Things went fairly well for a time on that memorable Thursday afternoon, when all the laboratory force made high holiday and scrambled for foothold on the locomotive for a trip; but the friction gearing was not equal to the sudden strain put upon it during one run and went to pieces. Some years later, also, Daft again tried friction gear in his historical experiments on the Manhattan Elevated road, but the results were attended with no greater success. The next resort of Edison was to belts, the armature shafting belted to a countershaft on the locomotive frame, and the countershaft belted to a pulley on the car-axle. The lever which threw the former friction gear into adjustment was made to operate an idler pulley for tightening the axle-belt. When the motor was started, the armature was brought up to full revolution and then the belt was tightened on the car-axle, compelling motion of the locomotive. But the belts were liable to slip a great deal in the process, and the chafing of the belts charred them badly. If that did not happen, and if the belt was made taut suddenly, the armature burned out--which it did with disconcerting frequency. The next step was to use a number of resistance-boxes in series with the armature, so that the locomotive could start with those in circuit, and then the motorman could bring it up to speed gradually by cutting one box out after the other. To stop the locomotive, the armature circuit was opened by the main switch, stopping the flow of current, and then brakes were applied by long levers. Matters generally and the motors in particular went much better, even if the locomotive was so freely festooned with resistance-boxes all of perceptible weight and occupying much of the limited space. These details show forcibly and typically the painful steps of advance that every inventor in this new field had to make in the effort to reach not alone commercial practicability, but mechanical feasibility. It was all empirical enough; but that was the only way open even to the highest talent.