By RICHARD JENSEN.
ALBERT MICHELSON AND EDWARD MORLEY were in the basement of a dorm at Western Reserve University in Cleveland with a curious contraption assembled on a stone floating in a donut-shaped trough of mercury.
It was April 1887. The two young men were trying to measure the speed of light. Or, more precisely, they were looking for variations in the speed of light. The actual speed of light had been determined with increasing accuracy beginning in 1676, and by the time of Michelson’s experiments, it was known within six tenths of a percent of the value we use today.
At that time there was no reason to believe that the speed of light was absolute, or that it represented some upper limit on how fast objects could travel. In fact, the prevailing theory about light itself implied that the speed of light would vary to an earthbound observer by a minute but detectable amount caused by the earth’s orbit around the sun.
In 1887, it was believed that light was propagated by waves and that these waves needed to propagate in a medium. By analogy, scientists believed that light behaved similarly to sound, which cannot travel through a vacuum (Alien was right: In space no one can hear you scream), or like ripples on the surface of a pond. Try to imagine ripples without the pond. It doesn’t make sense. Waves are disturbances in some underlying medium.
The name given to the proposed medium which carried light through apparently empty space was the luminiferous aether. Michelson and Morley were looking for evidence of the “aether wind,” and thus evidence of the aether itself. Because the earth revolves around the sun, scientists reasoned that the motion of the earth relative to this aether would cause a “wind” that would affect the speed of light.
IF YOU ARE standing still on a still day, you do not feel any wind. But if you start running in any direction, the air’s resistance to your movement will seem like wind. On the other hand, if there is already a wind present, if you run with the wind at your back, the same amount of effort will enable you to run slightly faster than if you are running into the wind, or if the wind is coming from the side.
The Michelson-Morley experiment was designed to look for the very subtle boosting or slowing effect that the earth’s orbit would have on the speed of light. Earth’s orbital speed was known to be about .01% of the speed of light, and that’s the size of the discrepancy they were looking for.
That they were able to design a test that could reliably obtain those results without relying on sophisticated electronics and computer assisted data analysis is a testament to just how smart these men were. That this experiment was small enough to assemble in the basement of a dormitory is, in hindsight, perhaps even more remarkable.
The test featured a partially silvered mirror that split a single beam of light into two beams that traveled down perpendicular pathways to mirrors which reflected them back and forth and then finally to an observation point.
The ingenious part was that the two pathways were of different lengths so that the light beam took longer to travel down one pathway than the other. This produced an interference pattern when the two beams were recombined at the observation point.
By placing one pathway parallel to the direction of the earth’s orbit and the other pathway perpendicular to it, any change in the speed of light would change the interference pattern by a predictable amount.
From April all the way into June Michelson and Morley performed this experiment over and over looking for any evidence of a change in the speed of light as the earth moved in its orbit around the sun.1
They found none.2
Instead of concluding that this test raised serious questions about the existence of the luminiferous aether, Michelson and Morley reasoned that it supported a less popular interpretation of the aether. Under this interpretation, some of the aether would stick to the earth as it passed through the interstellar medium, and because this aether was being dragged by the earth, it would move at the same rate as the earth. Thus, there would be no discernible variation in the speed of light due to the earth’s orbit. This interpretation of the aether would explain the results that Michelson and Morley recorded.
To us, almost 130 years after the fact, this may seem to be an egregious misinterpretation of data, one that calls into question their competence as scientists. However, these men were brilliant. The basic design of the Michelson-Morley experiment was used to discover evidence of gravity waves and refute the “holographic universe” hypothesis. What’s more, they were not isolated geniuses working in an idiosyncratic fashion: They were following the scientific method and publishing in peer-reviewed journals. Their instruments and language may strike modern scientists as quaint, but their method of stating a problem, offering a hypothesis, making a prediction and then testing that prediction will be familiar to anyone working in the sciences today.
IN 1859, 28 YEARS before Michelson and Morley published the results of their tests, Urbain le Verrier published his survey on the precession of Mercury. Le Verrier had analyzed the records of solar transits of Mercury beginning in 1697, and had discovered an anomaly in the precession of Mercury’s orbit.
Almost all recorded solar transits have occurred when an object in the solar system passes between the earth3 and the sun. The most obvious solar transit seen on earth is a solar eclipse. On other occasions, the planets Venus and Mercury can be observed crossing the sun. These transits provide useful information about the distance between the earth and the sun, the size of the object transiting the sun and the distance between that object and the sun.
The precession of the orbit of Mercury is the gradual but definite rotation of the planet’s orbit. Mercury’s orbit, like that of every other planet in the solar system is not perfectly round. It is elliptical, and the sun is not located at the center of the ellipse, it is located off to one side, at one of the focal points of the ellipse.
Because the orbit is not circular there is a point at which Mercury is closest to the sun, called perihelion, and a point at which it is farthest from the sun, called aphelion. This point of aphelion gradually rotates in counterclockwise fashion around the sun in a process called apsidal or orbital precession. In relatable terms, the point at which Mercury is most distant from the sun happens a little later with every orbit.
Based on previously recorded observations, Le Verrier estimated the orbital precession of Mercury at 569 arc seconds per century (about a sixth of a degree). Working without computers—or even adding machines—Urbain Le Verrier calculated the total impact that the other planets would have on Mercury’s orbit, based on their gravitational pull, and came up with a total of 531 arc seconds per century. There was a slight but significant discrepancy that could not be accounted for by any known mechanism or planetary body.
Le Verrier suggested a few possible solutions, the most popular of which proved to be the missing planet Vulcan.
IN DECEMBER 1859, just a short time after Le Verrier had proposed Vulcan as an explanation for the discrepancy in Mercury’s orbit, a physician and amateur astronomer in a village southwest of Paris, Edmond Lescarbault, contacted him and told him that he had observed an unexpected transit of the sun in March of that year.
Le Verrier rushed to Orgères-en-Beauce unannounced, carefully reviewed Lescarbault’s records and decided that he had indeed seen the planet Vulcan. Le Verrier returned to Paris, calculated the orbit and distance of this new planet, and announced it to a meeting of the French Academy of Sciences on January 2, 1860. For his part, Lescarbault was made a knight in the French Legion of Honor.
The speed with which Le Verrier acclaimed the discovery of Vulcan was almost certainly influenced by the controversy that had erupted 13 years earlier, when leading figures in the British scientific community had sought to undermine Le Verrier’s discovery of Neptune.
Based on Lescarbault’s observations, Le Verrier predicted several future transits involving Vulcan, but none of them occurred. Undaunted, Urbain tweaked his numbers with each failure. He remained convinced that some as yet undiscovered massive body was interfering with the orbit of Mercury. The evidence of Mercury’s orbital precession was indisputable. As a percentage of the overall precession of Mercury, the discrepancy Le Verrier arrived at is within 1% of the figure that is known today.
Despite the failure of Le Verrier’s predictions, the search for Vulcan continued throughout the nineteenth century, with astronomers often recording apparent observations that were not substantiated. Variants of the standard model of Vulcan as a single planet were also proposed. At one point, Lewis Swift, an American astronomer who had discovered thirteen comets, believed that there were two planets in the Vulcan system, not one.
As with Michelson and Morley’s mistaken interpretation of their experimental results, it may seem strange, with the benefit of hindsight, to see so many skilled astronomers chasing a red herring. We might think that the repeated failure to discover a planet within the orbit of Mercury should have raised questions about the limits of celestial mechanics—the equations that predicted the motion of the planets.
Doing so, however, would ignore just how successful those equations had been up to that point. When discrepancies were noted in the orbit of Uranus, Le Verrier and John Couch Adams had been able to use celestial mechanics to predict the location of the planet Neptune. The equations that Le Verrier used to predict the existence of Vulcan accounted for the motions of the rest of the planets precisely. There was no reason, apart from that small discrepancy, to doubt that these equations furnished a full and complete picture of gravity within the solar system. It made far more sense to continue to search for a non-existent planet than to reexamine the underlying tenets of a mathematical model that was nearly bulletproof.
AT ABOUT THE same time that Le Verrier was conducting his survey of Mercury and its transits and calculating the orbit of Neptune, William Thomson (eventually Lord Kelvin), James Prescott Joule and others were developing thermodynamics, the branch of physics concerned with heat and energy in large scale systems.
With the discovery that heat and energy are fundamentally equivalent and that they are conserved, the source of the sun’s energy became a serious question. Prior to the discovery of the conservation of energy, it could safely be assumed that the sun simply produced energy out of thin air, so to speak. There was no need to worry about the source of the sun’s energy, since it was not known that energy was conserved and that, therefore, the sun could not produce heat forever.
A variety of theories were proposed to furnish fuel to the sun, an early favorite being the belief that the gravitational energy produced by millions of meteorites falling into the sun would provide an ongoing source of heat. This theory was ultimately rejected because these meteorites would add to the sun’s mass, and that the effect on the earth’s orbit would have been noticeable even over the few thousand years during which man had been looking at the stars and keeping records.
In 1854, Hermann von Helmholtz proposed that the source of the sun’s energy was gravitational collapse, that is, that the sun was originally much larger than it is today, and that it is in the process of shrinking or collapsing, and that as it collapses in on itself it releases energy. The initial figures by Helmholtz, who supervised Michelson’s doctoral work, suggested an age of only about ten million years for the sun, while an adjusted calculation by Thomson suggested a figure closer to 20 million years.
In both instances, the number—although immense by human standards—was far too short for known geological processes. The figures also eventually put Thomson at odds with Darwin and naturalists who, in conjunction with geologists, were working with evidence that suggested that the earth was hundreds of millions of years old, at least.
Thomson also argued that the earth was about the same age as the sun, based on the assumption that the earth was not generating its own heat and was merely cooling down from its original molten state.
Given the dogmatic way with which Thomson declared his views, and the prominence which he had achieved by the end of the nineteenth century, he has frequently been held up as an example of a scientist who couldn’t adapt to changing times.
And yet, in so far as what was known at the time, he wasn’t wrong. From a certain perspective, scientists arguing that the earth had to be hundreds of millions of years old had no better explanation for the sun than Thomson had for people who furnished arguments in favor of the age of the earth. The disciplines of geology and biology were at loggerheads with physics. In this day and age, it’s easy to point fingers at Thomson and insist that he should have had the sense to yield, but in that time period, he was working with well-established theories of gravitation and heat energy that had held up to scrutiny and which were in use in many different disciplines. Geologists and biologists, on the other hand, were basing their arguments on speculative dating methods that could not, at that time, be verified experimentally.
The theories of geologists and biologists working in the late nineteenth century have been largely discredited. We still use the names that they gave to fossils and time periods, but that’s about it. By contrast, the equations Helmholtz and Thomson developed as a source for the sun’s heat, are still used to model early stages in the formation of stars, and the Kelvin-Helmholtz mechanism (as it is now known) serves as a primary explanation for the fact that large planets like Jupiter radiate heat.
IN 1900, LORD KELVIN gave a speech to the Royal Society in which he identified “two clouds on the horizon.” This was a somewhat discordant note in the realm of physics. The general attitude, as expressed by Albert Michelson in 1894, was this: “It seems probable that most of the grand underlying principles have been firmly established.”
Why were scientists of the 1800s so thoroughly convinced that they understood the “underlying principles” of the physical world?
First of all, the math worked. And, for the most part, it kept on working. The math was also very elegant. Earlier in the century, Karl Friedrich Gauss had published a series of equations that established a fundamental similarity between the way gravity, electromagnetism and (subsequently) light acted on objects.
Not only that, the theories expressed in these equations resulted in practical developments. Beginning with Heinrich Hertz’s experiments, almost the entire field of telecommunications can be traced back to James Clerk Maxwell’s field equations. Electrical power and the second industrial revolution were also made possible by the work of Maxwell, Michael Faraday and others.
There was an unquestionable trace of hubris in Michelson’s belief that his peers had found the ‘grand underlying principles,’ but it was a sentiment that should not seem unusual to anyone familiar with the way physics is presented today.
Great practical achievements have been made possible by the physics of the twentieth century. Certainly, we know much more about the universe around us than Lord Kelvin and his peers.
However, we do well to remember that Kelvin and his peers had a similarly advanced understanding as compared to the ‘natural philosophers’ of 1800.
INSOFAR AS WE KNOW, the universe follows laws that can be expressed mathematically. We also know that we have not yet come up with a complete mathematical model of the universe. We shouldn’t be surprised by this; mathematical models are just that: models. They are not the reality; they are an approximation of reality as best as we understand it based on the information we have at our disposal.
There are reasons why we may well be clinging to models that are in need of replacement. They are not too different from the reasons why 19th century scientists were so reluctant to accept the twentieth-century revolutions in physics.
The foremost reason we continue to use models that show increasing signs of inaccuracy is the track record these models have. The field of electronics exists because of the twentieth century revolution in the understanding of the atom. The so-called “Standard model” has been acclaimed as the most successful predictor of new discoveries in particle physics. Einstein’s theories have opened a door to the entire universe, making it possible for us to begin to understand the largest, strangest and oldest things that we can see.
There’s also considerable inertia behind prevailing models. With very good reason, these models are what students are taught, they are what theses and dissertations are written on, and they are what people in the field know. It would be irresponsible for universities to stop teaching the standard model and relativity on the basis that both are quite clearly incomplete representations of the world around us. Until something better comes along, this is what students should learn, but at the same time, this undeniably maintains an inaccurate status quo.
And aside from those pragmatic reasons for continuing to teach inaccurate models, there’s a human element. Human understanding may have increased by leaps and bounds over the past century, but human nature hasn’t. The pursuit of answers to basic questions is undertaken by individuals every bit as imperfect and every bit as prone to confirmation bias and partiality as their nineteenth century forebears.
The math is also increasingly difficult to master and manipulate. Any new model of the universe—or some as yet inadequately explained portion of it—needs to provide answers that are at least as good as those provided by the current models. This does not suggest that the math is going to get easier.
Possibly the time is not ripe for conceptualizing the next generation of physics. Although Albert Einstein is credited with developing the theory of relativity and the fixed speed of light, his paper used equations published by Hendrik Lorentz years earlier. In fact, certain aspects of Einstein’s theory of relativity were advanced by George FitzGerald over 25 years before Einstein’s paper—and they were advanced as an alternative explanation for the unexpected results of the Michelson-Morley experiment. Relativity would almost certainly have been discovered by someone else, if not Einstein, but not until the mathematical framework had been put into place.
Is the mathematical framework in place for new physics? Time will tell. Will this new physics bring along with it new concepts? To the layman, the most intriguing aspects of relativity and quantum mechanics are the counterintuitive pictures they paint of the world around us. It seems almost impossible that a new physics will not bring along with it a new interpretation of the structure of the universe. Frankly, it would be disappointing if this were not the case.
While nothing can be known for certain, it seems hard to believe that—despite all their successes—relativity and quantum mechanics are anything approaching the last word in describing the world around us, and how fortunate that this should be the case. Better to live in a world where Shakespeare’s view that “there are more things in heaven and earth, Horatio, than are dreamt of in your philosophy” holds true, than the one described by Albert Michelson where “the future truths of physical science are to be looked for in the sixth place of decimals.”
Richard Jensen is a writer and historic preservation consultant in Sioux Falls, South Dakota. He has written for Aviation History, American History and South Dakota Magazine. He has also written on Walker Evans and Ted Jung, and for The Fortnightly Review, on the subject of Chernoybl.
- Their findings were published in the November 1887 issue of the American Journal of Science as “On the Relative Motion of the Earth and the Luminiferous Ether“. ↩
- This is not true, strictly speaking, for they found a variation that was so slight that it was well within the range of error for their apparatus. ↩
- There have been a few transits recorded off earth. ↩