Astronomers have used observations from Hubble’s CANDELS survey to explore the sizes, shapes, and colours of distant galaxies over the last 80% of the Universe’s history. In the Universe today galaxies come in a variety of different forms, and are classified via a system known as the Hubble Sequence — and it turns out that this sequence was already in place as early as 11 billion years ago.
The Hubble Sequence classifies galaxies according to their morphology and star-forming activity, organising them into a cosmic zoo of spiral, elliptical, and irregular shapes with whirling arms, fuzzy haloes and bright central bulges. Two main types of galaxy are identified in this sequence: elliptical and spiral, with a third type, lenticular, settling somewhere between the two.
This accurately describes what we see in the region of space around us, but how does galaxy morphology change as we look further back in time, to when the Universe was very young?
"This is a key question: when and over what timescale did the Hubble Sequence form?" says BoMee Lee of the University of Massachusetts, USA, lead author of a new paper exploring the sequence. "To do this you need to peer at distant galaxies and compare them to their closer relatives, to see if they too can be described in the same way."
The astronomers used Hubble to look 11 billion years back in time to when the Universe was very young, exploring the anatomy of distant galaxies.
While it was known that the Hubble Sequence holds true as far back as around 8 billion years ago , these new observations push a further 2.5 billion years back in cosmic time, covering a huge 80% of the past history of the Universe. Previous studies had also reached into this epoch of the cosmos to study lower-mass galaxies, but none had conclusively also looked at large, mature galaxies like the Milky Way. The new CANDELS observations confirm that all galaxies this far back — big and small alike — fit into the different classifications of the sequence.
"This is the only comprehensive study to date of the visual appearance of the large, massive galaxies that existed so far back in time," says co-author Arjen van der Wel of the Max Planck Institute for Astronomy in Heidelberg, Germany. "The galaxies look remarkably mature, which is not predicted by galaxy formation models to be the case that early on in the history of the Universe."
The galaxies at these earlier times appear to be split between blue star-forming galaxies with a complex structure — including discs, bulges, and messy clumps — and massive red galaxies that are no longer forming stars, as seen in the nearby Universe .
Galaxies as massive as the Milky Way or more are rather rare in the young Universe. This scarcity has prevented previous studies from being able to gather a large enough sample of mature galaxies to properly describe their characteristics.
What was needed was a systematic set of observations such as those from Hubble's CANDELS survey, which was large enough to allow the astronomers to analyse a larger number of these galaxies consistently, and in detail . With Hubble's Wide Field Camera 3 (WFC3), the astronomers were able to observe in the infrared part of the spectrum to see how the galaxies appeared in their visible rest-frame , which is easier to compare with galaxies in our neighbourhood.
"The huge CANDELS dataset was a great resource for us to use in order to consistently study ancient galaxies in the early Universe," concludes Lee. "And the resolution and sensitivity of Hubble's WFC3 is second to none in the infrared wavelengths needed to carry out this study. The Hubble Sequence underpins a lot of what we know about how galaxies form and evolve — finding it to be in place this far back is a significant discovery."
 Previous studies have looked at the proportions of the different galaxy types back in time (heic1002). The mix of spiral, elliptical, lenticular and peculiar galaxies is different from today, with a great many more peculiars in the distant Universe than we see nearby.
 In a related recent paper, Alice Mortlock and collaborators took a different but complementary approach by classifying these distant galaxies by visual inspection. They found that the types of galaxies we see in the Hubble Sequence are well defined in terms of colour, structure, and star formation rates at very large distances from us, but that their morphology is still developing. While the morphology of a galaxy may be the final property to settle, the fundamentals of the Hubble Sequence are set much earlier on.
 CANDELS, the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey, is the largest project in the history of Hubble, with 902 assigned orbits of observing time. It is being carried out with two cameras on board Hubble – WFC3 and ACS – and aims to explore galactic evolution in the early Universe, and the very first seeds of cosmic structure at less than one billion years after the Big Bang.
 Previous studies of this period of cosmic history were inconclusive as they were limited to visible light, showing only the redshifted ultraviolet emission of the galaxies, which highlights star formation. As this star formation dominated the observations, the galaxies appeared to be clumpy and messy, with no resemblance to the galaxy shapes we see around us today. By pushing into the infrared part of the spectrum the astronomers could observe how these distant galaxies appear in their visible rest frame (which is now redshifted).
Notes for editors
The Hubble Space Telescope is a project of international cooperation between ESA and NASA.
These results are described in a paper entitled "CANDELS: The correlation between galaxy morphology and star formation activity as Z~2", to appear in The Astrophysical Journal.
 The international team of astronomers in this study consists of B. Lee (University of Massachusetts, USA), M. Giavalisco (University of Massachusetts, USA), C. C. Williams (University of Massachusetts, USA), Y. Guo (University of California, USA), J. Lotz (Space Telescope Science Institute, Baltimore, USA), A. van der Wel (Max Planck Institute for Astronomy, Heidelberg, Germany), H. C. Ferguson (Space Telescope Science Institute, Baltimore, USA), S. M Faber (University of California, USA), A. Koekemoer (Space Telescope Science Institute, Baltimore, USA), N. Grogin (Space Telescope Science Institute, Baltimore, USA), D. Kocevski (University of Kentucky, USA), C. J. Conselice (University of Nottingham, UK), S. Wuyts (Max Planck Institute for Extraterrestrial Physics, Garching, Germany), A. Dekel (The Hebrew University, Israel), J. Kartaltepe (NOAO-Tuscon, Arizona, USA), E. F. Bell (University of Michigan, USA).