Intelligence and Miscellaneous Articles, 317 



exploding distance, the nature of the electrodes, and the resistance 

 of the circuit traversed by the electricity are varied. 



Two methods have hitherto been employed to render appreciable 

 the duration of a spark. One, devised by Wheatstone, consists in 

 causing the image of the spark to be reflected upon a mirror move- 

 able about an axis parallel to the length of the spark. Although the 

 spark of an ordinary electrical machine exhibits no duration appre- 

 ciable by this means, yet, with a certain velocity of rotation, battery- 

 discharges give images elongated in the direction of the rotation of 

 the mirror, a proof of a sensible duration, but with diminution of 

 luminous intensity. With the aid of this mode of experimenting, 

 M. Feddersen has been able to study the constitution of the dis- 

 charge, and even its subdivisions. 



The other method, given by Arago for the purpose of getting a 

 limit of the duration of lightning-flashes, requires the employment 

 of a disk moveable about an axis perpendicular to its plane. This 

 disk is divided into sectors by radiating lines with equal intervals, 

 and appearing bright upon the darker ground of the disk. From 

 the widening of the lines produced by the light of the sparks their 

 duration may be estimated when the velocity of rotation of the disk 

 is known. This process of experimentation has been followed by 

 M. Felici, who studied by transmission the widening of the trans- 

 parent lines of an opaque disk, illuminated by the discharges of a 

 Leyden jar, when divers circumstances of their production are varied. 



MM. Lucas and Cazin have employed a method which permits 

 easier measurements, and in certain cases more precise, than the 

 preceding, but without distinguishing whether the sparks result from 

 one or several successive discharges. It consists in using a moveable 

 disk, the margin of which, intended to be viewed by transparency, 

 is interposed between the observer and the spark to be studied. This 

 disk, formed of plates of mica, bears on its margin transparent equi- 

 distant lines as fine as possible, obtained by photographic reproduc- 

 tion. It is placed in front of a second disk, opaque, of the same 

 diameter, which remains fixed, and has on its margin seven trans- 

 parent lines including six divisions, the breadth of which corresponds 

 to that of five divisions of the moveable disk, so that this second disk 

 forms a vernier by means of which ^ of a division of the first can be 

 estimated. This vernier constitutes the very ingenious novelty of 

 the method. 



The electric spark studied explodes in the focus of the lens of the 

 collimator, which sends rays, parallel to the axis of rotation of the 

 moveable disk, upon the lines of the fixed vernier. A telescope on 

 the other side of the disk permits the observer to examine the lumi- 

 nous appearances. 



If the spark has an inappreciable duration, the observer will see 

 only a single bright line or none. In the first case the spark has 

 flashed at the moment of coincidence of a line of the moveable wheel 

 and a line of the fixed vernier ; in the second case the spark took 

 place between two coincidences. 



The probabiUty of co'ncidence, which depends on the breadth of 



