104 C. Barus — Interferometry of Air 



Bere J-V is the displacement of the micrometer correspond- 

 ing to the pressure difference p — p, at the constant tempera- 

 ture given, and fi the index of refraction of the air contained. 



The induction coil was now attached and a current from 3 

 to 4 storage cells sent through the apparatus. Not the slightest 

 shifting of ellipses or interference lines could be detected when 

 the secondary current was alternately made and broken. The 

 lines were very clear. The experiment was then repeated and 

 the air pump (G-eryk) connected directly with the tube without 

 accessory apparatus between. A coarsely stratified column and 

 some cathode dark space was obtained. But the effect at the 

 interferometer was again definitely negative. 



The absence of an effect due to the current was to be 

 expected, since the positive and negative currents and the 

 direct beam of light both traveled in opposite directions. The 

 same is true for the ionization, which is too slight; but it was 

 supposed that some evidence of a temperature effect might 

 be obtained ; hut there was none. 



Since p = Cfc- 1)0 = C^JYO/e, 



where f is constant and denotes absolute temperature, there- 

 fore at constant pressure 



d(AJV) _ (M 

 AiV" ~~ 



iicl^N') = 5 X lu- 5 cm., MST — 25 x lCT'cm., 



= 293°, d$ = -59° ; 



so that whatever rise of temperature may have occurred must 

 have been much less than *6° Centigrade. 



4. Second Experiments. Apparatus. — In this experiment 

 the direct and return beams m and n were separated as shown 

 in fig. 2, into m and m! and n and n', the beams traveling along 

 contiguous sides of a rhombus. J^and N are again the two 

 opaque mirrors, the former with micrometer ; C is the colli- 

 mator, open as in fig. 1. The grating above the tripod of the 

 interferometer has been removed and replaced by a plate of 

 glass P of about the same thickness, while the grating is at G 

 diametrically opposite, with its ruled face toward 7V. The 

 normals of the mirrors are shown at p and p', and the telescope 

 is either at R (direct ray) or at D (diffracted ray). It is here 

 convenient to have two telescopes, one set at R and the other 

 ati>. 



To secure this adjustment the interferometer of fig. 1 is pro- 

 vided with a cross arm of gas pipe under RG, which can be 

 clamped firmly in place on the standard of the tripod, below P 

 in fig. 2, or G in fig. 1. The end of the arm below G, fig. 2, 

 has a foot with a vertical set screw for imparting strain, and is 



