Abstract :
During the summer of 1904 a series of tests was carried on for an operating company for the purpose of determining the most satisfactory method of metering the power delivered to certain consumers whose demands ranged from 200 to 1000 kilowatts. The aim was to determine, as far as possible, the nature and probable extent of the errors that occur in the established available methods of polyphase metering, and to find which of these methods would be on the whole the best to use for the given work — regarding as best that system in which errors are not only least liable to occur, but also in which, if they do occur, their presence can be most quickly and readily detected.
