370 M. Ampere's Letter to M. Gei'hardi on various 



short time before, his important discovery of the continuous 

 motion of rotation of a voltaic conductor round a magnet 

 and of a magnet round a conductor, and in which he states 

 that he was not enabled by the action of the latter to turn a 

 magnet on its axis, I endeavoured to produce that sort of 

 motion by causing magnets to act in every way I could ima- 

 gine upon the moveable conductor which I had hitherto made 

 use of in all my experiments, the two extremities of which 

 were placed in the axis of rotation. I soon arrived at this 

 general result, that so long as that circinnstance is allowed to 

 exist in a conductor, of which all the parts are connected in- 

 variahlij together, the continuous motion of rotation is impos- 

 sible; and I easily concluded that it was equally so, by the 

 mutual action of a magnet and a complete circuit of invari- 

 able form, since such a circuit may always be considei'ed as 

 the union of two portions of conductors, of which the extre- 

 mities are in the same axis of rotation taken at pleasure." 



And at page 356, in repeating that it is impossible to pro- 

 duce that sort of motion by employing magnets alone, or solid 

 conductors forming complete circuits, I explained in a note 

 at the bottom of the page the expression " solid conductors " 

 thus : " It is to be understood by this expression, that all the 

 parts of the portion of that conductor which forms a complete 

 or nearly complete circuit are invariably connected together, 

 and cannot alter their respective situations. When that por- 

 tion is composed of two or several moveable pieces separately, 

 or is formed entirely or partly of a liquid conductor, the mo- 

 tion of continuous rotation becomes possible." 



You perceive, sir, that the limitation which establishes the 

 correctness of what I have advanced, in the case where a ro- 

 tatory movement becomes impossible, is pointed out in the 

 most express terms in the above note contained in my Re- 

 cueil, immediately preceding my letter to Mr. Faraday, and 

 which was published more than two years ago. 



The second observation relates to the remark that you make 

 at page 16 of your paper, in consequence of your having de- 

 duced from the expression, 



'^"^g" (cos fl'— cos fl"— cos 6/ + cos 9,"), 



which I gave in page 28 o^ my Precis delaTheorie des Phcnor 

 mencs Electro-dynarniques, to represent the rotatory momentum 

 produced by the action of an electro-dynamic solenoid on a 

 ccmductor (which action may be in general compared to that 

 produced on (he same conductor by a magnet), that, supposing 

 the two extremities of the conductor, and the two poles of the 

 solenoid, or of the magnet, to be at the same time in the axis 



of 



