Now, in duplex telegraphy the artificial current must be equal to the chief current in intensity all the time; so that, since in submarine telegraphy the current rises gradually to its full strength and as gradually subsides, the artificial current must do the same. Reverting to the illustration derived from the flow of water, if we had a small pipe the rapid flow through which was to carry as much water one way as the slow flow through a large pipe was to carry water the other way, then if the large pipe had a widening along one part of its long course the short pipe would require to have a similar widening along the corresponding part of its short course. And to make the illustration perfect, the widenings along the large pipe should be unequal in different parts of the pipe’s length; for the capacity of a submarine cable, regarded as a condenser, is different along different parts of its length. What is wanted, then, for a satisfactory system of duplex telegraphy in the case of submarine cables, is an artificial circuit which shall not only correspond as a whole to the long circuit, but shall reproduce at the corresponding parts of its own length all the varieties of capacity existing along various parts of the length of the submarine cable. Several attempts have been made by electricians to accomplish this result. Let it be noticed that two points have to be considered: the intensity of the current’s action, which depends on the resistance it has to overcome in traversing the circuit; and the velocity of transmission, depending on the capacity of various parts of the circuit to condense or collect electricity. Varley, Stearn, and others have endeavoured by various combinations of condensers with resistance coils to meet these two requisites. But the action of artificial circuits thus arranged was not sufficiently uniform. Recently Mr. J. Muirhead, jun., has met the difficulty in the following way (I follow partially the account given in the _Times_ of February 3, 1877, which the reader will now have no difficulty in understanding):—He has formed his second circuit by sheets of paper prepared with paraffin, and having upon one side a strip of tinfoil, wound to and fro to represent resistance. Through this the artificial current is conducted. On the other side is a sheet of tinfoil to represent the sheathing,[29] and to correspond to the capacity of the wire. Each sheet of paper thus prepared may be made to represent precisely a given length of cable, having enough tinfoil on one side to furnish the resistance, and on the other to furnish the capacity. A sufficient number of such sheets would exactly represent the cable, and thus the artificial or non-signalling part of the current would be precisely equivalent to the signalling part, so far as its action on the needle at the transmitting station was concerned. “The new plan was first tried on a working scale,” says the _Times_, “on the line between Marseilles and Bona; but it has