312 CARNEGIE INSTITUTION OF WASHINGTON. 



modified telephones. If, for instance, one of these mirrors vibrates 

 synchronously with the other towards and away from it, the ellipses 

 remain in the field without displacement; if, however, the mirrors 

 move in the same direction, the eUipses necessarily vanish. Hence, if 

 there is a gradually increasing retardation in the electric circuit of one 

 telephone only, the ellipses will alternately appear and vanish for iden- 

 tical intervals of retardation. The experiments attest the feasibility 

 of so mounting the telephone mirrors on the interferometer that the 

 balance may be realized. Finally, an attempt is made to study 

 double refraction resulting from dielectric polarization both in solids 

 and in liquids by the above method, while in allied sections the pris- 

 matic spectrum is studied on a Rowland spectrometer and certain 

 peculiar results are shown in the resolution of interference fringes 

 resembhng the elliptic type. 



The interferometry of highly exhausted air carrying electric cur- 

 rent is briefly treated in the next chapter. The adjustment used is 

 that originally proposed by Mach, in which the rays do not retrace 

 their path but move along the sides of a parallelogram. The long 

 exhaustion tube carrying current is placed in one of these sides, the 

 light passing from end to end, and the effect of presence and absence 

 of current on the interference pattern is determined. Thus far the 

 exhaustions available were insufficient, so that the experiment in its 

 final developments will be resumed at some time in the future. It 

 is worth an inquiry to ascertain whether the motion of the cathode 

 rays in a given direction may not produce some preponderating modi- 

 fication of the properties of the ether in that direction. 



Chapter IV deals with the refraction of air at high temperatures 

 investigated by the displacement interferometer (provided with water 

 circulation) by comparing the refraction of a vacuum and of a 

 plenum of air at each given temperature. As the system is not 

 very sensitive to pressure, very perfect exhaustion is superfluous, 

 but the need of sealed apparatus makes the experiments difficult at 

 high temperatures. This chapter is but a beginning of experiments 

 of this kind, which are naturally very difficult. It is curious that the 

 refraction of Bunsen flame through which a beam of light passes 

 symmetrically may be approximately found by this means. 



Finally, in Chapter V the displacement interferometer is applied to 

 the electrometer. A large number of forms of the instrument are 

 designed and the results of each tested in detail. All of these are 

 comprehended by the closed cylindric electrometer, in which a closed 

 cylindrical charged needle is capable of displacement along the axis 

 of a closed cylinder, consisting of two symmetrically insulated halves, 

 oppositely charged. The suspension of the needle is of the pendulum 

 type. If the ends of each cylinder are removed the instrument 



