238  Mr.  A.  Russell  on  the 
11.  The  maximum  electric  intensity  "between  two  infinitely  long  parallel 
cylinders. 
12.  The  disruptive  discharge  between  two  parallel  cylinders. 
13.  The  application  of  the  formulas  to  experimental  results. 
I.  With  Direct  Pressures. 
(i.)  Lord  Kelvin's  tests  with  large  electrodes. 
(ii.)  A.  Heydweiller.     5  cm.  spheres, 
(iii.)  J.  Algermissen.     5  cm.  spheres. 
(iv.)  J.  Joubert   and   G-.  Carey  Foster.     1  cm.   and  2  cm. 
spheres. 
(v.)  E.  Hospitalier.     1  cm.  spheres, 
(vi.)  Compagnie  de  l'lndustrie  Electrique.    Plate  and  sphere. 
II.  With  Alternating  Pressures. 
(i.)    C.  P.  Steinmetz.^    2  inch  spheres. 
(ii.)  Comp.  de  l'Tnd.  Elect.     2  cm,  spheres. 
(iii.)  C.  P.  Steinmetz.     1,  Oo,  and  0-25  inch  spheres. 
(iv.)  E.  Jona.     Point  and  Plate.     Two  spheres. 
(v.)  C.  P.  Steinmetz.     0313  inch  cylindrical  electrodes, 
(vi.)  C.  P.  Steinmetz.     I'll  inch  cylindrical  electrodes. 
14.  Table  of  the  numbers  obtained  for  the  dielectric  strength  of  air  from 
the  direct  pressure  experiments. 
15.  Table  of  the  numbers  obtained  for  the  dielectric  strength  of  air  from 
the  alternating  pressure  experiments. 
16.  Conclusion. 
Appendix   I.  The  Disruptive  Voltages  for  large  Spherical  Electrodes. 
„         II.  The  Capacity  Currents  to  the  Electrodes. 
Postscript. 
1.  Introduction. 
PHYSICISTS  generally  attempt  to  deduce  the  dielectric 
strength  of  air,  at  a  given  barometric  pressure,  from 
the  results  o£  experiments  on  the  disruptive  voltages  between 
equal  metal  electrodes  at  given  distances  apart.  They 
calculate  the  maximum  value  o£  the  electric  intensity  be- 
tween the  metal  electrodes  on  the  assumption  that  the 
electric  field  round  them  is  similar  to  that  existing  at  low 
voltages.  Figures  obtained  in  this  way  were  found,  greatly 
to  the  disappointment  of  the  early  experimenters,  to  vary 
widely  with  the  distance  apart  of  the  electrodes.  Lord 
Kelvin,  however,  as  far  back  as  1860  *  deduced  from  the 
results  of  his  experiments  with  large  electrodes  that  it  was 
"  most  probable  "  that  the  numbers  obtained  in  this  way  at 
higher  voltages  would  be  "sensibly  constant/'  An  ex- 
amination of  the  results,  which  are  given  below,  obtained 
recently  by  electricians  will  show  that  experiment  has  amply 
justified  Lord  Kelvin's  conclusion.  The  author  finds,  by 
considering  experimental  results  obtained  both   with  direct 
*  Proc.  Roy.  Soc.  April  12,  1860 ;  '  Reprint/  p.  259. 
