Keith Treschman published this article in the Asian Journal of Physics in 2014.
There were three astronomical tests of general relativity. Besides the gravitational bending of light, there were the anomalous advance of the perihelion of Mercury and gravitational redshift. The early history of these latter two tests is addressed here. For Mercury, data for its position were obtained principally from transit phenomena. Le Verrier was the first to account for the known perturbation effects on the elliptical orbit of Mercury and calculated an unexplained discrepancy. This was supported by Newcomb who revised the figure. With the use of his general theory of relativity, Einstein appeared to calculate this disagreement from Newtonian principles. Yet, other avenues needed to be explored before an acceptance of general relativity as a reasonable paradigm. This is part of a more general query of when should scientists endorse a theory. For the test of the redshift of radiation in the presence of a gravitational field, support for this phenomenon followed a winding route. Many factors, which could contribute to the redshift of spectral lines needed to be nominated, and their individual contribution, if any, had to be teased from the rest. Very small measurements had to be effected. This situation received some respite when measurements moved from the Sun to large mass objects such as white dwarfs which theory suggested should have a much larger redshift. 1928 was taken as the year in which the results could be interpreted as supporting general relativity. However, developments opened up subsequently and further confirmation has continued to the present day. The story is threaded with a theme that new ideas in science follow anything but a straightforward course and that real history is much more interesting. ©Anita Publications. All rights reserved.
Treschman, Keith. (2014). Early Astronomical Tests of General Relativity: the anomalous advance in the perihelion of Mercury and gravitational redshift. Asian Journal of Physics. 23 (1 & 2):171-188.