ELECTEICAL  DISCHAEGES  AS  OBSEEVED  IN  TOEEICELLIAN  VACUA. 
for  a few  seconds,  the  discharge  being  continued,  the  tube  will  (at  each  place  where  the 
mercury  has  rested)  be  found  coated  with  minute  particles  of  condensed  mercurial 
vapour,  which  clear  off  when  the  mercury  ascends. 
17.  I have  stated  (10.)  that  I had  several  of  the  apparatus  (fig.  1)  constructed,  but  in 
no  two  could  I obtain  precisely  the  same  result : in  some  the  stratification  was  more  or 
less  distinct,  in  others  scarcely  visible,  but  in  all  a residuum  of  air,  more  or  less,  could 
be  detected  when  the  mercury  was  allowed  to  ascend  the  tube ; but  it  appeared  sur- 
prising that  there  should  be  so  marked  a difference  in  the  discharge  when,  as  in  some 
instances,  so  very  minute  a quantity  of  air  (less  than  q th  of  the  contents  of  the  tube) 
was  present. 
Mr.  Casella,  who  had  made  aU  the  glass  apparatus  already  described  (with  the  excep- 
tion of  the  barometer),  placed  one  of  his  most  intelligent  workmen  at  my  disposal ; this 
enabled  me  to  proceed  with  my  experiments,  and  1 had  five  separate  tubes  prepared  in 
the  following  manner ; — three  were  of  the  same  length  and  dimensions  as  the  tube  of 
the  original  apparatus,  Plate  I.  fig.  1 ; two  were  made  of  the  usual  barometer  tubing, 
the  wares  of  the  one  set  being  10  inches,  and  of  the  other  16  inches  apart. 
Each  of  these  tubes  was  filled  with  pure  mercury,  carefully  boiled;  a tube  about 
34  inches  in  length  being  attached  to  each,  also  filled  with  mercury ; the  apparatus  was 
inverted  into  a basin  of  mercury,  thereby  forming  the  usual  barometrical  vacuum,  and 
the  tubes  were  then  sealed  about  4 inches  below  the  lower  platinum  wire. 
18.  Designating  these  tubes  (Plate.  I.  fig.  2)  as  A,  B,  C,  D and  E,  the  stratification 
in  A is  nearly  equal  to  what  I obtain  in  my  best  mercurial  apparatus  (12.).  In  B the 
discharge  is  of  conical  form,  and  the  stratification  visible,  but  confused  and  indistinct. 
In  C the  discharge  is  in  a wa'vy  line,  without  any  stratification.  In  D the  discharge  is 
clear,  bright  and  white,  but  no  appearance  of  stratification.  In  E the  stratification  is 
very  distinct  and  well-defined. 
19.  On  examining  the  apparatus  B,  a minute  bubble  of  mercury  can  be  perceived  in 
the  stem,  and  in  C there  is  evidently  moisture  in  the  lower  part,  below  the  platinum 
wire.  Taking  advantage  of  this  accident,  I inserted  the  stem  of  C in  a freezing  mixture 
of  pounded  ice  and  muriatic  acid.  The  change  in  the  character  of  the  discharge  was 
very  striking ; instead  of  the  straight  luminous  line,  the  discharge  assumed  the  conical 
form,  exhibiting  the  striae  in  a very  much  more  clear  and  distinct  manner  than  in  B, 
though  not  quite  so  good  as  in  A. 
20.  I could  not  but  feel  assured  that  there  must  have  been  something  wrong  in  the 
mode  of  obtaining  vacua  which  exhibited  such  irregularity  of  action,  and  I determined 
on  trying  an  experiment  for  which  I had  long  since  made  preparation,  viz.  to  examine 
whether  any  different  effect  would  take  place  in  a barometrical  vacuum  obtained  by 
the  non-boiling  process  first  suggested  by  Mr.  Welsh,  and  which  he  so  successfully 
carried  out  in  filling  the  large  barometer  of  the  Observatory  at  Kew ; an  apparatus  so 
exhausted  might  also  effectually  test  Mr.  Welsh’s  process  of  filling  a large  barometer, 
the  value  of  which  those  who  are  practically  acquainted  with  the  difficulty  of  boiling  such 
