Top critical review
4 people found this helpful
on 31 January 2009
I got this, and other sources, as a reference for background material, so I have not read it cover to cover, but it seems as though every time I refer to it I find more to complain about.
An early indication that the authors understand very little about orbital dynamics, or even about dynamics in general, is found on page 18, where they say "we might reasonably conclude that the non-circular motions in the Milky Way are induced by a (similar) galactic bar". Indeed, the assumption that galactic orbits should be circular or near circular runs through the book. But there is no reason that orbits should be circular. A circular potential does not imply circular motion. To solve a dynamical problem, we must know initial conditions. The initial conditions for Galaxy formation in the early universe where hot and turbulent, as great gas clouds collapsed under gravity. It is highly improbable that orbits would be circular. In the solar system we see near circular orbits because they have been settling down for billions of revolutions. But the period of a galactic orbit is typically hundreds of millions of years. Typical stars have orbited their galaxies a few tens of times. There is no way a competent dynamicist would have expected to see mainly near circular orbits. Recently observations by the Keck telescope have shown that early galaxies were indeed more chaotic. This was described as a surprise to astronomers, [...]. But it would have been no surprise if astronomers had not been misled by incompetent dynamicists, touting theories such as those described in this and Binney's earlier books.
Why does a book which is supposed to be up to date have a section on the convergent point method of determining cluster distance? The convergent point method was only ever useful for the Hyades, and it was rendered obsolete by the accuracy of Hipparcos parallaxes. Although the Hipparcos results were published at about the same time as the book, this should have been obvious during preparation.
The section on the Lutz-Kelker bias gives one the impression that, here as elsewhere, the authors are trying to hide their lack of understanding by blinding the reader with equations. They have conflated two quite distinct forms of bias using an inappropriate application of Bayes' theorem. The Lutz-Kelker bias arises because, in a population distance limited to a sphere, the number of stars outside the sphere but included in the population because of parallax error will exceed the number inside the sphere but excluded for the same reason. This is because the error shell outside a sphere has greater volume than the shell inside the sphere. It follows that the mean physical distance of the population is larger than the mean parallax distance. If such a sample is used to calibrate luminosity distance, a systematic bias will be generated. This is the bias which Lutz and Kelker studied in their papers. The authors of this book seem to somehow get everything backwards, and attribute the much simpler bias arising from the fact that distance is inversely proportional to parallax to an issue with magnitude. However, as they are almost completely incoherent, it is impossible to say what they really think.
A section is devoted to raving about how the MK system of stellar typing "*will have* enormous value for studies of galactic structure" (my emphasis). The authors have apparently overlooked that science the fact that science is concerned with what has *already* been established, and have failed to realise that as a theoretical tool the MK system was rendered obsolete by the advent of stellar evolutionary theory, fifty years ago, and that it always was a hopelessly inconsistent morass defined from over 1000 quite arbitrary "standard" stars. As a measure of surface temperature, stellar type is crude and inaccurate. Even among the standard stars, large numbers of dwarfs are classed as giants and large numbers of giants are classed as dwarfs. If stellar typing is so useful, why does Binney use colour index in preference to MK type in his papers?
I found the section on stellar evolutionary theory to be badly structured and incoherent. I was left wondering to what extent the authors' ideas on subgiants predate evolutionary theory. A very incomplete sample of stellar classes are discussed, and important details, such as dredge-up phases, are missing. An adequate account at this level is given by Carroll and Ostlie, and there are a number of specialist books.
Much of the "bringing up to date" taking place in chapters 9 & 10, consists of reporting on on Binney's own papers, and those of his collaborator, Dehnen. Unfortunately, this research has rapidly been superseded, and he has since had to admit that he and Dehnen got the value of the LSR (the local standard of rest, or idealised speed of circular motion) badly wrong. He describes Eggen's hypothesis on the origin of stellar streams, but this was almost immediately refuted by the ages of stream stars, and should have been rejected from the plots described in the book, as the sizes of the streams makes it quite unrealistic. In fact Binney's figure for the LSR is unreasonable. He could have avoided these mistakes by doing some elementary checking, and instead of repeating Stromberg's asymmetric drift relation in three books, he should have realised that it is refuted both by the existence of streams and by the answer to a regular question in GCSE mathematics and AS level statistics --- it is dangerous to extrapolate a correlation beyond the point for which there is data.