MEMOIR XXII.] THE ACTION OF GLASS AND QUARTZ. 



in a way entirely satisfactory. I resorted, therefore, to 

 the brilliant light obtained when a piece of metal, or, 

 what is better, the hard variety of carbon obtained from 

 gas-retorts, is lowered upon mercury entirely filling a 

 very small open porcelain cup, and the continuous dis- 

 charge of a voltaic battery passed. The battery used 

 contained fifty pairs of Grove's cells, but a smaller num- 

 ber would have been amply sufficient. 



As soon as the light was emitted, I marked on the 

 lime sulphide the beginning of the red, the centre of the 

 yellow, and the termination of the visible violet ray. 

 Then, stopping the current, I examined on what parts 

 the plate was phosphorescing. The commencement of 

 the glow was between the indigo and the blue ; towards 

 the blue it extended far beyond the visible boundaries 

 of the spectrum. I could not see any divisions or points 

 of maximum in it. The surface of the plate shone all 

 over except in the region of the less refrangible rays, 

 and there were traces of the negative action which M. 

 Becquerel has illustrated in the cases of the solar ema- 

 nations rays which, however, were first observed in the 

 last century. 



It is necessary to remark that the rays from the voltaic 

 discharge resemble those from an electric spark in their 

 inability to pass through glass. On this observation all 

 the value of the foregoing experiment depends. 



But it can nevertheless be easily proved that although 

 glass is impervious to the phosphorogenic emanation 

 coming from the voltaic deflagration of any metallic bod- 

 ies, the observation applies to transient discharges only. 

 A voltaic light which lasts but a moment fails to cause 

 phosphorescence through glass in the same way that an 

 electric spark does; but if the discharge be continued 

 the surface presently begins to glow, and if maintained 

 for several minutes it shines as brightly as though a 

 piece of quartz had been used. 



