the reality of the phenomenon, combined with a general explanation for what went wrong, was more than sufficient for scientists to reject the existence of the phenomenon.
The same principle of rejecting a finding even if no scientific flaw can be found in the experiment, on the ground that the result cannot be replicated, is universal in science. Science is littered with experiments reporting some particular result that, upon later attempts, fails to replicate.
A personal example makes the point. My master’s thesis (Hines 1976) examined a particular question in the field of hemispheric asymmetries in the human brain. In one of the experiments, I obtained the results that I predicted. I was quite excited by this. Then I went back and tried to replicate the findings. Two replications failed. To this day I have no idea why the initial experiment succeeded in providing the anticipated results. However, the two failures to replicate, along with similar failures that were later reported in the literature, convinced me that the initial positive result was incorrect.
In the cases of N Rays and polywater, one can not specify the exact date on which these aberrations came to the attention of the scientific community, much less the public at large. Such is not the case with cold fusion, the final scientific mistake to be considered here. On Thursday, March 23, 1989, chemists Stanley Pons and Martin Fleischmann held a news conference at the University of Utah that was widely covered on the national news programs that evening. They announced that they had discovered a method for producing nuclear fusion using simple equipment at room temperature. The term “cold fusion” was immediately coined to describe this phenomenon. Over the next few years great controversy raged over whether or not cold fusion was real or illusory. The history of the controversy has been examined in detail in three books (Close 1991; Huizenga 1992; Taubes 1993). Of the three, that by Gary Taubes’s is probably the most comprehensive. Rothman (1989–90) has published an article-length summary.
Nuclear fusion occurs when atomic nuclei are brought so close together that they stick to each other. The mass of the stuck-together nuclei is less than that of the separate individual nuclei. The “extra” mass is converted into energy, with the amount of energy determined by Einstein’s famous equation E = mc2, where E is energy,
If this was all there was to it, fusion would be easy. However, in order to get atomic nuclei close enough to actually fuse, the natural repulsive force that exists between nuclei when they are brought close together must be overcome. This takes a huge amount of energy. The conventional method of achieving fusion is to use heat on the order of hundreds of millions of degrees to sufficiently speed up the nuclei so that the repulsive forces will be overcome. Needless to say, this is both difficult and expensive. Still, since fusion would be a relatively clean source of energy (it produces little of the radioactive waste that nuclear fission does), it has been the object of considerable research funds. Given the clean nature of fusion energy and the great expense of obtaining it using the conventional superheated methods, one can understand how attractive a method would be for obtaining this sort energy using cheap apparatus at room temperature. This is exactly what Pons and Fleischmann claimed to have done at their news conference.
The news conference immediately raised suspicions on the part of many scientists. The absolutely standard way to announce a new scientific result is to first publish it in a professional scientific journal in the relevant field. Before a paper can be published in such a journal, it is usually carefully reviewed by other scientists who are experts in the area of research the paper concerns. These experts point out any flaws in the paper in terms of methodology, statistical analysis, citation of the relevant scientific literature, and the like. Only after a paper has passed this type of scrutiny, called
What did Pons and Fleischmann point to at their news conference that led them to argue that they had actually achieved cold fusion? There were two measures that led them to the conclusion that fusion was taking place in their small fusion “cells,” which were really little more than large jars that held electrodes of palladium or platinum in a bath of heavy water (D2O). One was heat. It was claimed that the cells produced more heat than would be expected if no fusion was occurring. The other measure was gamma rays. Fusion produces gamma rays and Pons and Fleischmann claimed to have found gamma rays, coming from their fusion cells. One might think that it is an easy matter to measure these two variables, especially heat. One would be wrong. Both variables require very sensitive measurement requiring much experience. Scientists with real experience in the measurement of these two variables in the conditions under which Pons and Fleischmann were operating were nuclear physicists. Pons and Fleischmann, as chemists, had relatively little experience with such measurements. Working outside one’s area of specialization can be dangerous, because one will probably be less aware of the subtleties and pitfalls of experimentation and measurement in a field in which one is not an expert. The history of parapsychology, for example, is littered with scientists respected in their own fields who embarrassed themselves by making blunders when they assumed, incorrectly, that they were also experts in other areas of experimentation.
Pons and Fleischmann’s unfamiliarity with measurement of gamma rays is one striking example of how they went wrong. Fusion should result in gamma rays with an energy of 2,224 KeV (thousand electron volts). Pons and Fleischmann did claim to measure gamma rays being emitted by their cold fusion cells, but in the media and at scientific talks they gave, they reported the gamma rays to have an energy of about 2,500 KeV. This is a major discrepancy—but Pons and Fleischmann were not aware of it. In general, chemists didn’t spot the problem, but physicists did. The initial reports of gamma rays at 2,500 KeV were made before Pons and Fleischmann (1989) published their first report of their findings. The problem had certainly been pointed out to them before their paper was published. When their paper actually appeared in print, an amazing thing had happened. The paper reported gamma rays at 2,224 KeV, and there was no indication in that paper that gamma rays at 2,500 KeV had ever been detected. Other changes in the data took place between the verbal and the media presentations and the published paper. The reasons for these changes are still not clear, but it is difficult to interpret them charitably.
A second factor that caused scientists to be suspicious of Pons and Fleischmann’s claims at the news conference was the fact that if the claims of cold fusion were true, several basic laws of physics would have been in jeopardy. Recalling that extraordinary claims demand extraordinary proof, it was obviously going to take more than claims unsubstantiated by published results to convince scientists that cold fusion had been demonstrated. However, since the apparatus used was relatively straightforward and was visible in the background in a videotape shown at the news conference, and since Pons and Fleischmann did give a general description of their basic method, for the next several months numerous laboratories around the world tried to replicate the results.
When groups of researchers attempt to replicate an exciting new finding, even if that finding is artifactual, some will “succeed” in the replication, while some will correctly fail to replicate. The seeming successes may be due to various types of errors of experimentation. In the heated atmosphere of the first few months following Pons and Fleischmann’s news conference, it was the allegedly successful replications that got the lion’s share of attention, both from the popular media and from Pons and Fleischmann, who tended to dismiss failures as being due to “not doing it right.” Alas, they failed to be explicit about what “doing it right” consisted of in terms of exact methodology. As time passed, it was established that the seemingly successful replications were due to various sources of errors, some quite subtle. When these sources of error were eliminated, so, too, was evidence for cold fusion. Close (1991) discusses many of these replications and shows what went wrong to mislead the researchers into thinking that they had obtained fusion.
One laboratory in particular seemed to be especially able to replicate Pons and Fleischmann’s findings of excess heat in their fusion cells—and to find tritium as well. Standard physical theory requires that tritium be produced as a result of fusion. Pons and Fleischmann had never reported finding any tritium, but tritium, as well as heat, was reported in cold fusion cells in the laboratory of John Bockris at Texas A & M University. The combined finding of tritium and heat was taken as strong support for the reality of cold fusion. Alas, the tritium appeared only sporadically and, like other cold fusion findings, could not be reproduced. The actual pattern of the appearance of tritium suggested the possibility that it was fraudulently being added to some cold fusion cells (Taubes 1993).
The finding of excess heat in the fusion cells in Bockris’s lab has a particularly interesting explanation. It is a
