THE AID OF THE ACHROMATIC FRINGES. 



47 



The apparatus was now modified by inserting a stronger (commercial) 

 telephone. In this case the sensitiveness was so far increased that only cur- 

 rents not much exceeding milliamperes could be measured within the range of 

 the ocular micrometer. The sensitiveness So is 



In endeavoring to reach still higher frequencies, the phosphor-bronze wires 

 broke under the strain. 



It is difficult to account for this effect of frequency, so peculiarly marked 

 in the last instance, where the observations were very good. If different 

 harmonics are in action in case of the separate frequencies, the overtones 

 would have to respond in the case of the wires under less tension, and this 

 seems to be anomalous. Nevertheless, it is often possible to detect two cases 

 of resonance at different tensions, which I hold to be the fundamental, and the 

 octave, the latter obtainable for very low tension. Some suggestions will be 

 given in the two succeeding paragraphs, moreover. As the vibrators, inter- 

 rupter, etc., are damped systems, the periods will vary with their amplitudes, 

 so far as resonance is concerned. 



31. Steel wires. A number of trials were made with the bronze wires to 

 secure higher tension, but without avail. A steel wire was therefore inserted, 

 0.036 cm. in diameter, and the spring of the interrupter stiffened. Curiously 

 enough, the results with steel proved to be the reverse of the results with 

 bronze wire. In case of frequencies estimated in the ratio of 20 and 40 per 

 second, the sensitiveness was but 19 and 10 scale-parts per milliampere. 

 With the tense wire but a single harmonic could be found. The less tense 

 wire, however, responded once again on further decreasing the tension, with 

 an increase of sensitiveness of about 26 scale-parts per milliampere. 



In the work with these stiff wires, the tension is difficult to control to the 

 nicety required. One should expect a greater amplitude for the less tense 

 wire, as here found. 



A thin steel wire 0.015 cm. in diameter was also tried in contrast. Though 

 it showed considerable sensitiveness (up to 30 scale-parts per milliampere), 

 the system was inconveniently subject to tremor during adjustment. 



