Where Is the 1919 Data Proving General Relativity?

If I do a search on the Internet with the expression “Genius of the 20th Century”, almost certainly the first entry on the list will be Albert Einstein (1879-1955). Einstein won the Nobel Prize in Physics in 1921 “for his services to Theoretical Physics, and especially for his discovery of the law of the photoelectric effect”¹. However, he is best known for his theories of special relativity (1905) and general relativity (1915).

Einstein became world-famous as a result of observations made during the May 29th, 1919 solar eclipse, which was visible in the tropics. There were two expeditions: the first, led by Arthur Eddington (1882-1944), took place in Príncipe, an island in the gulf of Guinea off the coast of Africa, and now part of the country called São Tomé and Príncipe; the second, led by Andrew Claude de la Cherois Crommelin (1865-1939) and Charles Rundle Davidson (1875-1970), took place in Sobral, in the northeast Brazilian state of Ceará. The oversight of the entire project was undertaken by Frank Watson Dyson (1868-1935), at the time the Astronomer Royal in Greenwich. The results of the analysis were announced in a joint meeting of the Royal Astronomical Society and of the Royal Society that took place on November 6th, 1919: the eclipse observations confirm the theory of general relativity.

Soon after, Einstein was hailed as a genius in The Times (of London), The New York Times and the Berliner Illustrirte Zeitung. For example, the latter had a full-page picture of Einstein², with the caption:

...Eine neue größte der Weltgeschichte: Albert Einstein, dessen Forschungen eine völlige Umwätzung unserer Naturbetrachtung bedeudet und den Erkentnissen eines Kopernikus, Kepler und Newton gleichwertig sind.

A Deepl translation gives:

A new greatest in world history: Albert Einstein, whose research meant a complete re-evaluation of our view of nature and is equivalent to the discoveries of Copernicus, Kepler and Newton.

December 14th, 1919 front page of Illustrirte Zeitung

The 1919 expeditions are written about extensively by Daniel Kennefick, in several research papers, as well as in his book No Shadow of a Doubt: The 1919 Expedition That Confirmed Einstein’s Theory of Relativity.³

The Príncipe expedition used one telescope, while the Sobral expedition used two telescopes. Here is a summary from an article published by the Royal Observatory, Greenwich (a suburb of London):

Both parties took identical 13-inch astrographic lenses of 11 feet 3 inches focal length. That taken to Sobral came from the Astrographic Telescope at Greenwich. That taken to Principe came from the Astrographic Telescope at Oxford. The object glasses were attached on arrival to hollow steel tubes (made by Harvey & Co.), with the lens at one end and the photographic plate holder at the other and then mounted horizontally so that the sunlight could be reflected into them by a 16 inch diameter flat mirror.

The flat mirror or coelostat, was slowly rotated to follow the Sun, using a clock driven by a heavy weight, that ran for about half an hour before the weight had to be rewound. Both astrographic lenses were stopped down to 8 inches diameter to improve their image quality.

The Greenwich Sobral group, also took a 4 inch diameter lens with a focal length of 19 feet, loaned to them by Father Cortie. This was fed by an 8 inch diameter coelostat loaned by the Royal Irish Academy….

Mounting the telescope horizontally avoids the need to transport and align a heavy equatorial mount and makes it easier to keep the apparatus cool and free of wind shake, but the flat coelostat mirror introduces an extra optical element, capable of spoiling the image quality.⁴

The analysis for the Príncipe observations was conducted by Eddington, while that for the Sobral observations was overseen by Dyson. It was concluded that the observations from the Príncipe telescope and from the 4-inch diameter lens in Sobral confirmed Einstein’s predictions. The 13-inch diameter lens from Sobral was excluded by Dyson, as the images seemed to be out of focus.

So this is what I learned when I was young. Now, you can imagine my surprise when I read the following paragraph in Kennedick’s book:

The Principe team was led by Eddington, in his capacity as director of the Cambridge Observatory. Eddington’s role was largely personal. He understood the theory being tested and was experienced in the type of astrometry required to measure the predicted effect. But his observatory was not known for its expertise in eclipses, and the equipment he used was largely borrowed from the Oxford Observatory. He was accompanied not by one of his own staff but by Cottingham, who was familiar with the equipment to be used, as he counted both the Oxford and Cambridge Observatories among his clients. Cottingham’s role was largely to maintain the equipment in working order at the site. It is probable that Eddington alone handled the data analysis of the Principe expedition. We cannot determine how this was done because none of the data analysis sheets or photographic plates have survived. We do know that Eddington began the data analysis on Principe by himself, so it is almost certain that he continued on his own when back in England. [p.15, my emphasis]

Wait, what? Neither Eddington’s data analysis sheets nor his photographic plates have survived?

The obscurity of Principe extended further than just the day itself. Eddington began his data reduction alone while still on the island. It is probable that he never enlisted anyone’s help after his return to Cambridge. If he kept any notes or records, they were later lost. Even the plates he took have long since disappeared. They may have been disposed of when the observatory was cleared out after the Second World War. Eddington had passed away, and the university was anxious to remove Winifred, his sister, from the premises in order to rehouse needy academics in the postwar housing shortage. A fellow astronomer called Chubby Stratton was sent in to go through Eddington’s papers and seems to have thrown out a great deal of material, judging by the relatively small amount deposited at the Trinity College Library. Apparently, no one thought the eclipse plates were worth keeping. No one now can examine what Stratton did and second-guess his methods and decisions. In this sense, it is not surprising that doubts have been expressed about Eddington’s results. Apart from what he published, we have nothing else to go on. To be fair, he published in some detail, but it would be nice to have the plates. Apart from the fabulous published images of the prominence, which show no stars because of the clouds, we have no record of the eclipse at Principe other than the data and tables of calculations given in the report. [pp.220-221, my emphasis]

This is for the observations that proved without a doubt that classical physics was wrong, and made Eddington a star and Einstein a superstar! So today we have no physical evidence that Eddington actually observed what he claimed to have observed.

Spectacular image of the 1919 eclipse as seen from Sobral, the result of applying modern image processing techniques — including image restoration, noise reduction, and removal of artifacts. It unveils stunning details in the solar corona and a giant prominence emerging from the upper right part of the Sun. (https://www.eso.org/public/images/potw1926a)

So much for the Príncipe plates. Now, what about the Sobral plates? These were under the supervision of Dyson. Kennedick surmises that Eddington had no say in the exclusion of the main Sobral telescope:

It seems likely that Eddington was never present at Greenwich during the Sobral data reduction. In a letter to Dyson on October 21, 1919 (RGO Archive 8, fol. 150), he refers to having acquired a season railway ticket, rather suggesting he had not been traveling to see Dyson in the preceding weeks or months. We can be fairly confident that Eddington simply was not privy to the reduction of the Sobral data or to the crucial decision, recorded in the data sheets (in what appears to be Dyson’s hand), to reject the astrographic data and accept the four-inch data as “the result of the Sobral expedition” (RGO Archive 8, fol. 150). The eclipse expedition appears to be a reasonably good case of independence being preserved between the two wings of the collaboration. [p.209]

Wait a moment! When does one buy a season railway ticket? When one realizes that one has been travelling so often that it will be more convenient or less costly to have a season ticket. Furthermore, by train in 1919, it would have taken only an estimated 80 minutes to travel from Cambridge to London⁵. Kennedick’s speculation that Eddington never went to see Dyson when the latter was analyzing the Sobral plates seems to me to be completely farfetched.

Anyway, to reassure ourselves, at least the Sobral plates are available, right? And they can be reanalyzed, right? Here is what Kennedick wrote:

In June 2003 I made a trip to Cambridge, primarily to look over Eddington’s papers at Trinity College and the RGO’s manuscript archive, now housed at the Cambridge University Library. I also had the idea of making inquiries about the original plates. To my surprise, I learned from Adam Perkins, the curator of the RGO archives, that a modern reanalysis of the data had already been done. This had occurred in 1978, to commemorate the Einstein centenary of 1979. […]

In 1978 most of the Sobral plates still survived intact (in contrast to the Principe plates). One eclipse plate and one comparison plate taken with the four-inch lens were missing, and one of its eclipse plates was broken. A few of the astrographic plates were discolored. In general nothing stood in the way of the project. The plate measurements were made by E. D. Clements (known as Clem) on the Zeiss Ascorecard at the RGO, by then relocated to Herstmonceux Castle in Sussex. This plate-measuring machine permitted the operator to measure accurately and record the positions of every star on each plate, both eclipse and comparison. The data produced in this way could then be entered into an electronic computer. […]

In 1978 the plates were re-measured individually on the Ascore-card machine at Herstmonceux; each image was centered in a square graticule and the (X,Y) co-ordinates were recorded by means of moiré fringe gratings. The reasonable results obtained, particularly for the “inferior” Sobral astrographic images, would seem to indicate that the problem with the 1919 measurements was not so much in the quality of the images, but rather in the reduction method, which relied very heavily on the experimental determination of the scale constant e. [pp.241-242, my emphasis]

So there was a reanalysis of the Sobral plates in 1979, and it showed that with the appropriate reduction method, the observations made by both Sobral telescopes, not just the smaller one, are consistent with Einstein’s predictions.

So everything turned out just fine, right? Well, maybe not.

In a 2019 interview with newspaper Revista Pesquisa FAPESP⁶ of the São Paulo Research Foundation, Kennefick stated:

Unfortunately, for some reason I’m unaware of, no data from this expedition has survived. The photographic plates were lost. I’ve talked to many archivists and no one knows what happened. The loss must have occurred more than 50 years ago. The plates from Sobral survived and were used in a reanalysis of the eclipse data, which was done by other researchers in 1979. However, I’ve never seen them. I talked to some astronomers about this. They say that after 1979, the Sobral plates were moved and no one could tell me exactly where they are. They must be mixed in among other plates. [my emphasis]

What? Again? Given that the Príncipe plates were already lost, and the importance of the reanalysis of the Sobral plates, the latter were also lost? How is this possible?!?

One of the key ideas in modern science is the repeatability of experiments. For direct observations of ongoing phenomena in Nature, repeatability is not possible. The heavens will never be exactly as they were on May 29th, 1919. So if we wish to reëxamine the observations made that day and the interpretations that were made therefrom, it is the raw data that is of most importance. In this case, the raw data is the plates. If these have disappeared, this reëxamination is no longer possible.

Now, the fact that the original plates from the 1919 eclipse observations have gone missing does not mean that general relativity has been falsified. Far from it.

Nevertheless, when the raw data from one of the most famous set of observations in human history has gone missing, one naturally speculates, What is going on? So, the topic of relativity, both special and general, will likely be the subject of numerous future posts.

Acknowledgements: This post was written while being involved in numerous exchanges with Hans G. Schantz (https://aetherczar.substack.com) and his colleagues. I am of course responsible for every word that I wrote. Nevertheless, I would like to thank them for this help.

---

If you wish to donate to support my work, please use the Buy Me a Coffee app.

1. The Nobel Prize in Physics 1921. https://www.nobelprize.org/prizes/physics/1921/summary.

2. https://pubs.aip.org/view-large/figure/85175666/48_1_PT.3.5310.figures.online.f1.jpg

3. Daniel Kennefick. No Shadow of a Doubt: The 1919 Expedition That Confirmed Einstein’s Theory of Relativity. Princeton University Press, 2019.

4. Robin Catchpole and Graham Dolan. General Relativity and the 1919 Solar Eclipse. http://www.royalobservatorygreenwich.org/articles.php?article=1283.

5. https://en.wikipedia.org/wiki/Royston_and_Hitchin_Railway.

6. Revista Pesquisa FAPESP. Daniel Kennefick: The importance of Sobral. December 2019. https://revistapesquisa.fapesp.br/en/daniel-kennefick-the-importance-of-sobral-2