catches in a notch corresponding to the blank on the type-wheel. The adjustment can also be made by hand.
Lastly, the shaft (I) carries a third cam, which, at each revolution of this axis, engages with a very coarse-toothed wheel (T'), set on the same axis as the type-wheel, and pushes it a little forward or backward without detaching it from the driving-gear. Small discrepancies between the velocities of the type-wheel and chariot are thus corrected as often as a letter is printed. This contrivance serves to keep the receiving instrument from gaining or losing on the sending instrument during the transmission of a message. The type-wheel of the receiving instrument must be adjusted before the message begins, so as to make the two instruments start at the same letter.
Suppose a metallic cylinder, permanently connected with the earth, to be revolving, carrying with it on its surface a strip of paper freshly impregnated with cyanide of potassium. Also suppose a very light steel point permanently connected with the line-wire, and resting in contact with the paper. Every time that a current arrives by the line-wire, chemical action will take place at the point of contact, and the paper at this point will be discolored by the formation of Prussian blue. This is the principle of Bain's electro-chemical telegraph, which leaves a record in the shape of dots and dashes of Prussian blue. The apparatus for sending signals is the same as in Morse's system. The paper must not be too wet, or the record will be blurred; neither must it be too dry, for then no record will be obtained.
An autographic telegraph is one which produces at the receiving station a facsimile of the original dispatch. The best known-instruments of this class are those of Bonelli and Caselli. We shall describe the latter.
At the sending station a sheet of metallized paper, with the dispatch written upon it in a greasy kind of ink, is laid upon a cylindric surface (M, Fig. 16). At the receiving station there is a similar cylindric surface (R), on which a sheet of Bain's chemical paper is laid. Two styles, driven by pendulums which oscillate with exact synchronism, move over the surfaces of the two sheets, describing upon