Page:Popular Science Monthly Volume 24.djvu/267

Rh mind perceived that in the fact he named, which to so many of us might seem little more than a curious experiment, lay the principle which, if rightly developed, would make practicable the transmission of power.

If, now, we could call back this great electrical engineer, and ask him what recent discovery came next in importance to this, what would he reply? His answer would be the discovery that "a voltaic battery is reversible." The Gramme machine has given us means of transmitting power of electricity. The later discovery enables us to store up electrical energy as distinguished from electricity.

Electrical engineering, which embraces a knowledge of cables, telegraphy, electric lighting, electrical measurement, transmission of power, storage-batteries, and how to localize faults in cables, land lines, and telephone lines, has thus become a subject of the first practical importance.

A prominent department of the electrical engineer's work is the localizing of faults in ocean-cables, which may be of five different kinds: 1. Where the copper conductor makes a "perfect earth." 2. Where the copper conductor is broken, and yet the insulation remains unbroken. 3. Where an "imperfect earth" is made. 4. Faults arising from a hole in the gutta-percha sheath, making a connection between the conductor and the sea. 5. From the establishment of a connection between the iron sheathing and the copper core, by a nail or wire driven in.

The first kind of fault is easily located, because we know the resistance of the cable when it is in perfect working order. If, for instance, it has 10,000 ohms, Or units of resistance, a fault making a perfect earth midway in the cable would give us 5,000 ohms resistance. Or, we know how many ohms of resistance there are to a mile of cable when it is in perfect working order, and, by the use of delicate instruments and by mathematical calculations, we can easily locate the fault.

The location of the second class of faults, i. e., a complete breakage of the conductor, naturally followed by a total cessation of all communications between the two ends of the cable, may be detected in several ways. The charge which the cable will contain is first measured; and, when the charge per mile is known, the amount actually observed will directly give the location of the faults; and the exactness with which the position of the break can be determined is limited only by the accuracy with which the relative charges can be compared. Suppose, for instance, the discharge from a mile of the cable with a given battery, and reflecting galvanometer, is represented by a deflection of ten divisions, and the discharge from a cable containing a broken copper conductor is one hundred divisions, we know the fault is about ten miles from the shore.

A fault of the fourth kind is located very readily. There is a