Having placed the empirical foundation of the principles of the constancy of the speed of light in axiomatic form and recognizing that several of these axioms have not yet received conclusive experimental support, it will now be of interest to consider an experiment to confirm the light principle which I had not mentioned, but to which Born has referred.

The eclipses of the moons of Jupiter may be used for an astronomical measurement of the speed of light. It is well known that the delays in the eclipses of a moon will progressively increase over the course of a year; the resulting overall delay corresponds to the time that the light has taken to traverse the length of the axis of the Earth's orbit. Hence, we are measuring the speed of light in a one directional sense. Now Maxwell has already pointed out that the speed of light must be different depending upon whether it travels with or against the direction of the orbit of the Earth because the inertial system S, in which the sun and the elliptical orbits of the planets are at rest, will itself have a velocity V (which Born calls v) with respect to a preferred inertial system J, which according to the older theory is that in which the aether is at rest; the speed of light must therefore really be c + V in one direction and c − V in the other.