312 M. H. Becquerel's Experimental Investigations 



and which thus allowed the observed results to be reduced to 

 what they would have been if the electromagnetic intensity 

 had been perfectly constant. A mercurial commutator con- 

 veniently placed within reach of the observer, allowed the 

 direction of the electric current in the bobbins to be reversed 

 at pleasure, and the bodies studied to be subjected to inverse 

 electromagnetic action. 



The tube which contains the gases was originally 3 metres 

 long. It was constructed to maintain a vacuum and to sup- 

 port the pressure of several atmospheres ; for this purpose it 

 was terminated by very thick pieces of copper, which were 

 screened and soldered, and in which was drilled a rectangular 

 opening of O10 metre by 0*04 metre. 



These pieces presented a plane surface perpendicular to the 

 axis of the tube, and on which were fixed by means of screws 

 flat glasses enclosed in a frame of copper. The tube w T as pro- 

 vided with taps and a manometer; and the whole was so 

 arranged as to slip inside the bobbins. 



Various inconveniences, and amongst others a slight acci- 

 dent, which caused the simultaneous rupture of the two glasses, 

 led to the following modification of the apparatus. 



On the two plane surfaces which form the end of the 3-metre 

 tube (PL VII. fig. 1) are screwed and cemented cylindrical cop- 

 per tips 0*135 metre long, furnished at their ends internally with 

 a thread for a screw. The tube can be closed by two large copper 

 stoppers, which are screwed into the pieces just described, and 

 in which are cemented two flat glasses of about 0*005 metre 

 thickness, leaving for the light a rectangular opening of 0*09 

 metre by 0*06 metre. Washers of greased leather enable the 

 tube to be hermetically closed. On each side of these pieces 

 are arranged taps, which answer the purpose of either intro- 

 ducing the gas or putting the tube into communication with 

 a manometer. Before entering the apparatus, the gases are 

 passed through tubes containing substances necessary to 

 purify and dry them. 



The tube extends beyond the bobbins by about 0*135 metre 

 on each side, so that the glasses are subjected to only a very 

 feeble magnetic influence : we shall see afterwards how this 

 influence must be taken into account. 



The distance between the internal surfaces of the glasses is 

 3*27 metres ; the volume occupied by the gas submitted to the 

 experiment is about 37 litres. When the tube was hermeti- 

 cally closed it acted as a gas-thermometer, and the indication 

 of the internal pressure allowed the average temperature of 

 the gaseous mass to be deduced from it. The axis of the tube 

 made an angle of about 52° with the magnetic meridian. The 



