72 DISPLACEMENT INTERFEROMETRY BY 



taken in succession, so that an initial 5 is again in question. Since C = 6 X io~ 6 

 farad, if the mean value of Cu be inserted. 



960 



co = = loo 



6 



and the frequency therefore n= 160/6.28 = 25 per second. This is not in bad 

 agreement with facts. 



To measure a self-induction in the secondary is difficult, unless it is large 

 in comparison with the effect of the resistances there. (53.) It might 

 however, seem to be measurable in the primary if adequate provision were 

 taken to guard against the direct effect of the stray magnetic lines in the 

 vibrator, in the way suggested by the experiments of 45 ; for here the non- 

 inductive resistances are but a few ohms. I made a variety of interferometer 

 experiments of this kind, but without success. 



50. Ring transformer. A little ring transformer, consisting of about n 

 turns of wire per centimeter wound on an iron ring 7.5 cm. in average diameter 

 and i cm. thick, was used in place of the above solenoid. In this case there is 

 of course no external magnetic field and the vibrator is not influenced. With 

 an open secondary (i to 10 turns were used) the fringes are normal bands. 

 The results obtained with this apparatus were similar to the above and may 

 therefore be omitted here. 



In one respect, however, this apparatus is disadvantageous, as (owing to the 

 thickness of solid iron of the ring-body) the inductive effect of a reversal of 

 the magnetic lines within on the secondary is almost tumultuous. In other 

 words, in commuting the secondary current (even if induced in a single turn 

 of wire with high resistances, R, inserted), the ellipses enlarge enormously 

 (similarly to fig. 73) and then roll through bands several times. All this takes 

 considerable time, even when the final ellipse is of small range. In the case 

 of the solenoid with tubular core, the ellipses under like conditions merely 

 expand or contract with change of phase. A hollow annulus therefore would 

 best meet all conditions, or it may be even preferable to dispense with iron 

 altogether. Thus a coil was wound on a split 0.2 5-inch lead tube, which was 

 thereafter bent into a ring. A few turns of secondary may then be used and 

 the ellipses kept in the field without additional resistance. 



51. Magnetic screens. A number of special observations were made by 

 screening the primary linear solenoid with metallic tubes (capable of sliding 

 over it) from the influencing vibrator. For this purpose the primary was so 

 placed as to give a large deflection by direct magnetic induction, while the 

 secondary current was broken (^=00). A ^g-inch gas pipe slid over the prim- 

 ary then reduced the deflection, s, about 50 per cent, though a few centimeters 

 at one end of the primary had to be left uncovered. Such a result would, in 

 general, be expected from the theory of screening. 



On the other hand, when a thick brass tube was slid over the solenoid, the 



