346 Scientific Proceedings, Royal Dublin Society. 



site sides of the plane of division, and the combined prisms shifted 

 till the intensity of the light received at opposite sides of the 

 divisional plane is brought to equality — till, in fact, the prisms lie 

 in an equally illuminated field. In this position the discontinuous 

 appearance of the prisms vanishes, and they look like one con- 

 tinuous block of paraffin. While, however, ■ this complete dis- 

 appearance of the divisional line takes place when the sources of 

 light are emitting rays of like colour, in general comparison is not 

 between such, and a certain amount of judgment is required to 

 determine when the two differently tinted luminous bodies are 

 emitting equal quantities of light. This objection applies to all 

 photometers of visual comparison : in this form the difficulty is 

 much lessened by watching solely for the least conspicuous ap- 

 pearance of the line as the prisms are shifted. Tinted glasses 

 to cut off these rays not common to both sources of light can, 

 of course, be used. 



It is found that more sensitive measurements are obtained 

 from this form of photometer than from Bunsen's ; the operation 

 of adjustment also is less fatiguing, and is more rapidly effected. 

 The sensitiveness of the arrangement is still further increased by 

 interposing a silver leaf between the prisms, thus stopping all 

 passage of light from one prism to the other. The addition of a 

 lens for observing the divisional line adds to the facility and ease 

 of adjusting. In this form, suitably mounted and protected from 

 accidental sources of light, it will be found that the least move- 

 ment from its position of equilibrium is noticeable. This ex- 

 treme sensitiveness and ease of working is due to the very close 

 juxtaposition of the luminous surfaces undergoing comparison. 

 In Bunsen's photometer the surfaces compared are seen at some 

 inches removed from each other. 



Experiments have been made with alabaster, and with a highly 

 translucent marble known as Mexican agate. These substances, 

 however, proved inferior to paraffin. Good results would probably 

 be obtained from milk-glass. 



It is to be noted that if the prisms be not quite similar in 

 translucency (they should be cut from the same piece of paraffin), 

 it will be necessary to take two readings, with the faces of the 

 combined prisms reversed in position respecting the lights, the 

 mean of the two readings being taken as correct. 



