Page upon page of paradigm shifts

Our knowledge of stars, galaxies and planets has changed drastically in the past 100 years. This is hardly surprising. In this period, we have broken away from the blinkering bonds of the visible portion of the electromagnetic spectrum and have expanded into the infrared, radio, ultraviolet and X-ray regions. Our instruments have been elevated from the surface of Earth and transported into the clarity of space. The diameter of the largest telescope mirror has increased by a factor of ten, and we have taken full advantage of the many improvements in radiation detectors and computers.

In 1900, we thought that there was only one type of star and that all stars had a similar composition to planet Earth. We had no idea how stars generated their energy, no idea what they were like inside, and very dubious views as to their origin and evolution. We did not know why the average star mass is about 13 per cent of our sun's, or why there are no stars more massive than about 60 solar masses or less massive than about one 30th of the solar mass.

In the past 100 years, we have discovered that stars consist mainly of a 3:1 mixture of hydrogen and helium, that the transmutation of hydrogen into helium and E=mc 2 has solved the energy problem, and that the astronomical zoo contains not just one astronomical animal but a range of white, brown and ordinary dwarf stars, giant and supergiant stars, pulsars and black holes, together with a whole host of variable objects.

In 1900, we thought that there was only one galaxy and that our sun was near its centre. Now we know that the universe contains about 10 11 separate galaxies of different shapes, and we understand how the form of these galaxies is affected by mass, spin rate and age. Even more surprisingly, the static universe has been replaced by one in which the galactic system is expanding.

Cosmology, the study of the form, origin, evolution and end point of the universe, hardly existed in 1900. Not only have we discovered that space/the universe is forever getting bigger, we have also realised that about 15 x 10 9 years ago all the material was confined in an exceedingly small and hot volume. This went "bang". The big bang was of just the right energy and turbulence to ensure that the material in the universe eventually clumped together into stars and galaxies and did not just expand into an ever-rarefying gas cloud. The parameters of the big bang and the resulting universe also seem to have been finely tuned to ensure that some 15 x 10 9 years after the beginning, human life has developed sufficient intelligence to start to try to understand it. By studying stellar spectra, we have discovered that the elemental composition of the universe is dominated by hydrogen and helium. We have also found that all the material in the universe is not in the form of stars, planets or satellites. There are huge amounts of interstellar gas and dust. And what is more, there is "missing mass", which can be detected gravitationally by its influence on the orbits of stars and galaxies, but has a physical and chemical make-up that is unknown.

Studies of stellar evolution and nuclear astrophysics have revealed where the chemical elements come from, how one element is converted into another and why they have the abundances that they do. We have also discovered that our planetary system is not unique. But we have little idea where our planets came from, let alone the mechanisms that formed planets around other stars. Although 12 men have visited Earth's moon and brought back about 380kg of lunar material, the Moon's origin is still unclear.

Other planetary surfaces have been studied. Space probes have flown past many planets and satellites, orbited some and landed on a few. Instead of just knowing a bit about the surface of Earth, as we did in 1900, we can now compare our planetary features with high-resolution images of the surface of, say, Mercury, Venus, Mars, Eros, Io, Titan, Triton and Pluto. The space age has revealed a panoply of craters, canyons, solidified lava lakes, desiccated deserts, dry river beds, acidic clouds, sulphurous volcanoes, methane seas and fractured ice floes. We are on the way to understanding the role of continental drift, planetary differentiation, atmosphere formation, volcanism, and asteroidal and cometary catastrophic impacts.

The historical record is always selective, and the history of astronomy is no exception. To commemorate the first century of the American Astronomical Society, its centennial committee has done something that I have never seen done before. They asked a group of their members to trawl through the past 100 years of The Astronomical Journal and The Astrophysical Journal , the two world-famous publications produced by the society, and select the "fundamental" papers.

Fifty-three made the grade. Their authors come mainly from the US, the exceptions being astronomers such as Fred Hoyle, Vainu Babu, Yoshihide Kozai, Donald Lynden-Bell, Bengt Stromgren and David Bates, who happened to be taking their sabbaticals in the US at the time the research was done. The temporal distribution of the chosen work peaks very strongly in the 1950s (19 papers) and 1960s (13). The 1980s and 1990s score only one each. Whether this truly reflects the production rate of "fundamental" historical astronomical papers or is merely an artefact of the fact that the nominated group favoured the papers that impressed them in their youth, only time will tell.

The final collection abounds with the impressive and the ground-breaking. Paradigm shifts are ten a penny. Typical examples are the 1919 investigation of the positions of globular clusters by Harlow Shapley, a paper that moved the sun well away from the centre of the galaxy; the 1925 "proof" by Edwin Hubble that spiral nebulae were extra-galactic; and Hubble's 1929 analysis that showed how galactic recessional velocity varied with distance. There is George Hale's 1908 discovery that sunspots contained magnetic fields; the 1944 discovery of radio waves emanating from the direction of the galactic centre by Grote Reber; the 1964 discovery of extragalactic quasars by Jesse Greenstein and Maarten Schmidt; the 1965 discovery of the 3.5K universal background radiation by Arno Penzias and Robert Wilson; and the 1970 discovery of solar oscillations by Roger Ulrich. The list also includes Fred Whipple's 1951 model of the dirty-snowball cometary nucleus; Hoyle's 1954 work on the nuclear synthesis of the elements between carbon and nickel; and the 1974 paper by J. P. Ostriker, P. J. E. Peebles and A. Yahil on the mass of the universe.

Each of the 53 keynote papers is reproduced in its entirety and in facsimile. It is also accompanied by a short article, written by a US expert in the subject. This explains why the paper was selected, places it in its original context, describes how it led to change in the relevant field and brings the implications of the paper up to date.

The result is an absolute joy that provides a revealing historical window into the achievements of a subset of 20th-century US astronomy. Reading the original papers, one establishes a common bond with one's astronomical predecessors. The theoretical and observational problems that are tackled today might be different, but the methodology and aspirations are very similar.

It would be somewhat unfair to criticise the book for its obvious bias. It goes without saying that astronomy did not develop only in the US. Also, a huge amount of excellent 20th-century US astronomy was published outside the confines of The Astronomical Journal and The Astrophysical Journal . The book is also biased in a more typically "historical" fashion. The concentration on success stories and the inclusion of only the pinnacles of endeavour gives no impression whatsoever of the pitfalls, blind alleys and stumbling blocks of typical research life. The good is not balanced by the bad. It would be most revealing to return to the bulging library shelves and select the 53 "worst" papers from these two journals. Collecting these into a single volume, together with brief discussions as to what went wrong and why, might help us plough a straighter furrow in the future. It would certainly give a refreshingly different insight into the way in which a scientific subject progresses and into what the less eminent members of the astronomical profession get up to.

David W. Hughes is professor of astronomy, University of Sheffield.