When the trouble first appeared in Boston, in 1891, its cause was promptly identified. The electric-railway construction of those days was so crude, however, that many well-informed electricians fell into the error of assuming that heavier rails, more and larger return feeders, and better bonding (i. e., wire connections from rail to rail, around the joints, designed to decrease the resistance) would prove a panacea for all electrolytic ills. Indeed, this view is still held by a surprisingly large number of men versed in matters electrical. I am of the opinion that it is impossible, from a financial standpoint, to provide so satisfactory a legitimate return that considerable electricity will not seek a path through pipes, cable covers, etc.; for, in order to confine the electric current to the rails, the resistance of the earth and its contained pipes would have to be infinitely great, and this condition can be realized only by making the resistance of the rail infinitely small as compared with that of the earth. The cost of arriving at this condition is prohibitive, and the improved track return is, and always must be, a palliative merely, not a cure. [Illustration: LEAD SERVICE PIPE SHOWING THE EFFECTS OF EIGHT MONTHS’ ELECTROLYTIC ACTION, AND CLEARLY ILLUSTRATING THE FACT THAT DAMAGE OCCURS ONLY WHERE THE ELECTRICITY LEAVES THE CONDUCTOR. The interior surface is unattacked.] Assuming, then, that under the most favorable character of electric-railway construction some of the current may be expected to stray from the straight and narrow path, it behooves us to consider how it may best be cared for in order that it may not cause electrolysis. Since corrosion of this nature occurs only at those points where electricity _leaves_ the metal, one might suppose that the attachment of a conducting wire to the affected part would result in the harmless carrying away of the current. In isolated cases, in small towns, such a plan might accomplish the desired result. It is open to the objection, however, that it in a measure legalizes the conveyance of electricity on conductors other than those designed for the purpose. In larger towns, with more than one power house and with car lines radiating from and circumscribing the business center, the attachment of conducting wires entails a ceaseless disturbance of the electrical equilibrium, curing the evil in spots and developing new danger points. Furthermore, these connections tend to decrease the resistance of the total illegitimate return, thereby tempting a greater flow of electricity along other paths than the rails and track feeders. It has been generally believed that this increased current would develop electrolysis at the ends of the pipes, due to the jumping of the electricity around the presumably high resistance of the joints; and, indeed, many samples of such corrosion are in existence. I have found, however, that it is possible to calk a bell-and-spigot joint