hertz's EXl'EHIMENTS. 221 



language of tiilkingof oloctiic cliargcs on Ixxlio.s, and clcctdc currents 

 from one to the other, of ekM'trie eharges iieutraHzing one another, and 

 so forth, is not in accorchmce witli the most recent developments of 

 electro-magnetic theor.\. At the same time, those for whom these 

 articles are written are familiar with this language and with the view 

 of the subject that it is framed to suit, while they are unfamiliar with 

 a'ther electrically and magnetically strained and thereby the seat of 

 electric and magnetic energy, and conse<]uently it would have added 

 very nuich to their difticulty in grasi)ing the details of a complicated 

 (juestiou if it had been devSi^-ribed in unfamiliar terms and from an un- 

 familiar point of view. 



The electric force in the neighborhood of the vertical generator will 

 lie in vertical planes through it, and as A and B are alternately positive 

 and negative, the electric force will alternately be from above down- 

 Avards, and from below upwards. If then this foice is ])ropagated out- 

 wards in a series of waves, we may expect that all round our generator 

 waves of electric Ibrce will be diverging; \^aves in which the force will 

 be alternately down and uj). The state of atfairs might be roughly 

 illustrated l)y elastic strings stretched out in every direction from our 

 generator. If their ends at the generator be moved alternately down 

 and up, waves will be pro])agated along the strings, waves of alternate 

 motion down and up. 



lu order to reflect these waves we require a metallic sheet of consid- 

 erable area some two or three wavelengths away from the generator; 

 so far away in order that we may Imve room for our detector to find 

 the loops and nodes formed every half wave-length where the outgoing- 

 waves meet those reflected from the screen; not too far away or (mr 

 waves will be too feeble even at the loops to affect our detector. The 

 waves are thrown off all round, but are most intense in the horizontal 

 I)laue through the si)ark, so that our <letector had better be placed as 

 near to this plane as possible. TIh' detector may be either a very 

 nearly closed circle of wire or two condu<'tors, each somewhat longer 

 and thinner than the combined lengths of the generating conductors, 

 and placed vertically over one another, and separated by a minute air 

 gap. As the theory of this latter form of detector is simi)ler than that 

 of the circle, it will simplify matteis to consider it alone. The two con- 

 ductors should each have a period of electrical oscillation u]) and down 

 it, the same as that of the charges on the genei-ator. The generator 

 consists of two conductors certainly, but then during the time the S[)ark. 

 lasts they are virtually one conductor, being c(mnected by the spark 

 across which the electric charges are rushing altei'nately up and <lown. 

 Hence the period of oscillation of the charges on the generator corre- 

 s])onds to that on a single conductor of the same size as its two ]>arts 

 combined. 



Various experiments have been made as to the best form for these 

 conductors that form the detector They might be made identical 



