It was intended that great care should be taken in driving the spikes, that they were in the proper place, square with the rail, and left sticking up about an inch. The ties, of course, were all adzed down before the day of change. "Handspikes" were originally used to throw the rails, as were lining bars. We found, however, that small "cant hooks" were more easily handled and did better work. The first were made like Fig. 38, with a spike in the end of a stick, while the hook was fastened with a bolt about 10 or 12 inches above the foot. [Illustration: FIG. 38 and FIG. 39] We afterward made them of a 11/4 in. rod, 31/2 ft. long, pointed at one end, with a ring shrunk on 1 ft. from the bottom. Then the hook was made with an eye, as shown in Fig. 39, which slipped down over the top of the main rod. This was simple and cheap, and the iron was to be used for repair purposes when this work was done. Upon the system with which the writer was connected we had some branches where we could experiment upon the moving of the rail. Between Selma and Lauderdale the traffic was light, and at Lauderdale it connected with the Mobile & Ohio Railroad, which was narrow, and to which all freight had to be transferred, either by hoisting the cars or by handling through the house. By changing our gauge we would simply change the point of transfer to Selma. Here was a chance to experiment upon one hundred miles and cause little trouble to traffic. We could see the practical workings of our plans, and, at the same time, leave less to do on the final day. Upon the 20th of April we did this work. It had been our plan to do it somewhat earlier, but floods prevented. Most of the rail was old chair iron, short, and consequently more time was used in making the change than would have been required had our work been on fishplate rail. Our sections here were about eight miles long, and we arranged our men on the basis blocked out by the committee, viz., 24 to 26 men to the section, consisting of 6 spike pullers, 4 throwing rails, 12 spikers, 2 to push the cars and carry water. We soon found 5 ft. cars useless, and threw them into the ditch to be picked up at some future time. The men were spread out so as not to be in each other's way, and when the organization was understood and conformed to, it worked well. One gang changed 5 miles in 5 hours and 10 minutes, including a number of switches. We found, however, and it was demonstrated still more strongly on later work, that after 5 or 6 miles the men began to lag. We believed we had the best results when we had sections of about that length. It was arranged that two sections, alternately, commenced work together at one point, working from each other and continuing until the force of another section was met, working from the opposite direction. The foreman in charge was expected to examine the work and know that all was right. The push car which followed was a good test as to gauge.