80 Intelligence and Miscellaneous Articles. 



been led to investigate what was the effect of the nature of the iron 

 of the electromagnet upon the duration of the phases of magneti- 

 zation and demagnetization. For this purpose I have employed a 

 register in which the pieces of iron constituting the electromagnet 

 are removable, the bobbins, armature, style, &c. remaining the same, 

 so as to make evident the influence of the metal of the electromagnet. 

 To measure the duration of the phases, I have made use of the 

 method indicated in my first communciation on electric chrono- 

 graphs. 



The metallic portion of the electromagnets, which I placed 

 successively in the magnetizing helices, consisted of two cores of 2 

 millims. diameter and 13 millims. length. The coils into which the 

 current passed contained 14 metres of wire -J millim. in diameter. 

 The pile consisted of one Bunsen's element modified by M. Dulau- 

 rier. The varieties of iron tried were the ordinary iron of com- 

 merce, the soft iron used specially for telegraphs, malleable cast 

 iron, grey cast iron, and cast steel stretched and chilled. 



The results obtained were quite unexpected ; for the soft iron, 

 ordinary iron, malleable cast iron, and even the chilled steel gave very 

 nearly the same results for the duration of the phases of magneti- 

 zation and demagnetization, viz. second. 



Duration of demagnetization 0-00025 



,, magnetization (approximately) 0*00150 



The grey cast iron gave still better results ; for the duration of 

 the magnetization was reduced to about yitot5 °^ a second. It 

 would therefore be this last metal that would permit the greatest 

 possible rapidity to be obtained in the transmission of signals. 



In brief, with my present registers (such as will be described 

 soon) signals can be obtained perfectly distinct, succeeding each 

 other at intervals of -^5^- of a second, by employing no matter what 

 sort of iron for the electromagnets, and of -^-J-g- of a second when 

 the latter are of grey cast iron. It is essential to remark that I am 

 not speaking here of signals succeeding each other regidarly at inter- 

 vals of 3-J-3- or 3-^5- of a second, which would constitute a system. 

 In this latter case, indeed, the number of signals which can be 

 transmitted exceeds by far 350 or 500 per second. 



I am disposed to believe that the superiority of cast iron depends 

 on its molecular texture, and not on the quantity of carbon which 

 it contains. I intend also to try soft iron cast and not forged, which, 

 I think, will exceed in rapidity all that I have hitherto obtained. 

 I purpose, moreover, in another communication, shortly to revert to 

 the details of my experiments, and to the application of my regis- 

 ters to electric chronographs intended for artillery. 



The durations above indicated, it must be observed, do not in- 

 clude the time employed by the style in traversing its trajectory ; 

 it is by adding this to the durations of magnetization and demag- 

 netization that 3-J^- or -g-J-jj of a second, according to the case, is 

 found for the total duration of a signal, comprising the demagneti- 

 zation, the time of fall of the style, the magnetization, and the re- 

 turn of the style to its initial position. Besides, these are the 

 numbers when only one pile-element is employed ; the number of 

 signals transmitted per second increases with the intensity of the 

 current. — Comptes Bendus de VAcad. des Sciences, vol. lxxx. p. 1353 



