ELECTEICAL  DISCHAEGES  AS  OBSEEVED  IN  TOEEICELLIAN  VACUA. 
8.  The  conditions  necessary  to  enable  the  experimentalist  to  produce  the  phenomena 
of  the  striee  or  band  discharge,  have  been  stated  by  different  authors  to  be  as  follows : — 
1st,  that  the  vacuum  in  the  receiver  should  be  as  perfect  as  can  be  obtained  by  the  air- 
pump  ; 2nd,  that  care  should  be  taken  to  absorb  all  trace  of  moisture ; and  lastly,  that 
means  should  be  used  to  introduce  the  vapour  of  naphtha  or  phosphorus,  or  other  similar 
bodies. 
9.  The  induction  apparatus  used  in  the  following  experiments  was  constructed  for 
me  by  Mr.  Ladd  ; in  form  it  is  similar  to  that  of  M.  Ruhmkoeff  ; the  primary  coil 
consists  of  three  layers  of  No.  12  copper  wire,  total  length  50  yards ; the  secondary, 
about  three  miles  of  No.  35  covered  with  silk,  each  layer  of  the  secondary  coil  being 
carefully  insulated  by  folds  of  thin  gutta  percha ; with  four  or  five  cells  of  the  usual  size 
nitric  acid  battery,  this  apparatus  gives  a discharge  in  air  of  4^  inches ; but  my  experi- 
ments, except  when  other^vise  described,  were  made  with  only  one  cell,  and  generally 
without  any  condenser ; this  is  separate  from  the  apparatus,  it  consists  of  two  sections, 
one  of  50  and  the  other  of  40  square  feet  of  coated  surface,  which  can  be  used  either 
separately  or  connected. 
10.  While  pursuing  my  experiments,  it  occurred  to  me  that  an  apparatus  similar  in 
some  respects  to  that  used  by  Da\t,  could  without  much  difficulty  be  constructed,  which 
would  enable  me  not  only  to  make  experiments  in  a Torricellian  vacuum,  but  also  with 
great  facility  in  any  gas  which  does  not  act  on  mercury : Plate  I.  fig.  1 represents  this 
apparatus.  In  the  glass  tube,  two  platinum  mres,  a and  5,  are  carefully  sealed  about 
6 inches  apart;  the  tube  is  filled  Avith  pure  mercury.  A stopcock,  fixed  at  C,  can, 
by  means  of  a flexible  tube,  be  connected  with  an  air-pump.  When  the  air  is  extracted 
from  the  ball  of  the  apparatus,  the  mercury  sinks  in  the  tube,  and  in  this  manner  the 
Torricellian  vacuum  is  formed,  the  mercury  in  the  tube  descending  to  “ d” 
11.  In  my  first  experiments  the  discharge  did  not  exhibit  the  uniform  white  light  of 
the  Torricellian  vacuum,  but  striee  in  confused  or  irregular  forms ; and  although  there 
was  not  the  slightest  appearance  of  even  a trace  of  air  in  the  tube  when  first  filled  with 
mercury,  a minute  bubble  constantly  appeared  after  the  discharge  had  been  continued 
for  three  or  four  minutes,  and  the  mercury  was  allowed  to  rise  in  the  tube.  On 
incautiously  letting  in  the  mercury  after  the  lower  wire  had  been  heated  by  the  negative 
discharge,  the  tube  broke. 
12.  This  experiment  having  convinced  me  that  under  some  conditions  the  stratification 
could  be  produced  in  a Tonicellian  vacuum,  I ordered  other  similar  apparatus  to  be 
made,  one  of  which  I directed  should  not  only  be  very  carefully  filled,  but  that  the 
mercury  should  be  boiled  in  the  tube.  The  effect  I obtained  fully  compensated  me  for 
the  trouble  I had  taken.  In  this  apparatus  the  discharge  from  the  coil,  when  excited 
by  a single  cell  of  Grove’s  battery,  the  upper  wire  being  negative,  consisted  of  eight  or 
ten  distinct  stratifications,  extending  from  the  positive  wire  to  the  dark  space,  while  the 
usual  blue  flame  surrounding  the  intense  red,  which  has  the  appearance  of  red  heat,  is 
visible  on  the  negative  wire.  On  reversing  the  direction  of  the  primary  current  by  the 
B 2 
