APPARA TUS FOR SCIENTIFIC DEMONSTRA TION 159 



is a good arrangement forgone fixed position in a theatre, if the 

 optician understands his business and arranges matters so that 

 the arc-light is at a focus which gives an even disc at the 

 given distance ; but if this is not attended to, a good disc of 

 clear diagram will rarely be obtained. It is also best suited 

 for old-fashioned lamps, with the carbons in perpendicular line, 

 and was in fact first designed for spectrum-work, the nozzles 

 being so arranged that, 

 without moving the lan- 

 tern, the rays deflected 

 by the prisms should fall 

 on the same screen as 

 diagrams projected direct. 

 To accomplish this, how- 

 ever, the same dispersive 

 power must always be 

 used : otherwise the lan- 

 tern must be made to re- 

 volve upon the circular rim 

 supported by the pillars. 



When thus modified, 

 however, it is better to 

 have at once a tri-unial 

 lantern, which is the most 

 complete instrument of 

 all, when properly con- 

 structed. The first at- 

 tempts in this direction 

 were not successful, the 



three nozzles being arranged on three adjacent sidgs. 

 hexagon, copying an old exhibition 

 Beech ey. The objection to it is, 



directed towards the screen, another, or t^o oUi^rs, afeqjf qj< 

 in front, and interfere with manipulatio^. ,Tto f&n is a 

 one, and such lanterns have never coi&e 



FIG. 85. Double Electric Lantern 



