My thesis advisor has often lamented the rise of email in society, mainly because the electronic correspondence isn’t as permanent as the written letter, and lots of interesting historical anecdotes can get lost.

I delved into one of these recently while reading H.P. Lovecraft’s letters. Lovecraft was, in many ways, a ‘dream citizen’ for scientists. He had an active interest in all branches of science and attended popular lectures on many subjects. He also vigorously defended science against the encroachment of pseudoscience on one occasion (more on this in another post).

The particular letter which I want to describe is dated May 9, 1936 to James F. Morton:

Recent lectures of interest have been on Plato’s Republick, modern art, Gilbert Stuart, Rhode-Island silversmiths, archaick Greek art, Philosophy and Poetry, early classical sculpture, Mayan ruins, and the Michelson-Morley experiment. The last-named, deliver’d at the college Monday night, was by Prof. Dayton C. Miller, former colleague of Morley and present continuer of the experiment. He furnish’d startlingly convincing proof that the real results of the experiment do NOT shew that total absence of effect of the observer’s motion on the speed of light which forms the underlying assumption of the Einstein theory. Instead, there is merely a lack of the full difference which the observer’s motion ought (according to the old theory of time and space) to make… Miller himself offers no dogmatic solution, but suggests that a drift in the luminiferous aether (assuming, contrary to Einstein, that such exists) in the direction of the earth’s motion would account — on the basis of the old pre-Einstein universe of non-relativity — for the fact that the observer’s change of place in space gives some of the effect demanded by the old concept, but not all of the required amount. If Miller is right, the whole fabrick of relativity collapses, and we have once more the absolute dimensions and real time which we had before 1905.

Here Lovecraft offers a clear, concise contemporary account of one of the early major experimental challenges to Einstein’s relativity theory.

For those not familiar with relativity and its history, let me give a quick background of the pertinent details. A full discussion of the history and theory will come in later posts.

Physicists had understood, since the time of Newton, that motion is in some sense a relative concept. When I drive my car down the expressway, an observer by the side of the road states that I am moving forward at 65 mph. However, from my perspective, the observer is moving 65 mph backwards relative to me. The observer and I disagree about who is actually doing the moving: in a nutshell, that’s relativity — only relative motion matters. Intuitively, we would all probably argue that the observer actually has the ‘correct’ point of view. He’s standing still relative to the Earth and the atmosphere, while the car is burning gas, feeling wind resistance, and vibrating due to bumps in the road. Newton, however, correctly surmised that all of those factors are due to the relative motion with respect to the Earth: the fundamental laws of physics work the same if we imagine instead that the car is standing still and the Earth is moving underneath it. Another way to say it is that the laws of physics are the same for the person in the car and the observer by the side of the road: no scientific test can detect ‘absolute motion’, which evidently does not exist.

This notion of relativity held until the late 19th century. When Maxwell’s equations for electromagnetic fields, and light waves, were discovered, it was observed that these equations did depend on the motion of the observer. Because all waves that had been studied up to that point (sound, water) traveled in some sort of medium (air, water respectively), scientists assumed that light traveled in its own medium, the aether. All motion could be measured by motion with respect to the aether, which would apparently be ‘absolute motion’.

Scientists then worked to detect motion with respect to the aether and, consequently, detect the aether itself. Since the Earth moves in a circular orbit around the Sun, scientists reasoned that it must be moving relative to the aether at some point in its orbit. This motion could be detected by looking at the speed of light as the Earth moves.

In 1887, Albert Michelson and Edward Morley performed just such an experiment (now known as the Michelson-Morley experiment). Their instrument was precise enough to detect any change in the speed of light due to the Earth’s motion, but they found no such change. This observation dealt a serious blow to the idea of an aether — if there was an aether, it didn’t behave the way it should!

A number of solutions to the problem were proposed, but it was Einstein in 1905 who came up with the conclusive answer. He argued for a change to Newton’s principle of relativity: relativity should be reformulated so that the laws of electromagnetism are independent of any absolute motion. Doing this requires that the speed of light be the same for any observer, no matter how they are moving. One of the most shocking consequences of this conclusion is that time is no longer an absolute quantity: different observers will disagree on when events happen. Einstein’s result was so elegant, and explained so perfectly the Michelson-Morely experiment, that it was quite readily accepted.

We’ll talk more about the specifics of Einstein’s relativity in later posts; for now, we return to the subject of Lovecraft’s letter, Dayton Miller.

Miller began working with Morley in 1900, and they began to refine the original M-M apparatus to make even more delicate measurements of the speed of light and (presumably) more sensitive searches for the aether. In 1904, they published another negative result. Miller continued to work on the project, and in 1926 and 1933 reported detecting a small amount of what appeared to be aether drift. These results received quite a lot of attention, and even troubled Einstein somewhat, for if true, they cast doubt on the entire relativity framework. The aether drift Miller detected, however, was much smaller than what would be expected due to the motion of the Earth.

Clearly, Lovecraft attended a lecture of Miller’s relating to the 1933 work, and found the results impressive. Other scientists performing similar experiments, however, did not detect the same anomalous aether drift that Miller observed. Most considered Miller’s work to be a statistical anomaly or experimental error, and relativity studies continued apace. In 1955, Robert Shankland reevaluated the Miller experiments and concluded that systematic errors had resulted in incorrect results.

As an aside, it is worth mentioning that this conclusion does not cast strong aspersions on Miller’s scientific credentials. Systematic errors, which are in essence mistakes caused by subtle problems with experimental methodology, plague all experimenters and can be very hard to track down. The fact that it took until 1955 for someone to clearly explain the likely problem with Miller’s work is evidence that this wasn’t an obvious blunder.

Unfortunately, Miller now has a legacy as a hero to modern-day relativity denialists, who are still going strong. Despite hundreds, if not thousands, of experiments which demonstrate the validity of Einstein’s results, there will always be a few people who can’t let go of their simplistic notions of ‘absolute motion’.