speed at some part of the course before the end. The average speed of a record-making Olympic runner on a 100-meter course is given in the table as 9.26 meters per second. In the 1,500-meter race, Sheppard averaged, as we have already seen, 6.16 meters-per-second. Suppose that he commenced, say, by running at 9.26 meters-per-second. This would have been only 50 per cent, more than his average speed. It is clear that, had he done so, he would have been run out in 10 seconds. Again, if he had commenced by running at 7.09 meters-per-second, his average speed over the 800-meter course, and not quite 15 per cent, above his average speed over the 1,500-meter course, he would have been run out after 112.8 seconds, or only about half way.
It seems possible, however, to combine the incentive of shoulder-to-shoulder competition with uniform pace-making, and without loss of interest to the spectators, by running a light flag or pennant by the side of the track, on a slender wire of steel or phosphor-bronze. It would only be necessary to set short posts beside the track, each supporting a light metallic guide-pulley. Over all these pulleys would run the wire alongside the track, making a complete loop or endless chain. The wire would be propelled at some point in the course by a small electric motor, driven by a portable storage battery, as in the outfit of an electric automobile. An attendant at the motor would be charged with the duty of keeping the speed of the motor and wire uniform at that corresponding to the record for the particular event. By means of a stroboscopic fork, i. e., a tuning-fork carrying slotted wings on its prongs, through which a rotating target carried by the motor appears to stand still, it is readily possible to keep the speed of such a motor and wire constant to within a small fraction of one per cent.
When the runners were placed and ready to start, a small flag would be gripped on the running wire a few paces behind the men. As this flag reached the starting line, the starter would fire his pistol. Owing to the starting inertia of the men, the flag would gain a few feet at the first, and the runners would get under way with the flag slightly ahead. Since the flag would reach the winning post in record time, it would be the object of the men to outdistance it at that point. According to the reasoning above presented, they should best be able to do this by keeping close to the flag, which would serve as pacemaker. They should certainly be advised thereby if they started off at too high a speed. The spectators would have the advantage of seeing not only the contest of the actual runners; but also a contest with the "ghost" of the best runner that heretofore had made the record of that event his own, as impersonated in the flag running beside the track.
In raising the ghost of the record runner as above, there might be a danger of hurting the race by the runners losing heart if they failed to keep up with the flag. There might also be a danger of the specta-
