Finally, in the Paper referred to, he emphasises the well-known fact that long-distance signalling can only be accomplished by the aid of an aerial wire and a "good earth." Summing up his results, he concludes: (1) That intervening land of any kind reduces the practical signalling distance between two ships or stations, compared with that which would be obtainable over the open sea, and that this loss in distance varies with the height, thickness, contour, and nature of the land; (2) material particles, such as dust and salt, held in suspension in a moist atmosphere also reduce the signalling distance, probably by dissipating and absorbing the waves; (3) that electrical disturbances in the atmosphere also act most adversely in addition to affecting the receiving instrument and making false signals or _strays_, as they are called; (4) that with certain forms of transmitting arrangement, interference effects may take place which have the result of creating certain areas of silence very similar to those which are observed in connection with sound signals from a siren. It is clear, therefore, from all the above observations, that Hertzian-wave telegraphy taking place through the terrestrial atmosphere is not by any means equivalent to the propagation of a wave in free or empty space; and that just as the atmosphere varies in its opacity to rays of light, sometimes being clear and sometimes clouded, so it varies from time to time in transparency to Hertzian waves, the cause of this variation in transparency probably being the presence in the atmosphere of negatively-charged corpuscles or electrons. If there are present in the atmosphere at certain times "clouds of electrons" or "electronic fogs," these may have the effect of producing a certain opacity, or rather diminution in transparency to Hertzian waves, just as water particles do in the case of sunlight.