Barus — Displacement Interferometer. 69 



facing the ruled side of the grating plate usually appear as sep- 

 arated images on the mirror, about 1/8 inch apart, and the 

 one yellowish in color may be blotted out by a small screen 

 near the mirror. As a result of this the stationary interfer- 

 ences which are due to the two beams from the same mirror 

 also vanish. The spectrum is now clear and contains only the 

 interfering superpositions, giving the ellipses their maximum 

 definition. 



In other respects the mounting on the interferometer is the 

 same as before, and very little difficulty was experienced in 

 obtaining the ellipses. The top of the case DD, here necessa- 

 rily a square box, may usually be removed without disadvan- 

 tage. If liquid damping is necessary in case of a heavy 

 cylinder kk, the wire with vane may be dropped into a vessel 

 of oil from the rear end of tt, the front carrying the mirror. 



The usual method of connecting was used for potential 

 measurements, one pole of the experimental cells being alter- 

 nately earthed as well as one pole of the charging battery. A 

 Mascart key was found very useful. When the needle swings 

 to and fro the ellipses are visible at the extreme elongations of 

 the deflection (if not too large) of the vibrating cylinder. They 

 flash into the field of the telescope at the two ends of it, but 

 are invisible between. The distance apart of the positions of 

 visibility gradually diminishes until they coincide and the cyl- 

 inder is stationary. If the ellipses are clear, slight motion does 

 not interfere with the measurements, as the mean position is 

 readily determinable in the displacement method. This would 

 not, however, be the case had the evanescence of rings been 

 made the basis of measurement. 



4. Aquations. Cylindrical electrometer. — The equations 

 may be found in the usual manner, and as so little of the field 

 is displaced during deflection they apply with acceptable accu- 

 racy. Virtually a small cylindrical shell of uniform field is 

 supposed to disappear on one side to reappear on the other. 

 Hence, let V 3 , V t , and V i be the potentials of the two fixed 

 and a movable cylinder (needle) respectively. Let f lt denote 

 the electric force between cylinders (2) and (1), and f 13 the 

 force between cylinders (3) and (1), and R the common radius 

 of the fixed cylinders, r that of the movable cylinder. Hence, 

 if R-r is small, 



/»=( y~ V,)/(X-r) ; /» = (*V- V,V(X-r) 



If the displacement of the cylinder is dx the changes of poten- 

 tial energy on the two sides are 



j> + r V — V E + r V —V 



