400 Dr. F. P. Kerschbaum : Interference 



In fact, with the conception of the unit alone as sketched 



above. — i. e. as a train of oscillations moving along the tube 



© © 



of force in something like a straight line, — we cannot account 

 for interference. 



So we have only to make sure for our experiments that 

 the source of light which we are going to use does not emit 

 the units with any regularity, so that the different units do 

 not arrive at the interference apparatus with a fixed phase 

 relation. We shall show later that this is the case in our 

 arrangement. 



The Source of Light. 



The condition that light units of corpuscular cross-section 

 can escape from a gas which has to act as source of light and at 

 the same time necessarily fills the whole of the interference 

 arrangement without practically being stopped by the mole- 

 cules of the gas, is fulfilled, if the total cross-section of the gas 

 molecules, contained in a column of the length of the apparatus, 

 is small compared with the cross-section of this column. We 

 shall see later that the length of a suitable interference 

 apparatus is about 50 cm. From the known numbers and 

 dimensions of the gas molecules we can therefore conclude 

 that e. g. for H 2 the pressure of the gas in the apparatus 

 would have to be about 0"0001 mm., if only 0*1 of a given 

 cross-section is filled with H 2 molecules ; in other words, at 

 a pressure of 01)001 mm. only 10 per cent, of the emitted 

 light units are liable to hit H 2 -molecules. The order of 

 magnitude of this pressure is about the same for all the gases. 



Now it is known that canal-rays in a gas of or under 

 0"01 mm. pressure already emit light which is too feeble for 

 interference experiments of finite duration. The direct pro- 

 duction of light in gases at such low pressures, as required 

 above, seems therefore not to be practicable by any sort of 

 electric discharge. Moreover, such light is never homo- 

 geneous, and therefore not suitable for interference experi- 

 ments in w T hich the result depends upon the distinctness of 

 the fringes. 



In his recent work on the fluorescence of vapours Wood * 

 has discovered a very interesting source of ultra-violet light. 

 Hg-vapour in a vacuum at room temperature, i. e. at a partial 

 pressure of about 0*001 mm., can be excited to fluorescence 

 radiation by illumination with light- waves of the wave-leno;th 



o«/ © ^> 



X=2536 A.U. This fluorescence light is entirely mono- 

 chromatic and of the same wave-length as the exciting light, 

 it is scattered resonance light. It corresponds to the sharp 

 * R. W. Wood. Phil. Mag. xxiii. p. 689 (1912). 



