140 Mr. R. Beattie on the 



in the primary was always 25 amperes, supplied from a 

 battery of secondary cells. The electromotive force could 

 thus be altered by varying the number of cells, and the 

 current was adjusted to its propsr value by the introduction 

 of sufficient non-inductive resistance. The primary, there- 

 fore, formed part of a circuit carrying, when closed, always 

 the same current of 2'5 amperes, and having always the 

 same inductance (of about 0*4 henry), but whose resistance 

 varied in proportion to the electromotive force acting in it. 



Tn place of the usual vibrating interrupter a special 

 contact-breaker was employed, by means of which the 

 primary poles could be separated either slowly or quickly as 

 desired. The essential parts of this arrangement are shown 

 diagrammatically in fig. 1. A light lever OPi, hinged so 



Fig. 1. 



O P 2 a 



I * 



a 



4 



that it can turn about one end 0, carries at the other end one 

 of the primary poles P x . Immediately underneath is the 

 second pole P 2 . The two are kept in contact and the circuit 

 closed by the weight of the lever. To break the circuit 

 a wooden disk, capable of rotating round its axis D, is 

 furnished with a radially-projecting metal lug L, and has a 

 point on its rim connected by cord with a rubber band I. 

 By rotating the disk counter-clockwise till the lug engages 

 the catch 0, the rubber is stretched, so that, on release, the 

 disk flies round and the lug delivers a blow to the end of the 

 lever, thereby separating the poles with a rapidity governed 

 by the initial tension of the rubber. 



In what follows, three different rates of break are referred 

 to, viz. slow, moderately quick, and very quick; and in the 

 curves afterwards given these are denoted by the numerals 

 I., II., III. respectively. The slow break was produced by 

 raising the lever rather slowly by hand, no great care being 

 taken to separate the poles always at precisely the same rate, 

 as it was soon found that, up to a certain point, rate of break 

 has little influence on spark-length. The quick breaks were 

 obtained by the mechanical throw-up arrangement, the mod- 

 erately quick and the very quick breaks corresponding to 

 lug-velocities of about 8 and 18 metres per second at the 

 moment of impact. 



As secondary poles the blunt polished points of the coil 



