In defense of Lord Kelvin

You’ve probably heard the story about how in 1900 Lord Kelvin gave a speech claiming that, with the exception of a couple of minor unresolved issues (“clouds”), physics is basically wrapped up, and what’s left is just computing things to more decimal places.  Shortly thereafter, the solution of these two issues gave rise to the relativity and quantum revolutions, proving Kelvin spectacularly wrong.  Here’s the thing, though–I’ve read the speech, and Kelvin doesn’t say anything of the sort.  In fact, it’s clear that he regarded the unresolved issues as quite fundamental.

The speech itself is an interesting snapshot in time; it allows us to see the world through the eyes of a very intelligent person in 1900, without the foresight of knowing what was to come.  The first cloud is the Michelson-Morley experiment:  why do measurements of the speed of light in different directions not reflect the Earth’s motion through the ether (as, for example, the velocity of the air adds to the propagation velocity of sound waves)?  Kelvin admits no satisfactory explanation was yet at hand, but he relates an idea of Fitzgerald and Lorentz that perhaps motion with respect to the ether somehow causes matter to contract.  In other words, the Lorentz transformations have already been derived, but no one knows what they mean, and with benefit of hindsight we can see that Kelvin’s interpretation is heading down a blind alley.  However, putting oneself in his shoes, these do seem like reasonable (and, in their own way, quite radical) ideas to pursue.  Physics textbooks speak as if the Michelson-Morley result leads straightforwardly to special relativity, but in fact even with Lorentz having come up with the equations, Einstein and Minkowski had to make incredible conceptual leaps before people could understand what these equations meant.

The second cloud is the specific heat of molecular gases (not blackbody radiation, as some sources say).  Here, the story one hears about complacent nineteenth century physicists is completely unjust.  A full half century before Kelvin’s speech, Maxwell, in his own seminal work on the kinetic theory of gases, had proved that classical atomic theory couldn’t possibly explain observed reality.  Except for very simple (diatomic) molecules, the number of degrees of freedom will be too large, because each should contribute equal average kinetic energy, leading to heat capacities much higher than observed.  Boltzmann and Rayleigh also spoke of this as a severe problem.  In retrospect, we know what classical physics was missing–it was assuming all these degrees of freedom would be continuous.  In fact, there was already a hint that this wasn’t always true–if any had been able to interpret it as such–in the known existence of discrete spectral lines.  However, even with such a dramatic falsification of classical physics, the idea of discrete energy levels was a radical (one might say “quantum”) leap, something that wouldn’t have occurred to most of us.  Kelvin himself, from what I can tell from skimming, seems to have thought that Maxwell’s error came from the ergodic hypothesis–the long-term equiprobability of all states.  In fact, ergodicity breaking is an interesting topic in physics (cf. spontaneous symmetry breaking), but for this problem Lord Kelvin is once again not heading in the most fruitful direction.  However, one cannot accuse him of timidity; he’s calling the whole mathematical basis of statistical mechanics into question.

Kelvin is also attributed with the following quote “There is nothing new to be discovered in physics now.  All that remains is more and more precise measurement.”  As Wikipedia explains, this quote is found only long after the alleged statement, always without citation, and is very likely apocryphal, a misreading of an earlier statement by Michelson.  The actual 1900 speech is itself strong evidence that Kelvin believed no such thing.