Free Radicals - Michael Brooks [23]
In 1910, at the age of forty-two, he finally published a value for e. It was meant to be his career-defining publication. Eventually, it was – but thanks to scientists working in the German-speaking world, Millikan still had years of difficult and dirty work ahead of him.
The Austrian physicist Felix Ehrenhaft, a strange and intense character himself, quickly disputed Millikan’s result. Ehrenhaft had carried out a similar set of experiments and come up with a significantly smaller charge for the electron. In contrast to Millikan’s work, Ehrenhaft’s experiments seemed to show that electrical charge can be infinitely small. There is no fundamental, minimum unit of charge, Ehrenhaft said; there is no ‘electron’. Millikan now had to convince the world that he, not Ehrenhaft, was right. The series of experiments the desperate Millikan then performed were to cast a lasting shadow over his scientific integrity.
The task facing Millikan was to refute Ehrenhaft’s claim by showing that the value of charge on the oil droplets was never less than e. He was working alone now; Fletcher had gained his doctorate and promptly left for a post somewhere else – possibly anywhere else. Millikan took three years to complete the experiment to his satisfaction, and the notebooks he used to record the data show an array of untidy scrawls, enthusiastic exclamations, and rickety rows and columns of numbers. It is clear that Millikan never expected them to come under close scrutiny.
Unfortunately for Millikan’s reputation, the historian Gerald Holton retrieved the notebooks from the Caltech archives in 1980. Holton wanted to examine how the clean public face of science compares with the messy process of laboratory work. He was not expecting to spark a controversy that would still be raging decades later.
Much has now been written on the subject of Millikan’s honesty. According to the Harvard biologist Richard Lewontin, Millikan ‘went out of his way to hide the existence of inconvenient data’. Goodstein, self-appointed defence counsel for Millikan, says that his hero ‘certainly did not commit scientific fraud in his seminal work on the charge of the electron’. So where does the truth lie?
The debate hangs on a phrase in Millikan’s 1913 paper. In 1910, Millikan had published a value for e that is only 0.5 per cent off the value we use today (the error was largely due to his choosing a plausible but mistaken value for the viscosity of air). The 1913 paper was an attempt to refute Ehrenhaft and show that every measurement of electric charge gives a value of e or an integer multiple of e. In his 1913 paper, Millikan says that his data table ‘contains a complete summary of the results obtained on all of the 58 different drops upon which complete series of observations were made’. The statement is written in italics, as if to give it special weight. The notebooks for the 1913 paper show that Millikan actually took data on 100 oil droplets. The question raised by those who seized upon Holton’s analysis was this: did Millikan cherry-pick the data in order to confirm his original result and crush Ehrenhaft underfoot?
He certainly had motive. In Millikan’s 1910 paper he had made the ‘mistake’ of full disclosure. In its pages, he made statements such as, ‘Although all of these observations gave values of e within 2 percent of the final mean, the uncertainties of the observations were such that … I felt obliged to discard them.’ Another one is more damning: ‘I have discarded one uncertain and unduplicated observation, apparently upon a singly charged drop, which gave a value of the charge on the drop some 30 percent lower than the final value of e.’ This admirable honesty about the selection of data points had given Ehrenhaft ammunition that he used enthusiastically in his long feud with Millikan. Perhaps,