For purposes of economy in investment and maintenance, Edison devised subsequently a plan by which the number of conducting lines was reduced to two, instead of five. The verticals of the letters were perforated only four holes high, and the four rollers were arranged in pairs, one pair being slightly in advance of the other. There were, of course, only four pins at the receiving instrument. Two were of iron and two of tellurium, it being the gist of Edison's plan to effect the marking of the chemical paper by one metal with a positive current, and by the other metal with a negative current. In the following diagram, which shows the theory of this arrangement, it will be seen that both the transmitting rollers and the receiving pins are arranged in pairs, one pair in each case being slightly in advance of the other. Of these receiving pins, one pair--1 and 3--are of iron, and the other pair--2 and 4--of tellurium. Pins 1-2 and 3-4 are electrically connected together in other pairs, and then each of these pairs is connected with one of the main lines that run respectively to the middle of two groups of batteries at the transmitting end. The terminals of these groups of batteries are connected respectively to the four rollers which impinge upon the transmitting drum, the negatives being connected to 5 and 7, and the positives to 6 and 8, as denoted by the letters N and P. The transmitting and receiving drums are respectively connected to earth.
In operation the perforated tape is placed on the transmission drum, and the chemically prepared tape on the receiving drum. As the perforated tape passes over the transmission drum the advanced rollers 6 or 8 first close the circuit through the perforations, and a positive current passes from the batteries through the drum and down to the ground; thence through the earth at the receiving end up to the other drum and back to the batteries via the tellurium pins 2 or 4 and the line wire. With this positive current the tellurium pins make marks upon the paper tape, but the iron pins make no mark. In the merest fraction of a second, as the perforated paper continues to pass over the transmission drum, the rollers 5 or 7 close the circuit through other perforations and t e current passes in the opposite direction, over the line wire, through pins 1 or 3, and returns through the earth. In this case the iron pins mark the paper tape, but the tellurium pins make no mark. It will be obvious, therefore, that as the rollers are set so as to allow of currents of opposite polarity to be alternately and rapidly sent by means of the perforations, the marks upon the tape at the receiving station will occupy their proper relative positions, and the aggregate result will be letters corresponding to those perforated in the transmission tape.
Edison subsequently made still further improvements in this direction, by which he reduced the number of conducting wires to one, but the principles involved were analogous to the one just described.
This Roman letter system was in use for several years on lines between New York, Philadelphia, and Washington, and was so efficient that a speed of three thousand words a minute was attained on the line between the two first-named cities.
Inasmuch as there were several proposed systems of rapid automatic telegraphy in existence at the time Edison entered the field, but none of them in practical commercial use, it becomes a matter of interest to inquire wherein they were deficient, and what constituted the elements of Edison's success.
The chief difficulties in the transmission of Morse characters had been two in number, the most serious of which was that on the receiving tape the characters would be prolonged and run into one another, forming a draggled line and thus rendering the message unintelligible. This arose from the fact that, on account of the rapid succession of the electric impulses, there was not sufficient time between them for the electric action to cease entirely. Consequently the line could not clear itself, and became surcharged, as it were; the effect being an attenuated prolongation of each impulse as manifested in a weaker continuation of the mark on the tape, thus making the whole message indistinct. These secondary marks were called "tailings."
For many years electricians had tried in vain to overcome this difficulty. Edison devoted a great deal of thought and energy to the question, in the course of which he experimented through one hundred and twenty consecutive nights, in the year 1873, on the line between New York and Washington. His solution of the problem was simple but effectual. It involved the principle of inductive compensation. In a shunt circuit with the receiving instrument he introduced electromagnets. The pulsations of current passed through the helices of these magnets, producing an augmented marking effect upon the receiving tape, but upon the breaking of the current, the magnet, in discharging itself of the induced magnetism, would set up momentarily a counter-current of opposite polarity. This neutralized the "tailing" effect by clearing the line between pulsations, thus allowing the telegraphic characters to be clearly and distinctly outlined upon the tape. Further elaboration of this method was made later by the addition of rheostats, condensers, and local opposition batteries on long lines.
The other difficulty above referred to was one that had also occupied considerable thought and attention of many workers in the field, and related to the perforating of the dash in the transmission tape. It involved mechanical complications that seemed to be insurmountable, and up to the time Edison invented his perforating machine no really good method was available. He abandoned the attempt to cut dashes as such, in the paper tape, but instead punched three round holes so arranged as to form a triangle. A concrete example is presented in the illustration below, which shows a piece of tape with perforations representing the word "same."