282 



KNOWLEDGE. 



[December 1, 1900. 



definite direction, the transmitter used by Marconi is 

 one devised by Professor Righi, of Bologna, and shown 

 in Fig. 3. The two large spheres, e, e, are 1 1 centimetres 



A A 



Fig. 3.— Bighi Oscillator for use with Eeflector. 



in diameter, and are separated by a space of a millimetre. 

 In order to concentrate the beam of rays in the required 

 direction the oscillator is placed in the focal line of a 

 parabolic cylindrical reflector, as shown in Fig. 4. 



The most important part of the receiver is the co- 

 herer, which consists of a small glass tube (Fig. 5), 

 about two and a half millimetres in internal diameter 

 and some four centimetres in length. Two silver pole 

 pieces are lightly fitted into this tube, separated by a 

 gap of about a millimetre, containing a mixture of 96 



Fig. 4. — Marconi Transmitter witli Parabolic Reflector. 



parts of nickel and 4 parts of silver, not too 

 finely granulated, and worked up with the merest 

 trace of mercui-y. This powder must not be 

 packed too tight, or the action will be irregular 

 and over-sensitive to slight outside disturbances, 

 while if too loose it will not be sufficiently sensitive. It 



Fig. 5. — Marconi Coherer. 



is found that the best adjustment is obtained when the 

 coherer works well under the actions of the spaa'ks from 

 a small electric trembler placed at a distance of about 

 a metre. The tube is then exhausted on a mercury pump 

 until the pressure falls to about a millimetre, when the 

 tubulure left for exhausting it is sealed off. The tubes 

 are tested over a distance of 18 miles before being put 

 into use, and when all the requisite precautions are ob- 

 served, Signer Marconi finds them as reliable as any other 

 telegraphic instruments, and not liable to get out of 

 order when in use. His experience in this is confinned 

 by that of Professor Fleming. If the tubes aj-e not 



exhausted they are found to grow gradually less sensi- 

 tive, probably from slight oxidation going on, and this 

 of coiu-se would not be permissible in commercial in- 

 struments. 



The general arrangement of the receiving apparatus 

 for long distance work without a reflector is shown in 

 is the coherer tube, with its silver pole 



Fig. 6. — Marconi Receiver witli Vertical Wire and Earth Connection. 



pieces, j', j-. The coherer forms part of a circuit con- 

 taining a local cell, g, and a sensitive telegraph relay. 

 When electric waves impinge upon the coherer 

 its resistance falls from a nearly infinite value to 

 something between 500 and 100 ohms, which allows 

 the cell, g, to energise the electro-magnet of 

 the relay, n, and close a circuit containing a 

 larger battei-y, r, together with a Morse recorder, 

 h, and a trembling electric bell, p, to act as 

 decoherer. The hammer, o, of the bell is so adjusted 

 as to tap the coherer tube and shake the filings in it. 

 If at the moment in which these actions took place the 

 electric waves in the resonator had died away, this tap 

 would restore the coherer to its normal condition of 

 practically infinite resistance, and a dot only would 

 be recorded on the tape of the Morse machine. If, 

 however, the key of the transmitter were kept depressed, 

 then waves would succeed each other at very short 

 intervals, so that the acquired conductivity of the coherer 

 would only be momentarily destd'oyed by the tap of the 

 bell-hammer, and immediately re-established by the 

 electric waves. Now the armature of the Morse recorder 

 is somewhat heavy, and therefore has considerable 

 inertia, so that it cannot follow the very rapid vibrations 

 of the tongue of the relay. The practical result, there- 

 fore, is that the Morse instrument gives an exact re- 

 production of the dots and dashes produced by the 

 movements of the key at the transmitting station, 

 although during each movement of the key, however 

 short, the armatures of the relay and of the tapper go 

 through a series of rapid vibrations dependent on each 

 other. 



Small choking coils, k', k' — that is to say, coils wound 

 so as to have self-induction or electric inertia — are intro- 

 duced between the coherer and the relay, their effect 

 being to compel the greater part of the oscillatory 

 current induced in the circuit by the electric waves to 

 traverse the coherer, instead of wasting the greater 

 portion of its energy in the alternative path afforded by 

 the relay. If these coils are omitted, other circumstances 

 remaining the same. Signor Marconi finds that the dis- 



