iron with bad joints, which gave the circuit a high resistance) that these repeaters would be inserted at Toledo, Cleveland, Buffalo and Albany in order to work from Chicago to New York. Under such conditions the transmission would necessarily be slow, because an armature time will be lost at each repeater. Regarding each repeater as a key, when Chicago depresses his key the armature of the next repeater must act, and then the next successively, and all of this takes time, although only a small fraction of a second. The repeater was a very delicate instrument and had to be handled by a skilled operator. Every wire must be in its place or the instrument would fail to operate. I remember on one occasion in Cleveland that along in the middle of the night the repeater failed to work. The operator knew nothing of the principle of its operation, so that when it failed he had to appeal to some of his superiors. At this time there was no one in the office who knew how to adjust it, so they had to send up to the house of the superintendent and arouse him from his sleep and bring him down to the office. He looked under the table and found that one of the wires had loosened from its binding-post and was hanging down. He said immediately, "Here's the trouble; I should think you could have seen it yourself." The operator replied, "I did see that, but I didn't think one wire would make any difference." He learned the lesson that all electricians have had to learn--that even one wire makes all the difference in the world. But this operator was no worse in that respect than some of his superiors. One of the heads of the Cleveland office at one time in the early days wanted to give some directions to the office at Buffalo. He told the operator at the key to tell Buffalo so and so, when the operator replied: "I can't do it; Buffalo has his key open." The official immediately said with severity: "Tell him to close it." He forgot that it would be as difficult for him to tell him to close it, as it would have been to have sent the original message. But let us go back to the telephone. While it is possible to send a message from New York to San Francisco by telegraph, it is not possible to telephone that distance, because as yet no one has been able to devise a repeater that will transfer spoken words from one line to another satisfactorily. But unless the printer and publisher bestir themselves some one may accomplish the feat before this little book reaches the reader. If this proves to be true, let the writer be the first to congratulate the successful inventor. CHAPTER XVII. SUBMARINE CABLES. The first attempts at transmitting messages through wires laid in water were made about 1839. These early experiments were not very successful, because the art of wire-insulation had not attained any degree of perfection at that time. It was not until gutta-percha began to be used as an insulator for submarine lines that any substantial progress was made.