In the 1920s German physicist Emil Rupp was widely considered the pre-eminent experimentalist of his time. By 1935, he had been exposed as a fraud and his reputation lay in tatters. The discovery of Rupp鈥檚 faked data threatened other reputations, too, most notably that of Albert Einstein. Keen to see experiments that confirmed his theories and his instincts, Einstein had relished Rupp鈥檚 results. The great theorist was in the midst of his efforts to understand the quantum world and his ideas were based partly on Rupp鈥檚 forged experimental evidence. But had Einstein been complicit in the fraud?
IS LIGHT a particle or a wave? In 1926, this was the central question in physics. Fortunately for Emil Rupp, then 28 years old and looking to make his mark on physics, his expertise lay in the best means available to investigate the question: canal rays.
Rupp had just published a landmark study of these beams of atoms and ions, which are emitted at one end of a gas discharge tube. His work showed that a hydrogen atom emerging from a discharge tube could emit a coherent, or unbroken, beam of light up to 15 centimetres long. This was an extraordinary result: a moving hydrogen atom was expected to stay coherent over a much shorter distance than the recorded stationary value of 3.5 centimetres. No one paid any attention to a paper pointing out that Rupp鈥檚 diagrams showed the vacuum pump was in completely the wrong place. The work caused something of a stir and secured Rupp a prestigious academic post.
Advertisement
Einstein was captivated by Rupp鈥檚 result. Already one of the world鈥檚 most celebrated scientists and deeply involved in developing quantum theory, Einstein realised that the long coherence length 鈥渄iscovered鈥� by Rupp would allow him to perform an experiment testing the wave-particle nature of light. The 鈥渨ire grid experiment鈥� required a light source 鈥� an excited atom 鈥� moving parallel to a grid of wires. If the light was emitted over a length of time as a wave, the light as seen from the other side of the grid would be 鈥渃ut up鈥� by the wires, just as the view from a train is momentarily interrupted by telegraph poles alongside the track. If, on the other hand, the light was emitted instantaneously, as a particle, the grid should have no effect.
It was not an easy experiment to do. Studying the properties of the light involved the use of an interferometer, an array of mirrors that would cause the light to create a series of dark and light bands on a screen. The exact nature of that interference pattern would reveal whether the classical wave picture or the quantum particle picture was right, but creating the pattern relied on light with a long coherence length 鈥� much longer than most sources could achieve. Only Rupp鈥檚 light would suffice.
Rupp鈥檚 boss was an anti-Semite who had publicly agitated against Einstein, so Einstein could not visit him. Instead, he wrote letters with detailed instructions on how Rupp should carry out the experiment. Jeroen van Dongen of Utrecht University in the Netherlands, a member of the team editing the in Pasadena, has now pieced together . Einstein, he says, had doubts about Rupp from the beginning. Nonetheless, he continued to work with him until he got the experimental results he wanted.
The first problem Einstein saw with the experimental set-up was a Doppler shift. As the excited atoms moved along the beam鈥檚 axis, their light would have suffered a change in frequency that would reduce their coherence length. Rupp hadn鈥檛 mentioned how he avoided this in his paper. Einstein wrote to Rupp, asking whether he had perhaps made the light pass through slits to narrow the beam and thus circumvent the Doppler shift problem. Three days later Rupp replied, saying he had done exactly what Einstein suggested.
Einstein wasn鈥檛 convinced that the slits would really solve the problem, though. The shift could only be avoided if one of the interferometer鈥檚 mirrors was rotated by just the right amount. Rupp must have done this accidentally, Einstein somewhat generously assumed. 鈥淭he success of your earlier experiments can certainly only be explained by an unconscious rotation of the mirror,鈥� he wrote.
It was of little consequence to Einstein anyway because the wire grid experiment would be insensitive to this effect. Einstein told Rupp he should do this experiment first, then go back and check that the mirror鈥檚 rotation preserved the coherence length.
Rupp鈥檚 first set of 鈥減ositive鈥� results disappointed Einstein, who saw that Rupp had conflated two different canal ray experiments in his setup. Einstein urged him to start again. The next set of results was no more satisfying: Rupp had made a correction by moving a slit, for instance, which Einstein knew would have made no difference. Rupp鈥檚 next letter was unclear about what he had done to change things, but it was up to Einstein, he wrote, to 鈥渄ecide how these results compare to the theory鈥�.
Having seen Einstein applying rigorous criticism to the experiments thus far, van Dongen found his verdict on this set of results both surprising and revealing. Despite Rupp鈥檚 obvious confusion, Einstein chose to ignore all the previous problematic results and settled for the last results that Rupp had sent 鈥� the ones which confirmed his expectations. 鈥淚nitially, he is very critical of the results. Then, when he sees results he was expecting, he no longer questions them,鈥� notes van Dongen.
According to Einstein, Rupp鈥檚 experiments were 鈥渇ully satisfying鈥� and 鈥渁 convincing confirmation鈥� of his thinking. Now all that remained was to redo the original experiment in which Rupp had 鈥渦nconsciously鈥� rotated a mirror. If this were performed explicitly, it would be confirmation of everything Einstein already believed to be true.
Rupp鈥檚 next report stated that he had gained perfect results with the rotating mirror more than a dozen times. By this time, though, Einstein had identified another problem with the experiment: that hydrogen atoms at room temperature jiggle about and collide with one another. This would also reduce the coherence length to no more than 3.5 centimetres 鈥� and no amount of mirror rotation would cure it. In theory, there was no way the interference effects could work. 鈥淚t is not at all clear how interferences with such long path differences are possible,鈥� Einstein wrote to Rupp, who clearly felt exposed and began to backtrack. But Einstein pushed on and he had their two papers published back to back.
Even with Einstein鈥檚 uncritical support, time was running out for Rupp. Many physicists were suspicious of his results and Rupp was eventually forced to admit that he had superimposed and recopied photographic images to achieve many of the pictures he had published. He also had to defend himself against critical papers that pointed to other discrepancies in his publications. This defence required ever more elaborate deceptions and, in the end, Rupp overplayed his hand, persuading printers to alter one of his journal papers at the last minute. The journal鈥檚 editor found out and told the eminent physicist Walter Gerlach, who in turn contacted Einstein. Though aware of the growing controversy, Einstein kept silent.
In 1933, desperate to shore up his ailing reputation, Rupp went too far again, publishing a set of ground-breaking results that set alarm bells ringing among suspicious colleagues. When they pointed out that Rupp鈥檚 lab did not contain the equipment necessary to produce such results, Rupp had to admit that they, too, were fraudulent. He retracted his publications, suffered a nervous breakdown and was soon languishing in a sanatorium.
Gerlach put the final nail in Rupp鈥檚 coffin by pointing out that Einstein had made a fundamental error when drawing the rotating mirror experiment. The great theorist had indicated the wrong direction for the rotation. As Rupp鈥檚 experimental paper had followed Einstein slavishly, Rupp could never have seen the interference effects he had claimed. It was suddenly clear to all that every one of Rupp鈥檚 results had been fabricated. The German Physical Society advised its members it would no longer publish Rupp鈥檚 work, and that they should no longer cite it.
In 1936, faced with suggestions that his theory should be withdrawn, Einstein defended his brief collaboration. 鈥淚 do not consider my considerations of those days to be superfluous or false,鈥� he said. So was Einstein complicit in the fraud? Van Dongen thinks not. He believes Einstein鈥檚 only crime lay in 鈥渢heoretical prejudice鈥�: he was too keen to see his beliefs backed up by experiments. 鈥淚 find it hard to blame Einstein,鈥� van Dongen says. 鈥淗e was misled: that鈥檚 the most honest assessment one can give.鈥�
鈥淓instein was too keen to see his beliefs backed up by experiments鈥�
