If it is difficult to write a good and intelligible textbook on thermodynamics, it is next to impossible to write one that is not only good and intelligible, but also entertaining. Donald Haynie's book doesn't quite achieve that, but he makes a valiant effort, with numerous boxes to provide biological illustrations of the principles he is discussing. The first of these is called "hot viviparous lizard sex", and this gives a good indication of what is to follow in the others. The chapters become longer and more biological as the book proceeds, but the boxes become less frequent. This is a pity, because it is in the later chapters, when one sees the application of thermodynamics to biochemistry, that students will appreciate some light relief, and I doubt whether a footnote about Yesterday and Paul McCartney's knighthood will be enough to convince students of the importance of the Michaelis-Menten equation. At least one opportunity has been missed: the relevance of osmosis to everyday life is discussed in relation to the lysing of red blood cells if blood is diluted with water, but there is no mention of cooking. Proper treatment of blood in hospitals is, of course, very important, but it is hardly a matter of everyday experience for people who don't work in hospitals; on the other hand, it can be a valuable learning exercise to analyse the effect of adding a little salt when boiling carrots, even though virtually none of the salt (far too little to taste) remains when the water is discarded.



At times, one sees evidence of a failure to think things through. After telling us that Japanese honeybees can kill a hornet by forming a compact ball around it and raising its temperature by more than 10°C, the books asks "what if humans could somehow turn up their temperature at will?", forgetting that we were told (correctly) earlier on that whereas bumblebees can generate heat metabolically, honeybees cannot, so they use vigorous exercise and crowding, the same methods that work perfectly well (albeit not by 10°C) for humans.



There is little to criticize in the first half of the book, as the author clearly has a thorough understanding of chemical thermodynamics and a gift for presenting it in a way that is far more accessible than one finds in almost any other textbook on the subject. The main thing that I didn't like is that he wasn't able to decide whether to express energies in kJ, kcal or cal, switching arbitrarily between the three through the whole book. I haven't seen the first edition, but my guess is that the inconsistency is the result of a rather late and incompletely implemented decision to change from one system to another. So far as the problems are concerned (many thought-provoking problems at the end of every chapter), this may be a good thing, as readers of the biochemical literature need to be capable of understanding information presented in different symbols, terms and units from those they prefer. For the text, however, I can see no justification for it.



Although the more biological part of the book is in general correct and well written, there are some very unfortunate faults. The van 't Hoff plot is illustrated with a graph in which the zero on the abscissa is not labelled as such (the ordinate zero also, but that has no importance as it has no fundamental meaning). This is serious, because in the overwhelming majority of published van 't Hoff (and Arrhenius) plots, the zero is very far to the left of what is drawn, and one must not assume that the point where the axes cross is the origin. In the example given, the lowest abscissa value is, however, intended to represent zero, because the ordinate intercept is labelled as DeltaS0/R. This means that the "experimental" points span a more than 15-fold range of absolute temperature, so if the highest temperature used was 40°C, the lowest was about -250°C: this doesn't seem very plausible for an experiment of biological relevance, but if we assume that the lowest temperature was 0°C the highest was more than 4000°C, which seems even less credible. Why does it matter? It matters because estimation of entropy from a van 't Hoff or Arrhenius plot involves extrapolation of typically more than ten times the range of observations and thus results in a estimate with such a large statistical uncertainty that it is completely meaningless. Failure to understand this has resulted in a whole industry of nonsensical papers about entropy-enthalpy compensation. This is not to say that entropy-enthalpy compensation can never be a meaningful property, only that the entropy needs to be measured in a meaningful way. The idea of compensation appears later in the book, but without a clear indication that the thermodynamic parameters must come from calorimetric measurements, not from van 't Hoff plots.



Later in the same chapter, we read that "glucose phosphorylation is coupled to ATP hydrolysis", but that is again nonsense, because no hydrolysis is involved in glucose phosphorylation. It is true that the difference between standard Gibbs energies of ATP hydrolysis and glucose 6-phosphate hydrolysis tells us about the standard Gibbs energy of glucose phosphorylation, but that is as far as it goes; we can say nothing useful about the ratio of Gibbs energies. Haynie doesn't quite commit this sin, but he skates on thin ice when he talks about "efficiency" without being totally clear about what he means. Later on, we realize that the revolution in our understanding of metabolic control has passed him by, as he states that phosphofructokinase controls glycolysis, without any hints about why that might be a misleading statement. In the same context, he presents the DeltaG for the combination of aldolase and triose phosphate isomerase reactions in terms of ln([GAP]^2/[FBP]) without any mention of the implications of writing the logarithm of a concentration: what does it mean, and how can we make it acceptable? This would have been a good opportunity to remind the reader of the idea of a standard state, defined in the preceding chapter, but without any convincing explanation of why it is needed.



There is another missed opportunity in the next chapter, when the tired old cliché about being able to fit an elephant if you include enough parameters in your model is used to introduce the (perfectly valid) point that one should be cautious about adding parameters in a thoughtless way, followed by a comment that "entire books have been written on data analysis". True, but surely space could have been spared (in a chapter entitled "statistical thermodynamics") to mention that classical statistics developed to handle exactly this problem, and that that's what statistical tests are all about. The book returns to data analysis, but in an equally superficial and unhelpful way, in the later discussions of binding and kinetic data.



In discussing the Hill coefficient (symbolized as n), the book makes the usual error of relating the Hill equation to a model with n molecules of ligand binding simultaneously (despite the fact that Hill made it perfectly clear, a century ago, that that was not the right model), and then says that "n can in principle take any real value", providing only a half-hearted explanation of what non-integral values might mean. The confusion is compounded by equating negative cooperativity (n less than 1) with negative Hill coefficients (n less than 0). Negative Hill coefficients are possible, of course, in kinetic experiments, but not in binding experiments, the context in which the statement occurs. Another grotesque error occurs a few pages later, when the asymptotes of a curved Scatchard plot are drawn as if the straight parts of the curve lie along them; this error is responsible in the literature for some huge errors in estimates of binding constants.



In summary, there is much to like in this book, and as an introduction to chemical thermodynamics, it is excellent and readable, but for the more biological aspects it is more problematic, and I'd be reluctant to recommend it to students without some caveats. Nonetheless, a biochemistry student who read it and studied it thoroughly would certainly finish with a far better understanding of thermodynamics than is usual.