144 Mr. R. Beattie on the 



As regards the material of the poles, for instance, it is seen 

 that, other things (rate of break and electromotive force) 

 being the same, some substances (like platinum) give a much 

 longer spark than others (as carbon). This, of course, is a 

 well enough known fact, the explanation of which is not far 

 to seek: it is a result of carbon being a readily disintegrating 

 or easily arcing substance, and so setting up at the primary 

 break a heavy arc which, by bridging across the gap, renders 

 the rupture less abrupt. 



Again, with respect to rate of break, it is seen that a quick 

 break corresponds to a long and a slow break to a short 

 spark. This, too, can readily be understood; but it is not so 

 easy to comprehend the part played by the primary electro- 

 motive force in determining the length of the break-spark. 

 It might be expected that the existence of a high electro- 

 motive force in the primary, by causing the arc set up at the 

 primary break to be maintained for a longer period, would 

 result in a short secondary spark. Nevertheless, the ex- 

 periments show that this is not invariably the case; but that 

 while platinum and copper poles behave in the way antici- 

 pated, and give a shortened secondary spark for an increased 

 primary electromotive force, yet carbon, zinc, and lead poles 

 behave differently, for with them the spark-length sometimes, 

 ?'. e. when the break is slow, actually increases as the electro- 

 motive force is raised. 



The curves of figs. 2 and 3 further serve to bring out the 

 relation between the lengths of the make and break sparks, 

 and so help to indicate, amongst other things, the limitations 

 to the truth of the statement that the break-spark is longer 

 than the make- spark. 



Break-Sparks with Condenser. — By the introduction of a 

 condenser in parallel with the primary break the length of 

 the secondary spark is, as a rule, much enhanced. It has 

 long been established, however, that there is a best value for 

 the capacity of the condenser which gives a maximum 

 length of spark. Experiments have been made by Mizuno* 

 which show how the value of this " best capacity " and the 

 length of spark corresponding to it depend upon the magnitude 

 of the current interrupted in the primary. But it may still 

 be asked: For a given current broken in the primary, how 

 do the best capacity and the maximum spark-length depend 

 on the primary electromotive force, on the material of the 

 primary poles, and on the rate of break 'i An answer, in part 

 at least, to some of these questions is supplied by the results 

 contained in Table III., and illustrated in figs. 4, 5, and 6 

 * Phil. Mag., May 1898. 



