.” Although some cases turned out to be false identifications, most of these
objects were later confirmed by photography to be spiral galaxies.18,19 The nineteenth-
/>
century drawings captured the essential shape and important details of the galaxy structures
and of the satellite galaxies surprisingly well.
In a rather long-winded article in 1852, the American astronomer and mathematician
Stephen Alexander tried to make a grand synthesis of all nebular forms and star clusters:
“spheroid,” “elliptic,” “annular,” “spiral,” “globular” (see Fig. 2.11).20 He illustrated all his
examples with drawings from Birr Castle and others produced by John Herschel. He tried to
organize “nebulae” by their shape and even predicted hypothetical shapes. The main merit
of Alexander’s highly speculative article was the recognition of the spheroidal/ellipsoidal
shapes of non-galactic nebulae.
The British amateur astronomers Andrew Ainslie Common and Isaac Roberts used pho-
tography and their own observations based on telescopes with improved designs and glass
mirrors (Fig. 3.1). The advent of sophisticated telescope mounts as well as automatic drives
revolutionized the observation of faint objects, in particular the “nebulae.” The use of neg-
ative prints was favored: a negative print reproduced the details with a greater fidelity and
was better able to handle the wide contrast between the very bright nucleus (generally sat-
urated), the bright bulge and the faint outer arms or inter-arm regions. Roberts published
two spectacular books of his photographs of “nebulae” and star clusters.21 They did not
attempt to classify, nor were they atlases in the scholarly sense, but they defined the way
forward.
It was photography, not drawing, that became the obvious means for identifying cat-
egories. Max Wolf used photographic material to establish an early system of classes of
“nebulae” (Chapter 4). He did not differentiate between galactic and non-galactic nebulae
when proposing 23 classes distinguished by alphabetical letters a to w (with a h class).
o
Today, classes o to w can be recognized as being spirals with different inclinations and
strength of dust lanes, while a to n appear to be a mix of planetary nebulae and non-galactic
spheroidal systems. However, Wolf had no idea of the nature of the objects he was trying to
organize. His scheme remained a temporary “filler” and of marginal interest until something