First of all let me apologize for criticizing this work, not only because I wrote part of it and don't want to hurt my own feelings (any more than is absolutely necessary), but to you for any appearance of arrogance or impropriety. However, some more evidence in its favor has come to light since this book was written; in addition, there are a few mistakes (some corrected in proof which somehow Basic Books neglected to fix) and, more importantly, a basic potential misunderstanding about what the book does and does not say, which I see no reason not to address.
The main point of this book, which I cowrote, is that, although mutation leads to evolutionary change, all the best examples of speciation, including all that have actually been observed, have been through symbiosis. The greatest amount of biodiversity, including all basic metabolic modes from photosynthesis to oxygen respiration, evolved in the bacteria via mutation and gene transfer. But although given Linnean species names for the sake of convenience and via convention, speciation does not really apply to bacteria, which trade genes (via techniques borrowed by human beings practicing biotechnology) with little regard for species barriers. True speciation only evolved in the eukaryotes--protists, fungi, plants, and animals. These largely sexed beings pose the Darwinian problem of speciation proper. And here all the best examples of speciation involve symbiosis, the coming together of different kinds of organisms. Since Acquiring Genomes was written, more evidence has come to the fore to show that its central thesis--that the presence or absence of genomes, particularly those of microbes, can lead to speciation--is correct. In a recent Montreal conference on molecular biology and phylogeny, for example, John Werren from the University of Rochester in New York showed a picture of a chromosome of a sperm cell from a parasitic wasp: rod-shaped bacteria, Wolbachia, were nestled in the chromosome. Wasps can have their sex change due to the presence of bacteria, and antibiotics can make separate species of jewel wasps interbreed again. At this same meeting Professor Harold Morowitz (who is developing a Universal Metabolic Chart, on the model of the Periodic Table of the Elements) was impressed by the plasticity of ever-changing gene formations--emphasizing the need to look for metabolic pathways shared by most or all organisms to understand life's origins. Because life is an open thermodynamic system, as well as an open informational one, genomic transfer is rampant.
It is important to realize two things that Acquiring Genomes does not say. The book does not say that all bacterial diversity is the result of genome acquisition. As suggested above, and by Canadian biologist Sorin Sonea and others, despite the bacteriological convenience of their species names, bacteria arguably do not have species due both to rampant genetic transfer as well as the premier, zoological definition of species as an interbreeding population; since all bacteria can theoretically trade genes with each other either directly or through through vectors (and do not need to reproduce to do so), the animal definition of species does not really apply to them. The original genetic and metabolic diversity in bacteria must owe significantly to neodarwinian-style mutations but, since bacteria arguably do not possess species, such mutations do not occur for speciation.
As Ernst Mayr suggests in his Foreword, the evidence for speciation by genome acquisition in birds and mammals is not compelling. The argument for genome acquisition here depends on the possible symbiotic status of the ends of chromosomes, called kinetochores. (Bacteria don't have true chromosomes, they have chromonemes.) Because chromosome arrangements differ slightly in closely related mammal species (e.g., dogs and wolves) that no longer breed with each other, and because the spontaneous splitting of these chromosomes may owe to their separate bacterial origin, we make the argument that even vertebrate speciation may owe to the symbiotic aftershocks of microbial genome acquisition. The main point to remember is for every example of speciation for which there is actual evidence, genome acquisition is the causative factor; and that, despite mountains of theory, this is not the case for mutations.
Finally, the thermodynamics section is only an at best tantalizing foretaste of a much more comprehensive argument and regrettably contains a couple of mistakes, such as the characterization of Benard cells as octagonal (they're hexagonal) and appearing from a chemical gradient (they don't; they appear in a temperature gradient). And one final comment: both Lynn and I read Stephen King's On Writing after A.G.'s composition and realized belatedly how much it could have been improved, despite the complexity of some of the arguments, by eliminating further needless words.