Herbert asks us to imagine an observer in a space laboratory (L') who wishes to synchronize a pair of clocks equidistant from a light source from which he sends, at the same moment, a signal to each clock. 'Since the light signals leave the midpoint simultaneously and travel at the same speed, it seems that they must arrive at, and activate, the two clocks Thus the clocks are synchronized. However, a second space laboratory (L) also plays a role in this scene. two laboratories lie on paths parallel to that of the light signals, and their distance from each other is changing rapidly and uniformly. Knowing this, one can reason as follows: The distance between (L') and (L) is changing. Two equally valid and useful accounts of this fact can be given. We can consider (L') to be at rest, with the consequence that the changing distance in question must be laid to the motion of (L). Or, with equal validity and arbitrariness, we can deem (L) to be at rest, with the consequence that the changing distance in question must be laid to the motion of (L'). Suppose that we choose the second account, in which the clock-carrying laboratory is in motion. Under this account it must be denied that the clocks are synchronized because, due to (L')'s 'forward' motion, the signals have unequal distances to travel: the forward clock is running away from its signal with the result that the signal must travel more than half the distance separating the clocks, while the signal to the advancing aft clock travels less than half the distance. result is that the clocks are not activated Suppose, on the other hand, we choose the equally valid and arbitrary first account of the changing distance between the laboratories, the account in which the clock-carrying laboratory is at rest. Under this account, it must be affirmed that the clocks are synchronized, for because the signals travel equal distances the clocks are activated simultaneously. Clearly, whether the clocks are activated simultaneously or