D O C . 4 4 T H E T H E O R Y O F R E L AT I V I T Y 5 9 We heard earlier that a body in a linear, uniform motion with respect to a Gali- lean frame of reference K will perform an accelerated, generally curvilinear motion with respect to an accelerated frame of reference K I (the box) (this can also be in- terpreted as the effect of a gravitational field). A new result of fundamental impor- tance is obtained when one carries out the corresponding deliberations for a beam of light. With respect to K, the beam propagates along a straight line at the velocity c. With respect to the accelerated box (the frame of reference K I ), the path of the beam is, however, no longer straight. Light beams thus in general propagate cur- vilinearly in gravitational fields. This result is so important because it can be verified during solar eclipses. The deflection of light beams that pass near the Sun (which produces an enormous grav- itational field) must in fact cause the fixed stars that can be observed during a total solar eclipse to appear to move away from the sun by a very small angle (immedi- ately at the sun’s rim by 1.7 arc seconds) compared to the positions that they would occupy for us in the sky when the Sun was at some other position in the sky. The observations made thus far have confirmed the light deflection as predicted by the theory. Closely connected to this is the fact that the law of the constancy of velocity of light propagation in empty space does not hold in the same sense as in special rel- ativity, for a curvature of a light beam can occur only if the velocity of light prop- agation varies with the place. With this, however, the theory of special relativity does not fail (as was once claimed), one of whose main pillars is this law rather, it means nothing more than that special relativity holds only insofar as one can ne- glect the influences of gravitational fields.[5] One of the greatest achievements made possible by the theory of general relativ- ity was the purely theoretical derivation of the general law of gravitation (which we can, however, not treat in detail here). At the end of our considerations we should not fail to mention Minkowski’s[6] important idea, without which general relativity would perhaps have stalled at its very beginning: it is the concept of the four-dimensional world. The following should suffice: “space” is three-dimensional, because the position of a point at rest in it is determined by three coordinates. Analogously, the world of physical events, called for short simply the “world” by Minkowski, is four-dimensional, since in addition to the three spatial coordinates x, y, and z, a temporal coordinate t, the time at a particular moment, must be specified in order to be able to describe an individ- ual event. Earlier, spatial and temporal notions were kept separate now, however, “space in itself” and “time in itself” have been “dropped down to the realm of