So fast, then, will Washington have been moving to meet the message from Greenwich. So fast will Greenwich have been retreating from Washington's message.
Now the ultimate effect of motion on the time-determination cannot be calculated along any such simple lines as these. Indeed, it cannot be exactly calculated at all, for we have not all the data. But there is certainly some effect. Suppose one rows four miles up a river against a current of two miles per hour, at a rowing speed of four miles per hour. This will take two hours, plainly. The return trip with the river's gift of two miles per hour will evidently require but forty minutes. Two hours and forty minutes for the round trip, then, of eight miles.
Now then, to row eight miles in still water, according to our supposition, would have required but two hours. But, some one objects, the current must help the return trip as much as it hindered the outgoing! Ah, here is the snare that catches rough-and-ready common sense! How long would the double journey have taken if the river current had been faster than our rowing speed? How shall we schedule our trip if we cannot learn the correct speed, or if it varies from minute to minute?
These explanations are necessarily symbolistic rather than demonstrative, but any one who will seriously follow out these lines of thought, or, still better, study the attitude of the hard-headed modern physicist towards our classical geometry and mechanics, cannot fail to realize how conventional, artificial—even phantasmal—are the limitations set by the primitive idea of flat space and straight time.
The inferences which we may draw from our hypothetical experiment are plain. The settings of the two chronometers would be defective, they would not show the same time, but each of them would mark the local