640 



be illustrated in tlie next fig. 4. Tlie gfeater the speed of the signals 

 so iiiiicli the gi-eater the ii)teiisit3' of the inagnelic field, i. e. so much 

 the larger the decrement has to l)e made. The maximum speed of 

 still readable signals being al>oul 600 words pei' minute is obtained 

 when the intensity- of tlie field is maximum, being 22.600 Gauss 

 in one of our instruments. 



If sufficiently strong signals be available, the allowable speed 

 could slill be increased by admitting air around the string. 



One of the diflicultios that had to be overcome on designing the 

 instrument was the adjustment of Ihe tension of the sliing. This 

 must be secured in an exceedingly precise and punctilious manner. 



We stretch the string by extending it. In the figures i-eprodnced 

 above a stiirrg has beerr used of a leiiglh of 6 inillim., stretched so 

 that it was in turre with a wave of 7,5 liiloni. Su|)pose that it then 

 be extended to an amount of 1 °/. ^"d 'h"^ be lengthened by 60;i. 

 If the current exciting the electromagnet is 1 Amp. I he decrement 

 of the string is 0.25 < -10 ^. b'r-oni this it can be calculated that an 

 increase of the elongation to arr anrorrnt of 4,8 ,(t(n, snfifices to briirg 

 the string so much out of turre thai ihe amplitude of its vibrations 



1 



will decrease irr the proportion of J : — - i.e. 30 7» • The same effect 



I 2 



is produced by changing the wave-lejrgth of the signal to an amouirt 



of 30 ceirtim. on 7,5 kiloin. 



For the above calculations formula (2) has been used : 



- = — (2) 



where ^., repi-esents the wave-length of corrtinuous waves indrrctively 

 coupled with a cirrrrit and being in tune with its proper period, 

 and P., the wave-length that is so much smaller- oi' larger than the 

 former orre, that the electric power' of (he circuit is r'educed to otre 

 half. Here the movement of the sirirrg is substitrrted for the coujrled 

 current, and the amplitude of its vibrafioirs for the square root of 

 the electric power. 



It rreed not be emphasized, that much smallei' changes of the 

 amplitude are measurable than the value which is mentioned above 

 for coirveriience' sake. The apparatus for stretching the sliing must 

 enable elongalion to Ite effected within certain limits with absolute 

 regularity, and by degrees smaller than 1 fifi. Both of our present 

 models comply with this reqiiireinent. 



The experiments performed with the galvanomeler- have brought 

 to light some [ihenoinena concerned with the movement of string 



