Old before their time

The discovery of massive, evolved galaxies at much greater distances than expected — and hence at earlier times in the history of the Universe — is a challenge to our understanding of how galaxies form.

Over the past two decades, astrophysicists have been spectacularly successful in explaining the early evolution of the Universe. Existing theories can account well for the time span from the Big Bang nearly 14 billion years ago until the Universe began to cool and form the first large structures less than a million years later. But detailed explanations of how the original stew of elementary particles subsequently coalesced over time, to form the stars and galaxies seen in the present-day Universe, are still being refined. As they report on pages 181 and 184 of this issue, Glazebrook et al.1 and Cimatti et al.2 have discovered the most distant ‘old’ galaxies yet. But the existence of these objects at such an early epoch in the history of the Universe seems inconsistent with the favoured theory of how galaxies formed.

That favoured theory is the so-called hierarchical model, in which smaller structures gradually accumulate into ever larger structures, ultimately forming galaxies of the sort we see today3. The most massive galaxies are expected to have formed relatively late in the process, with few existing before the Universe was half its present age. Such predictions can be tested, in principle, through the observations made of distant galaxies.

Nature has provided us with a powerful means of observing the history of the Universe: because the speed of light is finite, as we look out into space we actually peer back in time, seeing distant objects not as they are now, but as they were when their light was emitted millions or billions of years ago. Unfortunately, galaxies more than 6 billion light years away are not only exceedingly faint, but are also particularly difficult to identify. The visible galaxy spectra are ‘redshifted’ to longer, near-infrared wavelengths as a consequence of the expansion of the Universe; at these wavelengths, the Earth's atmospheric emission obscures the key spectral ‘fingerprints’ that are commonly used to identify galaxies.

For these reasons, virtually all of the galaxies known from the early days of the Universe are those that are still forming new stars, and hence emitting copious amounts of light4. Although easier to find, such galaxies are not particularly useful for testing theories of galaxy formation because it is impossible to set strong lower limits on how old they are. However, finding significant numbers of massive, evolved galaxies (which finished forming stars long ago) at distances that correspond to half the present age of the Universe would indicate that such galaxies formed much earlier than the leading theory predicts.

Several earlier studies5,6,7,8 have found evidence for a population of evolved galaxies in the distant past. These studies used the colours of galaxies as rough estimators of their distance — a method that is easier but also much less accurate than measuring galaxy distances directly through the redshift of their spectrum. By pushing some of the largest ground-based telescopes to their limits, Glazebrook et al.1 and Cimatti et al.2 have now provided the most compelling evidence yet that ‘old’ (that is, evolved) massive galaxies were numerous at early epochs. Using the 8-metre telescopes at the European Southern Observatory in Chile to survey a small region of the sky in detail, Cimatti et al.2 report the discovery of four massive, evolved galaxies, all of which are considerably more distant than the previous record holder. Supplementary images from the Hubble Space Telescope (Fig. 1) show that these galaxies appear to be massive and old, both structurally and spectroscopically.

Figure 1: Journey to the early Universe.


This image, taken in visible light by the Hubble Space Telescope, shows a plethora of galaxies billions of light years away in a random patch of sky called the Hubble Ultra Deep Field. As part of a survey of galaxies in this region, Cimatti et al.2 have found several massive galaxies that were already fully assembled billions of years before they should have been, according to current theories. A complementary survey in the northern sky by Glazebrook et al.1 has revealed similar examples of such galaxies, posing a problem for theories of galaxy formation.

Complementing this discovery is the ambitious Gemini ‘Deep Deep Survey’, which was performed using the Gemini observatory's Hawaii-based 8-metre telescope. This study1 is notable both for the exceptionally long exposures obtained (30 hours for each target) and for the use of an innovative observing mode, which reduces background noise exceptionally well. As a result, Glazebrook et al.1 were able to measure redshifts for far fainter galaxies than is possible by conventional means. As well as strengthening the evidence that massive, evolved galaxies were a significant component of the young Universe, the Gemini team has estimated the change in the abundance of such objects since that time, by observing additional galaxies at intervening distances. They conclude that, going from the present day back to the earliest epochs they probe, the abundance of massive galaxies decreases much more slowly than predicted by the hierarchical model.

With this first solid confirmation1,2 that as far back as 10 billion years ago there were already many old massive galaxies, it is clear that even the best models can't fully explain the evolution of galaxies. These studies are forcing astronomers to consider whether massive galaxies grew much earlier than predicted by the hierarchical model, or whether the stars in these earliest galaxies formed in a substantially different way from our expectations9. As well as providing the motivation to explore new models of galaxy evolution, this is a tantalizing first look at the type of science that will become routinely possible as the next generation of even larger telescopes come online in ten years' time.


  1. 1

    Glazebrook, K. et al. Nature 430, 181–184 (2004).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Cimatti, A. et al. Nature 430, 184–187 (2004).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Blumenthal, G. R. et al. Nature 311, 517–525 (1984).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Steidel, C. C., Adelberger, K. L., Giavalisco, M., Dickinson, M. & Pettini, M. Astrophys. J. 519, 1–17 (1999).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Dickinson, M., Papovich, C., Ferguson, H. C. & Budávari, T. Astrophys. J. 587, 25–40 (2003).

    ADS  Article  Google Scholar 

  6. 6

    Fontana, A. et al. Astrophys. J. 594, L9–L12 (2003).

    ADS  Article  Google Scholar 

  7. 7

    Drory, N. et al. Astrophys. J. 562, L111–L114 (2001).

    ADS  Article  Google Scholar 

  8. 8

    Franx, M. et al. Astrophys. J. 587, L79–L82 (2003).

    ADS  Article  Google Scholar 

  9. 9

    Somerville, R. S. in Proc. ESO/USM/MPE Workshop on Multiwavelength Mapping of Galaxy Formation and Evolution (eds Bender, R. & Renzini, A.). Preprint at http://arxiv.org/abs/astro-ph/0401570 (2004).

    Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Wirth, G. Old before their time. Nature 430, 149–150 (2004). https://doi.org/10.1038/430149a

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.