Conservative signaling is all in favor of the overlap, on account of the safety factor, in case the signal is accidentally overrun. Another consideration was the use of automatic train stops. These stops are placed at the home signals, and it is thus essential that a stopping distance should be afforded in advance of the home signal to provide for stopping the train to which the brake had been applied by the automatic stop. Ordinarily, the arrangement of overlap sections increases the length of block sections by the length of the overlap, and as the length of the section fixed the minimum spacing of trains, it was imperative to make the blocks as short as consistent with safety, in order not to cut down the carrying capacity of the railway. This led to a study of the special problem presented by subway signaling and a development of a blocking system upon lines which it is believed are distinctly in advance of anything heretofore done in this direction. [Illustration: REAR VIEW OF BLOCK SIGNAL POST, SHOWING TRANSFORMER AND INSTRUMENT CASES WITH DOORS OPEN] Block section lengths are governed by speed and interval between trains. Overlap lengths are determined by the distance in which a train can be stopped at a maximum speed. Usually the block section length is the distance between signals, plus the overlap; but where maximum traffic capacity is desired the block section length can be reduced to the length of two overlaps, and this was the system adopted for the Interborough. The three systems of blocking trains, with and without overlaps, is shown diagramatically on page 143, where two successive trains are shown at the minimum distances apart for "clear" running for an assumed stopping distance of 800 feet. The system adopted for the subway is shown in line "C," giving the least headway of the three methods. [Illustration: PNEUMATIC TRACK STOP, SHOWING STOP TRIGGER IN UPRIGHT POSITION] The length of the overlap was given very careful consideration by the Interborough Rapid Transit Company, who instituted a series of tests of braking power of trains; from these and others made by the Pennsylvania Railroad Company, curves were computed so as to determine the distance in which trains could be stopped at various rates of speed on a level track, with corrections for rising and falling to grades up to 2 per cent. Speed curves were then plotted for the trains on the entire line, showing at each point the maximum possible speed, with the gear ratio of the motors adopted. A joint consideration of the speeds, braking efforts, and profile of the road were then used to determine at each and every point on the line the minimum allowable distance between trains, so that the train in the rear could be stopped by the automatic application of the brakes before reaching a train which might be standing at a signal in advance; in other words, the length of the overlap section was determined by the local conditions at each point.