. which has not been compressed, and the roiat'iGn in the cont^'ary direc- 

 tion is less. The following are the numerical results. 



" In one experiment I obtained on a piece of hea^y glass not com- 

 pressed, 3° of rotation to the right or to the left, according to the 

 direction of the current : on slightly compressing the glass, I had to 

 turn to the right the eyepiece to 4°, 5°, and even to in order to 

 restore the image to its first condition. In closing the circuit, the 

 rotation produced in the same direction as that due to compression 

 •was or 4°, while the rotation produced in the contrary direction 

 was from 2° to 1|°. On ceasing to compress the glass, I obtained 

 the same phienomena as I had observed before the compression. 



" I have made in the same manner experiments with a piece of 

 flint-glass, which produced a rotation of 2° under the influence of the 

 magnet. When I applied the same magnet to pieces of compressed 

 flint-glass, I could not discover the slightest sensible rotation in 

 whatever direction I might make the current pass. Plates of quartz 

 cut perpendicularly or parallel to the axis, and compressed in various 

 dn-ections, did not acquire any rotator}' power under the influence of 

 the magnet, I think that the peculiarity exhibited by compressed 

 heavy glass is of some interest, in as fai' as it appears likely to lead 

 to a more satisfactory explanation of the wa.nt of rotatoiy power 

 communicated by magnetism in crystalline bodies. 



" I shall conclude by communicating the negative results of some 

 experiments I attempted with a view to discover the action of dia- 

 magnetic bodies on each other, and of magnetism on gaseous bodies. 

 I suspended small needles of bismuth between the poles of a very 

 powei-ful electro-magnet, and with a good chronometer I counted 

 the number of their oscillations, either alone or in the vicinity of 

 pieces of bismuth of various shapes and sizes. I repeated these ex- 

 periments with ail possible care, avoiding the slightest current of air, 

 reckoning the smallest oscillations, and those of the same extent in 

 the different cases. I never obtained any difi^erences beyond half a 

 second, which existed equally whether the pieces of bismuth were 

 near or not. The experiment therefore does not serve to show the 

 action of dia^magnetic bodies on each other ; an action which natu- 

 rally ought to exist, but which perhaps is overpowered by the stronger 

 action of the magnet. 



" I afterwards counted the oscillations of a small needle of bismuth, 

 which I succeeded in suspending by a silk fibre {fil de cocoa) inside 

 of a glass ball blown at the top of a barometer-tube. The ball was 

 placed between the poles of my electro-magnet. In this experiment 

 the bismuth needle was held sometimes in a nearly perfect vacuum, 

 at others in atmospheric air. The number of oscillations in both 

 cases was exactly the same. 



" We must therefore give up the idea of explaining diamagnetic 

 phenomena by a magnetic action, which would be stronger upon the 

 air than upon bismuth." 



