The high cost of very high-power vacuum tubes together with their low efficiency and limited durability is at present, however, likely to militate against their general use, but im provements in methods of construction of this apparatus will doubtless reduce its cost and in crease its durability. The existing need of high and very expensive aerial wires in long distance wireless telegraphy and telephony is also a limiting factor to the extensive employment of those arts to very long distances.
Some of the chief obstacles to the exten sive and successful use of wireless telegraphy and telephony in the past have been those due to statics and interference between or clashing of signals from different sending stations at nearby or even at remote stations. A great forward step toward the solution of the statics problem is that due to the ),Veagent anti-static devices. (See TELEGRAPHY, WIRELESS). The interference problem is still to be solved.
As already noted, by the use of selective tuning devices which only transmit and re spond to waves or oscillations of a certain frequency, stations may cut out waves of other frequencies and thus avoid interference. Thus by the use of a standard wave length for cer tain circuits, as, following land telegraph prac tice, the ether route used by different wireless stations may be termed non-interfering; ether circuits may be and in fact are established very generally at the present time. Thus the regular wave lengths for a given trans-Atlantic circuit may be set at, say, 10,000 meters. Other stations using that wave length in the vicinity of that circuit to which the 10,000-meter wave length is assigned, would set up interference or clashing of signals. It is known, however, that a difference of 300 to 400 meters in the wave length is ample to avoid interference between such stations. For instance, a circuit employing a 10,000-meter wave length could operate side by side with a circuit employing a 10,400-meter wave length. For long-distance wireless telegraph and telephone operation, or, in other words, for high-power station work, long wave lengths are essential. The present minimum and maximum available wave lengths are, say, from 10,000 to 20,000 meters for long distance signaling. Since the signals from the very high-power stations are now received all over the world, it is obvious that if a difference of, say 400 meters, is required to prevent in terference between any two such wireless cir cuits, the total number of available world wide circuits would be limited to say 25.
It may be noted that for short-distance wire less signaling a difference in the respective wave lengths of about 20 meters is found sufficient to prevent interferences between cir cuits. Even in short distance signaling, how
ever, were all stations to adhere rigidly to an allotted wave length in any designated zone the number of available wave lengths would ultimately be reached. While the foregoing conditions continue the limitations to wireless telegraphy and telephony will exist.
It is, however, not unlikely that improve ments in this branch of electrical signaling will render it possible to operate transoceanic wireless telegraphy and telephony on a much closer margin than 400 meters. Indeed it has been noted by Alexanderson in the paper re ferred to that by improved arrangements of apparatus and circuits it will be possible to operate such circuits with a difference of but 1 per cent of the total frequency, instead of 4 per cent as at present. This would make it possible to increase the number of trans oceanic circuits to, say, 100, roughly speaking. Again, by means of improved methods of di rective signaling the number of parallel wire less circuits on one wave length will be in creased by five. These improvements together with an increase in the speed of signaling to 100 words per minute generally, instead of 20 words as at present, will, it is estimated, increase the total capacity of the radio trans oceanic traffic of the world 1/5 times.
In wire telephony interference of this nature does not occur as measures are taken to prevent such interference (teamed inductive interference between circuits) by using two parallel wires suitably transposed for each telephone circuit. (See TELEPHONE). The ef fect of induction between parallel Morse and printing telegraph circuits is not pronounced inasmuch as the receiving instruments em ployed on those systems are comparatively in sensitive to weak induction currents from other lines.
Notwithstanding the real difficulties in the way of what may be termed universal wireless telephony for social, commercial and other pur poses it is confidently expected by competent telephone engineers and inventors that its ulti mate consummation is not beyond the possibil ity of practical realization. In the meantime wireless telephony if only available, com mercially, for comparatively short distances, obviously could be installed to great advantage in the officer's room of every ship that floats ocean, lake, river or harbor, because of the fact that telephony requires no specially trained operator.
Bucher,