268  Mr.  A.  Russell  on  the 
(ii.)  Compagnie  de  r  Industrie  Electrique.     2  cm.  spheres. 
The  Compagnie  de  Flndustiie  Electrique  et  Mecanique 
of  Geneva  have  published"*  a  curve  giving  the  sparking- 
distances  between  two  spherical  electrodes,  each  one  centimetre 
in  radius.  The  frequency  of  the  alternating  pressure  em- 
ployed was  50,  and  the  ratio  of  the  maximum  to  the  effective 
voltage  was  1*26.     Calculating  Emax.  by  the  formula 
Rmax.  =  {(l-26V-0-8)/a-}/ 
for  values  of  x  from  0*5  cm.  to  5  cms.,  we  find  that  the  mean 
value  of  Emas.  is  37*9?  which  practically  agrees  with  Stein- 
metz's  result  for  2  inch  spheres. 
(iii.)   C.  P.  Steinmetz.     1,  0*5,  and  0*25  inch  spheres. 
The  analysis  of  Steinmetz's  experiments  with  1,  0*5,  and 
0*25  inch  spheres  are  instructive,  but  for  reasons  explained 
in  §  9  they  do  not  give  much  assistance  in  obtaining  Emax. 
With  the  1  inch  spheres  the  mean  of  the  values  of  Emas. 
obtained  up  to  pressures  of  63 '7  effective  kilovolts  is  41*3 
kilovolts  per  centimetre.  With  the  half-inch  spheres  the 
mean  of  the  values  for  pressures  up  to  31*3  effective  kilovolts 
is  43'1. 
When  the  quarter-inch  spheres  were  28  cms.  apart,  the 
disruptive  pressure  was  112  effective  kilovolts.  It  we 
calculate  Emax.  for  this  pressure  as  if  the  spheres  were  in  a 
vacuum,  we  find  that  it  is  more  than  six  times  the  dielectric 
strength  of  air.  In  the  experiment  there  must  have  been 
coronse  round  each  of  the  electrodes  after  the  pressure  was 
about  17  effective  kilovolts. 
In  these  experiments  the  frequency  was  125  and  E/V  was 
equal  to  1*12. 
(iv.)  E.  Jona.      Point  and  Plate.      Two  spheres. 
E.  Jona  has  published  f  a  table  giving  the  sparking- 
distances  between  a  point  and  a  plate  and  also  between  two 
equal  spherical  electrodes  each  of  2  cms.  diameter,  for 
pressures  varying  from  15  to  240  kilovolts.  When  the 
electrodes  are  far  apart  it  is  obvious  that  the  distribution  of 
the  Faraday-tubes  is  considerably  affected  by  the  supporting- 
rods  connecting  the  electrodes  with  the  transformer  terminals. 
An  important  result  proved  by  these  experiments  is  that  for 
all    distances    greater   than   23   cms.   the   sparking-voltages, 
*  Turner  and  Hobart,  '  The  Insulation  of  Electric  Machines/  p.  35. 
t  E.  Jona,  Atti  delV  Associazione  Elettrotechnica  Italiana,  vol.  vi.  p.  3, 
1  Distanze  Esplosive  nell'  Aria,  negli  olii  ed  altri  Liquidi  Isolanti." 
