254 



practically constant. The following table shows the changes which occni'recl 

 as the speed of the machine was increased : 



Potential expi'essed in volts, currenL expressed 4n amperes. Distance 

 between points. 18.05 cm. 



Table 1. 



Speed 



Voltage 



before 



magnet 



was excited 



Voltage 



with 



Magnetism 



Current 



no 



Magnetism 



Current 



with 



Magnetism 







Type of discharge 

 Effect of magnetism on form of discharge 



1 



23000 



24000 



.00014 



.00014 



Glow discharge. 



2 



24500 



24500 



.00017 



.00017 



" 



3 



25300 



25300 



.00021 



.00021 



Small brush at anode. 



4 



26200 



26200 



.00025 



. 00025 



Increased brush at anode. 



5 



26500 



26500 



.00029 



.00029 



Occa.sional spark. 



6 



27000 



28300 



.00035 



.000.35 



Spark changed to brush by magnetism. 



7 



25300 



28300 



.00037 



.00037 





8 



24000 



28300 



.00044 



.00044 



" " 



9 



23000 



28300 



.00052 



.00049 



" 



10 



22800 



28300 



.•00059 



.00054 



" ' 



11 



22800 



28300 



.00058 



.00056 



" " " " " 



12 



23000 



28300 



. 00065 



.00059 







13 



22800 



28300 



.00072 



.00069 



" partially changed to ttriish by magnetism. 



14 



22300 



26500 



.00078 



.00076 



" " " " " •' " 



15 



22300 



25000 



.00083 



.00083 



Path curved but spark not stopped. 



16 



22300 



25000 



.00086 



_. .00084 



spark scattered. 



In the above table, the current wa.*^ measured by means of a Weston 

 milli-ammeter, and the potential by means of an electroscope. This elec- 

 troscope was made of two brass discs 10 cm. in diameter mounted in ver- 

 tical planes on ebonite supports which were fitted to a common base. The 

 distance between the plates could be varied by moving the supports. At 

 the top of one of the discs was soldered a support holding a small needle 

 upon which was suspended a brass vane which carried a ix)inter at the 

 lower end. The pointer moved in front of a scale which was calibrated 

 by connecting in multiple with the discs two No. 12 Thomas Harper needles 

 (sharps), measuring the critical spark length between them and comparing 

 with the table prepared by H. W. Fisher.' The position of the pointer was 

 read through a telescope placed two meters in front of the scale. The 

 potential read by this apparatus amounted to only a rough estimate, since 

 it could not safely be trusted nearer than 150 volts. This was especially 

 true when the sparks did not pass rapidly in succession because the vane 



* H. W. Fisher, Transactions of International Electrical Congress, Vol. 2, pp. 

 294-312, St. Louis, 1904. 



