. Carnegie Institution of Washington publication. 70 DISPLACEMENT INTERFEROMETRY BY 30,000 ohms inserted. The ordinates 5 (scale-parts) show the range of ellipses in the direction of the fringe vibration. The fringes being about a scale-part in width, the sensitiveness at high resistances was about io~7 ampere per fringe. In case of figure 76 the secondary consisted of but one turn (e= io~3 volts) and relatively low resistances were admissible. The sensitiveness here is about io-8 ampere per fringe. In case of the large fringes used in series 7 and 8 below, I found about ~8 ampere to the
. Carnegie Institution of Washington publication. 70 DISPLACEMENT INTERFEROMETRY BY 30,000 ohms inserted. The ordinates 5 (scale-parts) show the range of ellipses in the direction of the fringe vibration. The fringes being about a scale-part in width, the sensitiveness at high resistances was about io~7 ampere per fringe. In case of figure 76 the secondary consisted of but one turn (e= io~3 volts) and relatively low resistances were admissible. The sensitiveness here is about io-8 ampere per fringe. In case of the large fringes used in series 7 and 8 below, I found about ~8 ampere to the scale-part, which is probably the limit of the present apparatus. At R= 20,000 ohms the tele- phone was practically silent to the ear, but the vibrator gave distinct evidence of current even above 30,000 ohms. The curves (figs. 75 and 76) are so nearly hyperbolic that the inductance must be negligible compared with the high resistances. In the first series Rs is practically constant, a result which might be used for standardizing s in case of these particular fringes. In series (3) r is computed. from (R+r)s = const. For R below 500 ohms, the results for the effective resistance of the circuit come out between 400 and 500 ohms. To compute L2a)2=A(s(R-{-r))*/Asz would require a more accurate specification of r and better individual values of 5 than the method provides in the secondary. Thus, if r — 2$o ohms, Leo = 640 to 660; but this is clearly over five times too large. Probably 5 and i do not pass through zero together. In fact, if r = 280 ohms and if i/(R-\-r) and s be constructed as in figure 76, the relation is so nearly linear that L will be merged in the errors of observation. But the line suggests an initial s=i scale-part. In a later adjustment I reduced the induction till the ellipses remained in the field in the absence of extra resistance R. The resistance of the circuit itself with its 3 telephones was r = 28o ohms. Results were obtained, for example, as follows
Size: 2516px × 993px
Photo credit: © Book Worm / Alamy / Afripics
License: Licensed
Model Released: No
Keywords: ., bookauthorcarnegie, bookcentury1900, bookdecade1900, bookyear1902
