114 



NATURE 



[March 23, 191 1 



ing children quite as vigorous in pliysiquc AWfii much 

 better endowed mentally than the average native of 

 Africa. 



An interesting allusion is made in this book to the 

 origin of fire, interesting because the native tradition 

 quoted by Mr. Dennett is in accord with the observa- 

 tions and theories of several African explorers. On 

 p. 2\b he quotes the Voruba legend liuit before man 

 knew how to make fire, bush fires used nevertheless 

 to wcur almost yearly at the end of the dry season 



\\ tun natural combustion took place." In Africa, 

 .11 .my rate, this was how fire became an agent of 

 man. ' I have seen myself lightning set fire to a dead 

 tree or to the dry grass near a tree, and thus start a 

 l)ush fire in the dry season. Hush fires are very 

 tktrimental, in reality, to civilised agriculture in 

 Africa. Nevertheless, in regions where the natives 

 appreciate this fact and do not set fire to the grass 

 or brushwood, bush fires occur from time to time in 

 the dry season, and the natives assert that they are 

 due to some natural cause, more especially to light- 

 ning, but also, it has been suggested, to some action 

 of the sun, possibly of "a burning-glass" character 

 acting through silica or some other mineral substance 

 which concentrated the rays on to tinder. But light- 

 ning frequently starts a fire in Africa — as witness the 

 cathedrals, .barracks, hospitals, &c., which are burnt 

 u> the ground from this cause. The spread of the 

 bush fire 'proved to be of enormous benefit to early 

 man, since when he followed behind its ravages he 

 was presented by nature with a variety of cooked or 

 lialf-cooked beasts, birds, and reptiles. In this way 

 he learnt the charms of cooked food and the useful- 

 ness of fire, and no doubt began to count on the 

 annual opportunities offered to him in the dry season 

 for the renewal of his household fire before he learnt 

 to produce a flame artificially. 



The chapter dealing with totems (p. 175 cl seq.) 

 is particularly interesting, and useful information is 

 given on the laws and customs of land tenure. 



H. H. Johnston. 



HIGH-FREQUENCY GENERATOR FOR 

 WIRELESS TELEGRAPHY. 



11 is announced in thr daily Press that Dr. Gold- 

 schmidt has recently succeeded in sending wire- 

 less messages from Berlin to the South-Western 

 frontier of Germany using his new high-frequency 

 alternator to generate the electric oscillations. The 

 production of undamped waves by means of high- 

 frequency alternators has been the aim of numbers of 

 inventors, as it is hoped that by producing a suitable 

 generator it may be possible to avoid the defects of 

 working that are associated with the usual arc and 

 ^|)aik methods. 



A number of alternators have been built, but owing 

 to various reasons none of them have as yet come into 

 txieiided use. Their design on the usual lines is very 

 difficult, as even if the rotating parts are made to 

 revolve at the highest speeds permissible from 

 mechanical considerations, the number of poles re- 

 quired to produce the high frequencies necessary for 

 wireless telegraphy is so great as to leave very little 

 room for the windings, and the consequent cramping 

 of the windings and great leakage between the closely 

 spaced poles give rise to considerable drop of voltage 

 when load is put on the machine. 



Most of the machines that have been constructed 

 hitherto have been of the inductor tj'pe, consisting of 

 fixed windings placed under the influence of rapidly 

 rotating armatures of iron containing a large number 

 of projections or teeth, but the Goldschmidt machine 

 is built on a quite different principle. 



It is a phenomenon well known to those who have 

 NO. 2160, VOL. 861 



to work with single-phase alternators that when '■ 

 is put on such machines the armature reaction c.i 

 double-frequency currents to flow in the field wind 

 and that these double-frequency currents cause tr 

 frequency currents to flow in the armature 

 and so on. This may be explained by the c 

 tion that a stationary alternating flux can be 

 as being composed of two equal and constxn 

 rotating with equal speeds in ojiposite direct: 

 speed of the two fluxes being such that one ■ 

 revolution is made by them in the time of one peruMi 

 of the alternating flux. 



Applying this to the case of an alternator v.-'-'- 

 stationary alternating-current winding and rotxr 

 field system, it will be seen that if the rotating ;,. .>, 

 produces alternating currents of a frequency / in the 

 stator, the component rotating fields produced by ''"• 

 stator currents will rotate at the same speed as 

 field system, one in the same direction as the !, . 

 winding, and therefore having no inductive effect 00 

 it, and the other in the opposite direction, and there- 

 fore inducing a current of a frequency 2/ in it. A con- 

 tinuation of this process would cause currents of frf^- 

 quencies 2/, 4/, 6/, &c., to appear in the field wind' 

 and currents of frequencies 3/, 5/, 7/, &c., to ap; 

 in the stator w-inding. The production of the hii 

 frequency currents in ordinary alternators is lin 

 by the fact that the amplitudes of the series of ■ 

 monies decrease rapidly owing to the great imped, 

 opposed to their flow in the windings, but the tri 

 frequency harmonic superposed on the fundament.; 

 often sufficiently marked to cause undesirable disuT- 

 tion of the wave-shape of the electromotive force of 

 the machine. 



Dr. Goldschmidt has constructed a machine in 

 which the effect referred to is utilised to produce c"-- 

 rents of very high frequencies, although the fu! 

 mental frequency of the machine is comparati 

 low. In order to prevent the damping out of 

 higher harmonics, he connects in parallel with 

 stator w-indings a series of capacity-inductance shi 

 tuned to resonance with the odd multiples of 

 fundamental frequency of the machine, and in parallel j 

 with the field winding a series of such shunts tuned to j 

 resonance with the even multiples of the fundam* ; 

 frequency. Owing to the presence of these shi. 

 the high-frequency currents are able to attain cun- 

 siderable magnitudes, and the electrical energy is re- 

 flected backwards and forwards between the stator 

 and rotor of the machine a great number of times, 

 the frequency of the oscillations being increased at 

 each reflection until a frequency corresponding to the 

 free period of the radiating circuit is reached. 



In Dr. Goldschmidt's machine, oscillations of 

 120,000 cycles a second are produced, and the rated 

 output is 12 kilowatts. 



Practical experience of working will be necessary 

 before it is possible to say to what extent machines 

 of this type are likely to replace the present oscillation 

 generators, but it seems not unlikely that diffi- 

 culties will arise in keeping the frequency of the 

 oscillations constant enough to enable clear signals to 

 be received. It would appear that any small varia- 

 tion of the speed of the generator would cause each 

 successive harmonic to depart to an increasing extent 

 from its proper frequency, so that the final frequency 

 reached might be so far removed from its proper value 

 that signals would no longer be able to be received; 

 added to this the amplitudes of the harmonics might 

 be expected to be greatly diminished by the fact that a 

 change of the speed of the machine would put all the 

 capacity-inductance shunts out of tune with the har- 

 monics for which they are adjusted, and a consider- 

 able weakening of the signals emitted by the aerial 

 would result. A. J. Makower. 



