In the early days it was actually proposed to telegraph pictures by ordinary telegraphy, using this principle. The suggestion was to agree upon a code of twenty-six shades, each called by a letter of the alphabet. One shade was to be _a_, the next _b_, and so on. Then the picture was to be divided up into squares, and the particular shade of each square telegraphed by means of the corresponding letter. The shades thus communicated were to be put together at the receiving end, on a prearranged system, and so the picture was to be built up. Given plenty of time, that scheme might be moderately successful, but to get a really good reproduction the subdivision needs to be so minute, and the number of squares, therefore, so immense, that it would be quicker to send the picture by train than to telegraph it by such laborious means. In a fairly coarse half-tone block the squares are, say, 2500 to the square inch. That number of letters would therefore have to be telegraphed for every square inch of picture transmitted, to say nothing of the difficulty of building up a picture of such a great number of parts and giving to each the desired shade. That idea, abortive though it is in its crude form, illustrates very clearly the fundamental principle on which this work is done. The problem is really to devise a machine which will do that same thing rapidly and automatically divide up the original into a large number of squares, and then send an electric current to represent each square, such current by its strength to indicate the shade of the square: and finally a similar instrument is needed to act as receiver, and to reproduce those squares in the proper order, giving to each its correct shade. In practically all of them the mechanism is rotatory, the original being placed upon a drum which turns round under a stylus, or its equivalent, while the stylus gradually travels along from end to end after the manner of the needle of a phonograph, or else the same result being achieved by the drum itself having an endwise movement as well as a rotative one. The receiving instrument is of similar form, and both must start together, move at the same speed and indeed preserve a perfect correspondence with each other. If the distance be great between the two there may be difficulties due to the "retardation" of the currents passing between them. Electricity does not pass through long wires, particularly if they be under the sea, with anything like the quickness which we are apt to think. Over a short line and under favourable circumstances the receipt of a telegraph signal at the farther end is practically instantaneous, but on long lines, and under certain conditions, that is far from being the case.