This division between events whose order of occurrence is absolute and those whose order is relative is obtained as follows: Whenever it is possible for a ray of light to pass from one event to a second event, leaving the first event at the instant it is produced and reaching the second event before or even at the instant at which it is produced, the time sequence of the two events cannot be reversed by relative motion; it is absolute. On the other hand, if it is impossible for the ray of light to satisfy the preceding requirements, owing to the too great distance or the too small period of time separating the two events, the time sequence of the two events becomes indeterminate. One observer may claim that precedes , whereas the other may claim that precedes , and yet another may assert that and are simultaneous. If we remember that in Einstein’s theory no active influence, no propagation, can be transmitted with a speed greater than light, we see that when a ray of light cannot proceed from to , leaving when is produced and reaching when or before is produced, it is quite impossible for us to suppose that has influenced . No causal connection can exist between these two events. Hence we see that when two events are causally connected, their order of occurrence will remain the same for all observers. Only when no causal connections could possibly exist can the temporal sequence of the events be reversed by an appropriate motion of the observer. [Pg 218] It is not because a causal connection happens to exist between two events that the order is non-reversible; the non-reversibility is due entirely to the spatio-temporal separation of the events. As for the causal connection, it may or may not exist. The only reason we mention causal connections with reference to this problem is in order to show that there is nothing in the relativity theory to suggest a reversal of causality; hence, though Einstein’s theory entails the relativity of simultaneity and in certain cases that of temporal sequence, there is no danger of a cause appearing as an effect or vice versa. In particular, no relative motion of the observer could ever lead him to believe that the glass we had let fall had in reality leaped up from the floor into our hand. No danger of the black and white powders which we had shaken into a grey mixture, appearing to separate back into black and white, as a result of a change in our relative motion. In short, no danger of the principle of entropy being overthrown.