electrical oscillations to be set up in an aerial on a scale never before accomplished. This work involved, not merely the ordinary experience of an electrical engineer, but also the careful consideration of many new problems and the construction of devices not before used. Every step had to be made secure by laboratory experiments before the responsibility could be incurred of advising on the nature of the machinery and appliances to be ordered. Many months in the year 1901 were thus occupied by the author in making small-scale experiments in London and in superintendence of large-scale experiments at the site of the first power station at Poldhu, near Mullion, in Cornwall, before the plant was erected and any attempt was made by Mr. Marconi to commence actual telegraphic experiments. As this work was of a highly confidential nature, it is obviously impossible to enter into the details of the arrangements, either as made by the writer in the first instance, or as they have been subsequently modified by Mr. Marconi. The design of the aerial and of the oscillation transformers and many of the details in the working appliances are entirely due to Mr. Marconi, but as a final result, a power plant was erected for the production of Hertzian waves on a scale never before attempted. The utilisation of 50 H.P. or 100 H.P. for electric wave production has involved dealing with many difficult problems in electrical engineering, not so much in novelty of general arrangement as in details. It will easily be understood that Leyden jars, spark balls and oscillators, which are quite suitable for use with an induction coil, would be destroyed immediately if employed with a large alternating-current plant and immensely powerful transformers. [Illustration: FIG. 26.--WOODEN TOWERS SUPPORTING THE MARCONI AERIAL AT POLDHU POWER STATION, CORNWALL, ENGLAND.] In the initial experiments with this machinery and in its first working there was very considerable risk, owing to its novel and dangerous nature; but throughout the whole of the work from the very beginning, no accident of any kind has taken place, so great have been the precautions taken. The only thing in the nature of a mishap was the collapse of a ring of tall masts, erected in the first place to sustain the aerial wires, but which now have been replaced by four substantial timber towers, 215 feet in height, placed at the corners of a square, 200 feet in length. These four towers sustain a conical arrangement of insulated wires (see Fig. 26) which can be used in sections and which constitute the transmitting radiator or receiver, as the case may be. Each of these wires is 200 feet in length and formed of bare stranded wire.