438 



NATURE 



[March 30, 1876 



that he is not able exactly to assign the true cause of certain facts 

 relative to electric induction, in the phenomenon in question. 

 We shall see in what follows, that nothing remains obscure in 

 the theory developed by Mellon!, when we reflect on the exist- 

 ence of what is called curvilinear induction, which was dis- 

 covered by Faraday, and which is now acknowledged as a reality 

 by all those who have kept themselves abreast of electrostatic 

 science. 



All that we have indicated in order to render clear the explana- 

 tion given by Melloni of the electrostatic fact of which we have 

 spoken, will be proved with a much greater amount of evidence 

 by the experiments which follow. But, before describing them, 

 it will be useful to give a brief historic sketch of the researches 

 of different physicists as to the effect of electrical induction. 

 The greater part of these researches are favourable to the theory 

 of Melloni, while some are opposed to it. This theory will be 

 acknowledged as true when it is proved that the induced electricity 

 of the first kind does not possess any tension, or, what comes to 

 the same thing, when it is proved that the homonym of the in- 

 ductor exists also in the extremity B of the induced and insulated 

 cylinder. 



Historical. — Electrostatic induction, or electrical influence, 

 was first observed by Canton in 1753 (Phil. Trans, vol. 48, part i., 

 p. 350). Franklin continued his researches, but Wilke and 

 .(^ipinus gave a greater development to the discovery indi- 

 cated.^ We conclude from the works of Canton that this physi- 

 cist knew also the induction now called curvilinear ; for he 

 indicates several circumstances depending on the same pheno- 

 menon. The fiist who recognised, more than 100 years ago, that 

 induced electricity of the first kind, that is with the sign opposite 

 of the inductionj does not possess tension, was /Epinus.* Subse- 

 quently Lichtenberg clearly announced that induced electricity 

 of the first kind has no tension.^ De Luc was also of the same 

 opinion,* as also the celebrated Volta.^ This Italian physicist " 

 admitted the want of tension in tlie induced electricity opposite 

 to that of the inductor, and admitted moreover that the electrical 

 influence is exercised lay means of a ;5rt/'//<7/ dissimulation of the 

 inducing electricity, and the entire dissimulation of the induced 

 electricity with opposite sign, a fact which is always verified. 

 It seems that the question whether or not induced electricity of 

 the fir.st kind can have tension, was discussed for the fust time 

 by Lord Mahon and Volta, about 1787, to judge from what De 

 Luc says." 



Among no! able physiciss who afterwards admitted that in- 

 duced electricity of the first kind has no tension — before that 

 doctrine was reproduced in a more developed form by Melloni, 

 July 25, 1854 — we must also reckon Fischer. This will be seen 

 in reading Fischer's " Mechanical Physics," translated by Biot 

 (4th ed., Paris, 1830, p. 238-242). The physicist Pfaff admitted 

 completely the want of tension in induced electricity of the first 

 kind.* 



The celebrated Ohm, in a paper "On an unrecognised 

 property of latent electricity," criticises Pfaff, and concludes that 

 it is not true that induced electricity of the first kind has no 

 tension.^ Consequently if Ohm had known of the existence of 

 the influence named curvilinear, he would, by means of his 

 experiments, have arrived at the contrary conclusion. The cur- 

 vilinear influence discovered by Faraday, was unknown also to 

 Melloni, but, however, he did not fail to recognise the truth that 

 the induced electricity opposite to that of the induction does not 

 possess tension. 



C. F. Mohr, a Coblentz pharmacist, criticises Pfaff, stating 

 that in his experiments, the induced cylinder received the elec- 

 tricity by communication. ^^ But we know that by operating on 

 a very dry day, such communication does not take place ; and 

 yet by experimenting well, the result obtained by Pfaff is ob- 

 tained, which triumphantly refutes Mohr.^'^ 



» Fischer's " History of the Arts and Sciences," Gottingen, 1804, vol. v., 

 p. 726. 



" "Tentcimen theoria electricitatis et magnetismi," Petersburg, 1739 

 § 43, No 2 



3 See Erltben's work " Elements of Physics," sixth edition." Gottingen, 

 1794., p. 519. 



•♦"Ideas on Meteorology," vol. i., second part, p. 334, § 360-1 (Pans, 



X787)- 



5 See his collected works, vol. i., parti. Florence, 18 16, p. 258, line 4. 



fi Ibid, p. 200, hne 6 from bottom, p. 260, line 14, and pp. 222-277. 



7 " ideas of Meteorology," vol. i., part i., p. 292, § 324-5. 



** Zehler's " Physikalisches WCrterbuch," vol. iii., p. 311 (Leipsic, 1827, 



P- i)- 



9 " Neues Jahrbuch der Chemie und Physik," by Schweigger Seidel, 



vol. v., p. 129 (1832). 



10 " Pogg. Ann. der Phys. u. Ch.", vol. xxxvi., pp. 224-8 (1835). 

 »i " Op. cit. Val. 44, p. 332, and p. 334, line i (1838). 



M, Riess gives a general risumS of the question in the *' Re. 

 pertorium der Physik" (vol. ii., p. 29; Berlin, 1838). He 

 believes that by adopting the vertical position of the induced 

 cylinder, instead of the horizontal position commonly adopted, 

 we may be convinced that the pith balls, or even the gold leaves, 

 diverge by the tension of the induced electricity — opposite to that 

 of the inductor— which they possess. But this is not altogether 

 true, since the chief cause of this divergence consists in curvi- 

 linear induction, which is not impeded in the vertical position 

 of the inducted cylinder. Moreover, we cannot at all under- 

 stand the choice of an induced cylinder placed vertically, to 

 which, without any good reason, M. Riess has given the pre- 

 ference for the purpose of proving the phenomena of electrical 

 induction, since the^e phenomena are always the same, and are 

 equally well explained in an induced cylinder, whether it be ver- 

 tical or horizontal. M. Riess, in his memoir " On the power of 

 propagation of induced electricity,"^ produces some observa- 

 tions against the memoir which Pfaff published in reply to that 

 of Mohr. We shall see that the same observations are evidently 

 overturned by my experiments, which I shall shortly describe. 



In M. Riess's work " Die Lehre von der reibungse Electri- 

 citat" (Berlin, 1853, pp. 177-207), there is a very elaborate 

 theory of electrical induction entirely opposed to that announced 

 by Melloni, and agreeing with the old and commonly adopted 

 theory ; but the arguments and experiments of Riess are reduced 

 to nothing by the arguments and experiments which I shall after- 

 wards describe. 



Two memoirs were published by Knochenhauer in Poggen- 

 dorf s Annalen, in the first of which (vol. 47, p. 455, 1839) the 

 author treats explicitly of induced electricity, and denies that it 

 possesses any tension, at the same time also asserting that the 

 electrical influence cannot traverse the conductors ; all this 

 agrees with our point of view. In the second memoir (Pogg. 

 Ann., vol. 51, p. 125, 1840) he treats of the power of induction 

 of the Coibents, an argument which has a cIo->e connection with 

 electrical induction. We ought to observe here that Fischer, 

 long before Pfaff and Knochenhauer, asserted that induced elec- 

 tricity had no tension. It is really extraordinary that neither 

 Fechner nor Riess ever sought to examine the physics of Fischer 

 in connection with the subject of electrical induction. 



Knochenhauer, in a memoir in Pogg. Ann., 1843, vol. 58, 

 p. 31), replies to Fechner, maintaining against him that induced 

 electricity of the first kind must be regarded as entirely latent. 



The physicist Petrina, in a memoir the object of which is to 

 prove the erroneousness of the hypothesis that the electric in- 

 fluence can traverse a conductor,^ shows himself favourable to 

 the absence of tension for the induced electricity opposite to that 

 of the inductor, and concludes that Fechner had by no means 

 refuted the experiments of Knochenhauer which admit this 

 absence of tension. 



According to the inferences to be drawn from the memoir of 

 Petrina above referred to, it appears that this physicist was one 

 of the first to recognise, in 1844, curvilinear electrical induction, 

 already demonstrated by Faraday in 1839. This phenomenon, 

 and that of the inability of the electric influence to traverse the 

 conductors, are both closely connected with and comprised in the 

 fundamental phenomenon of electrical induction.^ Faraday's 

 researches referred to below are arranged in a series of thirty. 

 In the eleventh of this series he speaks of his experimental 

 researches on curvilinear induction,^ and expresses himself as 

 follows:— "I believe that of all the consequences which flowr 

 from the hypothesis of induction from molecule to molecule 

 curvilinear action is the most important of all. As the existence 

 of such an action has been established with certainty, I do not 

 see how the old theory of rectilinear action at a distance can be 

 maintained, or how anyone can oppose induction from molecule 

 to molecule." It is really astonishing that in no modern treatise 

 on Physics or on Electricity do we find any mention of cur- 

 vilinear induction, which may easily be tested by repeating the 

 experiments of Faraday, as also the other experiments which I 

 have published.' We must, however, except the treatise of De 

 la Rive and that of M. Gavarret ; in the latter there is a para- 

 graph entitled, " Induction through dielectrics can be ex«rted in 

 a curved line." 



{To be continued.) 



' " Pogg. Ann," vol. 44, p. 624. 



^ "Pogg. Ann.," 1844, v. 61. p. 116. . . , ^ 



3 See Faraday's "Experimental Researches— Electricity," and " Cata- 

 logue of Scientific Papers," vol. ii. (Lond., 1868). 



4 See also "Pogg. Ann.," 1839, vol.46, p. 537.— De la Rive, " Traite 

 d'Electricite " (Paris, 1854), t. i, p. 138-9- 



5 " Comptes Rendus," 1856, t. 43, p. 719. 





