"(All those who have been concerned with the practical application of electrical telegraphy readily agree on this point, namely, that the immense majority of the disturbances which affect electrical telegraphs arise from variations in the intensity of the currents employed. The cause of the variations lies either in the source of the currents or in the varying conditions of the conductor circuit. The first of these causes can be eliminated by making use of constant sources. I am content, in this regard, to observe that I give preference to the Daniell cell.)"