664  Prof,  Fleming  on  Oscillation  Valves. 
o 
oscillations  are  created  by  induction  in  this  circuit,  we  find 
that  the  galvanometer  gives  a  steady  deflexion  showing  the 
passage  of  a  continuous  current  through  it,  and  therefore  o£ 
the  unilateral  conductivity  of  the  space  between  the  glower 
and  the  metal  tube. 
The  distance  over  which  this  transference  of  electricity 
can  take  place  depends  very  much  upon  the  temperature  of 
the  glower,  and  the  amount  of  rectification  of  the  alternating 
current  obtained  upon  success  in  keeping  down  the  tempe- 
rature of  the  metallic  electrode.  This  is  best  achieved  by 
circulating  water  through  it. 
It  follows  as  a  consequence  from  the  above  facts,  that  there 
is  a  considerable  emission  of  negative  ions  or  electrons  from 
the  incandescent  portion  of  the  lime  cylinder  used  with  the 
oxy-coal  gas-burner  to  produce  the  lime-light,  and  that  the 
space  near  the  incandescent  portion  of  the  lime  cylinder  as 
well  as  the  space  near  the  Nernst  lamp-glower  is  highly 
conductive  by  reason  of  the  presence  there  of  negative  ions 
emitted  from  the  oxide  surface. 
In  the  practical  construction  of  oscillation  valves,  the  ad- 
vantage of  placing  the  heated  and  non-heated  electrodes  in  a 
vacuum  is  that  the  plate  which  acts  as  an  anode  can  be  placed 
at  a  greater  distance  from  the  incandescent  surface  and 
thereby  kept  cool,  since  the  electrons  ejected  from  the  heated 
surface  are  projected  to  a  much  greater  distance  when  the 
atmospheric  pressure  is  reduced.  Although  platinum  coated 
with  calcium  or  barium  oxides  undoubtedly  emits  a  much 
larger  electronic  current  per  square  millimetre  than  carbon 
at  the  same  temperature  and  under  the  same  surrounding- 
conditions  as  to  gas  pressure,  I  find  that  for  rectifying 
electric  oscillations  the  carbon-filament  oscillation-valve  as  I 
have  designed  it,  affords  more  conveniently  all  that  is  required. 
There  are  some  advantages  in  employing  a  thick  carbon 
filament  and  constructing  it  to  be  worked  at  about  12  volts 
and  take  a  fairly  large  current  of  2  or  3  amperes.  For  one 
thing,  the  filament  is  much  less  likely  to  be  destroyed  by  over- 
heating in  working,  and  hence  the  valve  lasts  longer.  In 
some  cases  an  advantage  may  ensue  from  working  valves  in 
parallel,  that  is  joining  up  a  number  of  such  carbon-filament 
valves  with  their  carbon  filaments  in  parallel  on  the  same 
heating  battery,  connecting  together  the  insulated  metal 
cylinders  contained  in  each  bulb  together,  and  then  using  the 
multiple  arrangement  as  if  it  were  a  single  valve. 
When  used,  however,  to  rectify  such  oscillations  as  are 
employed  in  the  receiving  circuits  of  wireless  telegraph 
apparatus,  a  single  valve  will  do  all  that  is  required  because 
