This paper has for its main objects:—(a) The survey of the various forms of power at high tension which have been used in connection with continuous-wave wireless telegraphy in the early stages of the development of that system ;(b) An examination of present practice in experimental working in connection with transmitting sets of small power; and(c) A statement as to probabilities regarding the future.The essential requirements for producing “pure continuous waves” are set out and a description is given of successful methods, and of others.The conclusion reached is that direct (i.e. unidirectional) current directly generated is inherently the best form of power, and that British engineers have now quite sufficient experience and data to design and construct generating machinery for large outputs at the high, tension required for world communication.