Thanks for the link to Jaki's paper - it's a very interesting body of science-history research (although there is a glaring error on the 6th page: "E = M/c^2", dividing by c^2 instead of multiplying).

However I find only 3 mentions in Jaki's paper of the corpuscular theory, 1 of which refers to Newton, and 2 to Soldner in the early 1800s. I don't see a statement that the theory still had support from Einstein or his leading contemporaries in the 1910s/1920s. On the contrary, Jaki complains that they must have been either ignorant or unfairly dismissive of Soldner's work on gravitational deflection of light corpuscles. And one must be aware of the difference between supporting a theory and considering a theory in order to basically test for its shortcomings. A year before embarking on the 1919 solar eclipse expedition, Eddington did just that, skeptically evaluating Newtonian gravity with respect to GR gravity ("Report On The Relativity Theory Of Gravitation", pages 54 - 56). His disposition could be summed up as being confident in the correctness of GR, but still withholding final judgement until experimental tests could put clear distance between it and the Newtonian competition. Good science should indeed spend time to weigh all serious theories against the evidence, even theories that already seem to have been overtaken by something better.

Meanwhile, Jaki's article portrays "Philipp Lenard, leader of the anti-Einstein crusade" as someone who delighted in Soldner's work (and by implication, entertained corpuscular theory) purely because it fitted with his "anti-Einsteinian polemics". But we cannot take his polemical position seriously as detached, evidence-based science.

I see that there was also an RG Q&A thread on this topic, to which you contributed recently: https://www.researchgate.net/post/How_can_Newtons_gravity_theory_predict_that_light_is_bent_when_passing_close_to_a_star

I like the clarity of K. Kassner's answer, which also points to why corpuscular theory still had some hangers-on in the 1910s: "The answer is half the value predicted by Einstein in 1915 (but Einstein predicted the same value as von Soldner in 1911, albeit based on the wave picture of light)". The likes of Lenard took sustenance from the fact that particles in Newtonian gravity deflect to the same extent as waves in GR (at least in Einstein's earlier calculation).

Thanks Ray Butler for your opinion. Maybe people kept an eye in the old Newtonian theory for a "just in case" situation. What is not clear to me is when Kassner, like others I know, use the idea of "zero mass particles". I saw this idea as a way to consider waves as particles, a "zero mass particle" being a wave. To me one big difference is the absence of collision. An e-m wave just colides with particles that have charge, and never with another e-m wave.

Newton predicted that the refraction of light in denser fluids was caused by his corpuscles moving faster in those fluids than in air, while Huygens' wave and wave-front theory predicted that light must move more slowly in dense liquids than in air. That happens to be correct, and as a result, when I first started studying physics I was given the impression that Huygens' theory had been accepted right from the start. But that was wrong. The wave theory was rejected right from the start, because if light was a wave, it seemed necessary that there be something for it to wave, the (incorrect) idea of the luminiferous ether was a couple of centuries away, and therefore Newton's corpuscular theory was the preferred one until it became possible to measure the speed of light in different materials. That was impossible with the technology available in the late 1600's, all of the 1700's, and the first half of the 1800's, so throughout that time, the corpuscular theory reigned supreme. It was only when Foucault finally succeeded in showing that light moves slower in dense liquids in 1850 that Huygens' theory of wave-fronts finally replaced Newton's corpuscles.

My knowledge, although maybe incomplete, makes me think differently regarding the subject: from my reading of the book "Opticks" by Newton, in my memory, he considered light as impulsions, such that vibrations in ressonance were the explanations for the interference rings that he studied. Light being a particle could explain refraction because the velocity of a particle in a fluid is reduced by friction, causing the bending as in Snell's Law. But such a path can not be considered as a straight line, energy is being lost all along it. To my knowledge, the wave theory was accepted after Arago's experiment showing that at the center of a shadow a bright region appears, as Poisson enunciated (as not being possible, certainly he was shocked with the experience).

An alternative view: With logical explanations about constituents, creation, sustenance and mechanisms of actions of corpuscles, corpuscular theory of light is correct even today. See: Nature of light ( http://vixra.org/abs/1312.0130 ) , ‘MATTER (Re-examined)’ - www.matterdoc.info .

Nainan Varghese, thanks for contributing. I will not read the article because it starts stating light as matter in a filosofical-mistical way, out of any law of physics, classic or modern. I need to spare time, that is another reason. My assertion is that light can not colide with light, as it would do if it would be a "corpuscle". I am curious in this days thinking on the interaction of electromagnetic waves with neutrons: without charge, there would be no interaction, I guess, and the wave traverses neutrons as if they were not there.