Although Hjorth had, as stated before, put forward the idea that a dynamo generator might itself furnish currents for magnetizing its own magnets, this valuable suggestion was not apparently worked out until 1866, when a machine was constructed for Sir Charles Wheatstone. This appears to have been the first self-exciting machine in existence. Wheatstone read a paper before the Royal Society in February, 1867, “On the Augmentation of the Power of a Magnet by the Reaction thereon of Currents Induced by the Magnet Itself.” This action later became known as the reaction principle in dynamo machines. As often happens, the idea occurred to other workers in science almost simultaneously, and Dr. Werner Siemens also read a paper in Berlin about a month earlier than that of Wheatstone, clearly describing the reaction principle. Furthermore, a patent specification had been filed in the British Patent Office by S. A. Varley, December 24, 1866, clearly showing the same principle of action, and he was, therefore, the first to put the matter on record. The time was ripe for the appearance of machines closely resembling the types now in such extended use. Gramme, in 1870, adopting a modified form of the Paccinotti ring and commutator, and employing the reaction principle, first succeeded in producing a highly efficient, compact, and durable continuous-current dynamo. The Gramme machine was immediately recognized as a great technical triumph. It was in a sense the culmination of many years of development, beginning with the early attempts immediately following Faraday’s discovery, already referred to. Gramme constructed his revolving armature of a soft iron wire ring, upon which ring a series of small coils of insulated wire were wound in successive radial planes. These coils were all connected with a continuous wire and from the junctions of the coils one with another connections were taken to a range of copper bars insulated from each other, constituting the commutator. In 1872 Von Hefner Alteneck, in Berlin, modified the ring winding of Gramme and produced the “drum winding,” which avoided the necessity for threading wire through the centre of the iron ring as in the Gramme construction. The several coils of the drum were still connected, as in Gramme’s machine, to the successive strips of the commutator. In modern dynamos and motors the armature, usually constructed of sheet-iron punchings, is a ring with projections as in Paccinotti’s machine, and the coils of wire are in most cases wound separately and then placed in the spaces between the projections, constituting in fact a form of drum winding. In the early 70’s a few Gramme ring and Siemens drum machines had been applied to the running of arc lights, one machine for each light. There were also some Gramme machines in use for electroplating.