While I was doing an unrelated search on the Nature website, I came across this intriguing debate about why body temperature is maintained around 36 degrees and not some other value. (Nature, Vol 324, December 4, 1986, p. 418)
The discussion was initiated by a letter from John Paul, a scientist in Australia who contended that the specific heat capacity of water is lowest at 36 degrees, and therefore heat loss would be minimal at that temperature
But he neglected a fundamental principle of physical chemistry; the rate of heat loss is proportional to the difference between the temperature of the body and that of the surroundings and is independent of the specific heat capacity (remember high school and Newton's Law of Cooling?). More importantly, the specific heat capacity of a body can be thought of as a measure of how well the body offers "resistance" to fluctuations in temperature. The reason why water works so well as an essential life fluid for example is because its specific heat is so high; there is minimal fluctuation in the temperature of water when heat is injected or taken away from it.
Thus, an optimal substance for maintaining a given temperature would be one whose specific heat capacity is as high as possible under the given circumstances, not one whose specific heat capacity is minimum at the given temperature.
These facts were pointed out by William Calder from the University of Arizona and by Steven Benner and Jack Dunitz at the ETH, Zurich. Dunitz as is known is an extremely versatile scientist, a veteran researcher and one of the greatest structural chemists and technical writers of the last century.
Dunitz and Benner make their objections to Paul's explanation clear and offer an alternative partial explanation; that 36 degrees is the optimum compromise between viscosity and hydrophobicity. It's high enough for the viscosity to not become so low as to impede diffusion-limited processes, and low enough that hydrophobic molecules do not "dissolve" too easily.
Natural selection must have taken a remarkable number of factors into account in optimizing this property.
Modular drug design software?
2 hours ago in The Curious Wavefunction