258 



rays are supposed to pass tlu-ough the window W and fall on some feebly 

 luminous object such as a heated platinum wire or a calcium sulphide 

 screen at S. Both the platinum wire and the sulphide screen were 

 used and Avlien viewed by the eye thrcugh ground glass at various 

 angles and positions relative to the source nothing definite was noticed. 

 The feebly luminous spot at times appai-ently brightened, then moved 

 around in a circle and went through a series of displacements. This 

 proved that nothing definite can be arrived at by viewing directly with 

 the eye. 



The most reliable method of recording the action of a feebly luminous 

 source is photography. With tliis method, direct and indirect vision is 

 eliminated, as well as the error due to the increased sensitiveness of the 

 eye after being in the dark for some time. A number of photographs 

 were taken, on Seed's regular "gilt edge" plates, with the light from a 

 heated platinum wire, a luminous calcium sulphide screen, and a feeble 

 spark. 



The Platinum Wire. 



The platinum wire was a very thin strip cut from a piece of foil 

 .03 mm. thick, so that in no place was the wire more than .05 mm. 

 broad. Only one place along it was allowed to be heated and the ap- 

 proximate breadth of this place was .08 mm. The wire was heated by 

 a current approximately .9 amperes from three or five Edisou-Lalande 

 batteries. In some of the latter experiments a storage battery was used. 

 The relative position of the different parts of the ai)i)!iratus is shown 

 in Fig. II. 



==i 



B is a cardboard l)ox in wliicli is p'aced the plntinuni wire. The 

 platinum wire is soldei-ed to two coppi'r wires wliicli are fastened to a 



