The mirror instrument has this drawback, however--it does not 'record' the message. There is a great practical advantage in a receiving instrument which records its messages; errors are avoided and time saved. It was to supply such a desideratum for cable work that Sir William Thomson invented the siphon recorder, his second important contribution to the province of practical telegraphy. He aimed at giving a GRAPHIC representation of the varying strength of the current, just as the mirror galvanometer gives a visual one. The difficulty of producing such a recorder was, as he himself says, due to a difficulty in obtaining marks from a very light body in rapid motion, without impeding that motion. The moving body must be quite free to follow the undulations of the current, and at the same time must record its motions by some indelible mark. As early as 1859, Sir William sent out to the Red Sea cable a piece of apparatus with this intent. The marker consisted of a light platinum wire, constantly emitting sparks from a Rhumkorff coil, so as to perforate a line on a strip of moving paper; and it was so connected to the movable needle of a species of galvanometer as to imitate the motions of the needle. But before it reached the Red Sea the cable had broken down, and the instrument was returned dismantled, to be superseded at length by the siphon recorder, in which the marking point is a fine glass siphon emitting ink, and the moving body a light coil of wire hung between the poles of a magnet. The principle of the siphon recorder is exactly the inverse of the mirror galvanometer. In the latter we have a small magnet suspended in the centre of a large coil of wire--the wire enclosing the magnet, which is free to rotate round its own axis. In the former we have a small coil suspended between the poles of a large magnet--the magnet enclosing the coil, which is also free to rotate round its own axis. When a current passes through this coil, so suspended in the highly magnetic space between the poles of the magnet, the coil itself experiences a mechanical force, causing it to take up a particular position, which varies with the nature of the current, and the siphon which is attached to it faithfully figures its motion on the running paper.