K. Honda and S. Shimizu — String Alternator. t>5 



its direction once in a complete vibration. Thus, suitably 

 changing the length (by sliding two wedge-shaped blocks) or 

 the tension (by changing the suspended weight) of the string, 

 an alternating current of desired frequency is obtained. 



Now, in the above method, only a part of the current 

 employed can be turned to the alternating current ; in case where 

 a strong alternating current is required, it is sufficient to use two 

 equal sets of accumulators ; two poles of the one set being 

 connected to M. x and P, and those of the other to M s and S, so 

 that the currents through the portion of the circuit C are 

 directed in opposite sense. A sensitive galvanometer, inserted 

 in the circuit, shows whether the mean strength of the alter- 

 nating current in both directions is equal or not. The small 

 difference in the current-strength can easily be effaced by 

 adjusting the spark gap of the mercury contact. 



To cut off the electrical connection at the middle point of 

 the wire, the following method proved to be the most satis- 

 factory. A silk-covered wire of suitable length is cut into two 

 parts. The ends are twisted together like a rope for the length 

 of l cm and bent at right angles to the wire as shown in the 



annexed cut. The twisted por- 

 tion of the wire is firmly bound 

 together by a silk thread, and 

 the remaining portion of the 

 cover is scraped off. Though 

 different specimens of wire 

 were tried, copper and alumi- 

 num wires gave the most satisfactory result. 



The simplicity of the present arrangement for obtaining an 

 alternating current of definite frequency is one of its chief 

 merits. The actual frequency, if the tension of the string be 

 not small, was found not to differ by more than 1 per cent 

 from the value calculated by the formula 



n - y _ 

 I P 



where I is the length, T the tension and p the linear density of 

 the wire. It was not difficult to obtain vibrations ranging 

 from 30 to 1000 per second. 



It may not be out of place to remark that in using the 

 present arrangement, the self-induction of the circuit is to be 

 reduced as small as possible, inasmuch as it diminishes the 

 strength of the current. Hence in case the self-induction of 

 the circuit be not small, a large number of accumulators must 

 be used ; otherwise the string would not maintain its vibra- 

 tions. Measuring the mean strength of the current by an 

 ampere-balance, we found but a small fluctuation, if the current 

 does not exceed 1*5 amperes. 

 , Physical Laboratory of the Imperial University, Tokyo, Japan. 



Am. Jour. Sci. — Fourth Series, Vol. X, No. 55. — July, 1900. 

 5 



