THE AID OF THE ACHROMATIC FRINGES. 69 



field and on the degree of resonance, while the phases change with the latter. 

 Thus if the resonance is not adequately perfect, linear fringes only are obtained 

 and these change through inclined fringes, through normal fringes, into fringe- 

 lines with an inclination opposite to the first, while the inducting solenoid 

 changes its position continuously through the symmetrical position. Thus 

 the solenoid may either be moved parallel to the vibrator, between positions 

 on opposite sides of it ; or it may be correspondingly rotated in any given direc- 

 tion, so that the effects of its opposite ends respectively supervene. 



When the degree of resonance is sufficiently accurate, the axes of the fringe 

 ellipse are parallel to the sides of the vibration rectangle. In other words, 

 there is a phase difference of 90 between the vibrating objective of the tele- 

 scope and the vibrator (cc, fig. 61). This may be used as a sharp criterion for 

 resonance. Thus, I placed the solenoid in a non-symmetrical position, so 

 that the vibrating magnetic field was seen at the vibrator. Then by very 

 gently changing the tension of the wires (system v, t, fig. 61) from excess to 

 deficiency, fringe-lines inclined upward on the right with a range of 5 scale- 

 parts changed to the orthogonal ellipse with a range of 20 scale-parts, into 

 fringe-lines inclined downward on the right, ranging 5 scale-parts. When the 

 tension was thereafter gradually increased, the same sequence was observed 

 in the opposite direction. Again, if the resonance is adequate, the orthogonal 

 ellipse does not change in phase when the solenoid is moved, as specified. The 

 passage is through normal fringe-lines; but if resonance is only slightly im- 

 perfect, the oblique ellipse passes through oblique fringe-lines into a symmetri- 

 cally inclined ellipse. 



The criterion is sufficiently sensitive to indicate a change of period with the 

 amplitude of the interrupter. Thus an orthogonal resonance ellipse was pro- 

 duced with a range of 25 scale-parts. An inductance was now put into the 

 primary, decreasing the image band-width (vibration of the objective) and 

 decreasing the fringe-lines to a range of 5 scale-parts. The tension of the wires 

 of the vibrator was now carefully decreased, little by little, until an orthogonal 

 ellipse with a range of 25 scale-parts was obtained. Thereafter the inductance 

 was again removed. The band- width of vibration increased again, but the fringes 

 became lines of symmetrically opposite inclination. Increasing the tension of 

 the wires slightly brought back the original orthogonal ellipse with a range of 

 25 scale-parts. Many other experiments similarly striking might be instanced. 



The orthogonal ellipse is clearly compatible with maximum fringe displace- 

 ment. For the magnetic field changes sign at the maximum elongations of the 

 vibrator. If the ellipse is permanent, the periods of field and vibrator are 

 rigorously equal. 



49. Resistance, capacity, inductance, in case of the compensated vibrator. 



With the fringes for i = o in the secondary reduced to bands, a number of 

 experiments on the effect of resistance were made, examples of which are 

 given in figures 75 and 76. In the former case a coil with 10 turns (e lo- 2 

 volt) was slid over the primary and large additional resistances up to R = 



