Suppose a sheet of paper so prepared as to be a conductor of electricity, and that a message is written on the paper with some non-conducting substance for ink. If that sheet were passed between the knobs at _a_ (the handle H being pressed down by a spring), whilst simultaneously a sheet of Bain’s chemically prepared paper were passed athwart the steel pointer at the receiving station, there would be traced across the last-named paper a blue line, which would be broken at parts corresponding to those on the other paper where the non-conducting ink interrupted the current. Suppose the process repeated, each paper being slightly shifted so that the line traced across either would be parallel and very close to the former, but precisely corresponding as respects the position of its length. Then this line, also, on the recording paper will be broken at parts corresponding to those in which the line across the transmitting paper meets the writing. If line after line be drawn in this way till the entire breadth of the transmitting paper has been crossed by close parallel lines, the entire breadth of the receiving paper will be covered by closely marked blue lines except where the writing has broken the contact. Thus a negative facsimile of the writing will be found in the manner indicated in Figs. 8 and 9.[32] In reality, in processes of this kind, the papers (unlike the ribbons on Bain’s telegraph) are not carried across in the way I have imagined, but are swept by successive strokes of a movable pointer, along which the current flows; but the principle is the same. [Illustration: FIG. 8. FIG. 9.] It is essential, in such a process as I have described, first, that the recording sheet should be carried athwart the pointer which conveys the marking current (or the pointer carried across the recording sheet) in precise accordance with the motion of the transmitting sheet athwart the wire or style which conveys the current to the long wire between the stations (or of this style across the transmitting sheet). The recording sheet and the transmitting sheet must also be shifted between each stroke by an equal amount. The latter point, is easily secured; the former is secured by causing the mechanism which gives the transmitting style its successive strokes to make and break circuit, by which a temporary magnet at the receiving station is magnetized and demagnetized; by the action of this magnet the recording pointer is caused to start on its motion athwart the receiving sheet, and moving uniformly it completes its thwart stroke at the same instant as the transmitting style.