Thomson 1904: Difference between revisions

From Mass Spec Terms
← Older editNewer edit →
Line 93: Line 93:
THE    ATOMIC    STRUCTURE    OF    ELECTRICITY
THE    ATOMIC    STRUCTURE    OF    ELECTRICITY


HITHERTO  we  have  been  dealing  chiefly  with  the properties  of  the  lines  of  force,  with  their  ten- sion, the  mass  of  ether  they  carry  along  with  them, and  with  the  propagation  of  electric  disturbances along  them ;  in  this  chapter  we  shall  discuss  the nature  of  the  charges  of  electricity  which  form  the beginnings  and  ends  of  these  lines.  We  shall  show that  there  are  strong  reasons  for  supposing  that these  changes  have  what  may  be  called  an  atomic structure ;  any  charge  being  built  up  of  a  number of  finite  individual  charges,  all  equal  to  each  other : just  as  on  the  atomic  theory  of  matter  a  quantity of  hydrogen  is  built  up  of  a  number  of  small  par- ticles called  atoms,  all  the  atoms  being  equal  to each  other.  If  this  view  of  the  structure  of  elec- tricity is  correct,  each  extremity  of  a  Faraday  tube will  be  the  place  from  which  a  constant  fixed  num- ber of  tubes  start  or  at  which  they  arrive.
[[Thomson 1904/Chapter 4]]
 
Let  us  first  consider  the  evidence  given  by  the laws  of  the  electrolysis  of  liquids.  Faraday
 
 
y2  ELECTRICITY    AND  MATTER
 
showed  that  when  electricity  passes  through  a liquid  electrolyte,  the  amount  of  negative  electric- ity given  up  to  the  positive  electrode,  and  of  posi- tive electricity  given  to  the  negative  electrode,  is proportional  to  the  number  of  atoms  coming  up to  the  electrode.  Let  us  first  consider  monovalent elements,  such  as  hydrogen,  chlorine,  sodium,  and so  on ;  he  showed  that  when  the  same  number  of atoms  of  these  substances  deliver  up  their charges  to  the  electrode,  the  quantity  of  electricity communicated  is  the  same  whether  the  carriers are  atoms  of  hydrogen,  chlorine,  or  sodium,  in- dicating that  each  atom  of  these  elements  carries the  same  charge  of  electricity.  Let  us  now  go  to the  divalent  elements.  We  find  again  that  the  ions of  all  divalent  elements  carry  the  same  charge, but  that  a  number  of  ions  of  the  divalent  ele- ment carry  twice  the  charge  carried  by  the  same number  of  ions  of  a  univalent  element,  showing that  each  ion  of  a  divalent  element  carries  twice as  much  charge  as  the  univalent  ion ;  again,  a  tri- valent  ion  carries  three  times  the  charge  of  a  uni- valent ion,  and  so  on.  Thus,  in  the  case  of  the  elec- trolysis of  solutions  the  charges  carried  by  the ions  are  either  the  charge  on  the  hydrogen  ion  or twice  that  charge,  or  three  times  the  charge,  and
 
 
THE  ATOMIC    STRUCTURE  OF  ELECTRICITY        73
 
so  on.  The  charges  we  meet  with  are  always  an integral  multiple  of  the  charge  carried  by  the  hy- drogen atom  ;  we  never  meet  with  fractional  parts of  this  charge.  This  very  remarkable  fact  shows, as  Helmholtz  said  in  the  Faraday  lecture,  that "  if  we  accept  the  hypothesis  that  the  elementary substances  are  composed  of  atoms,  we  cannot  avoid the  conclusion  that  electricity,  positive  as  well  as negative,  is  divided  into  definite  elementary  por- tions which  behave  like  atoms  of  electricity."
 
When  we  consider  the  conduction  of  electricity through  gases,  the  evidence  in  favor  of  the  atomic character  of  electricity  is  even  stronger  than  it  is in  the  case  of  conduction  through  liquids,  chiefly because  we  know  more  about  the  passage  of  elec- tricity through  gases  than  through  liquids.
 
Let  us  consider  for  a  moment  a  few  of  the  prop- erties of  gaseous  conduction.  When  a  gas  has been  put  into  the  conducting  state — say,  by  expos- ure to  Rontgen  rays — it  remains  in  this  state  for a  sufficiently  long  time  after  the  rays  have  ceased to  enable  us  to  study  its  properties.  We  find  that we  can  filter  the  conductivity  out  of  the  gas  by sending  the  gas  through  a  plug  of  cotton- wool,  or through  a  water-trap.  Thus,  the  conductivity  is due  to  something  mixed  with  the  gas  which  can
 
 
74  ELECTRICITY  AND  MATTER
 
be  filtered  out  of  it;  again,  the  conductivity  is taken  out  of  the  gas  when  it  is  sent  through  a strong  electric  field.  This  result  shows  that  the constituent  to  which  the  conductivity  of  the  gas  is due  consists  of  charged  particles,  the  conductivity arising  from  the  motion  of  these  particles  in  the electric  field.  We  have  at  the  Cavendish  Labora- tory measured  the  charge  of  electricity  carried  by those  particles.
 
The  principle  of  the  method  first  used  is  as  fol- lows. If  at  any  time  there  are  in  the  gas  n  of  these particles  charged  positively  and  n  charged  nega- tively, and  if  each  of  these  carries  an  electric  charge e,  we  can  easily  by  electrical  methods  determine  n  e, the  quantity  of  electricity  of  our  sign  present  in the  gas.  One  method  by  which  this  can  be  done is  to  enclose  the  gas  between  two  parallel  metal plates,  one  of  which  is  insulated.  Now  suppose  we suddenly  charge  up  the  other  plate  positively  to  a very  high  potential,  this  plate  will  now  repel  the positive  particles  in  the  gas,  and  these  before  they have  time  to  combine  with  the  negative  particles will  be  driven  against  the  insulated  plate.  Thus, all  the  positive  charge  in  the  gas  will  be  driven against  the  insulated  plate,  where  it  can  be  meas- ured by  an  electrometer.  As  this  charge  is  equal
 
 
THE  ATOMIC  STRUCTURE  OF  ELECTRICITY  75
 
to  n  e  we  can  in  this  way  easily  determine  n  e :  if then  we  can  devise  a  means  of  measuring  n  we shall  be  able  to  find  e.  The  method  by  which  I determined  n  was  founded  on  the  discovery  by C.  T.  E.  Wilson  that  the  charged  particles  act  as nuclei  round  which  small  drops  of  water  condense, when  the  particles  are  surrounded  by  damp  air cooled  below  the  saturation  point.  In  dust-free air,  as  Aitken  showed,  it  is  very  difficult  to  get  a fog  when  damp  air  is  cooled,  since  there  are  no nuclei  for  the  drops  to  condense  around ;  if  there are  charged  particles  in  the  dust-free  air,  however, a  fog  will  be  deposited  round  these  by  a  supersat- uration  far  less  than  that  required  to  produce  any appreciable  effect  when  no  charged  particles  are present.
 
Thus,  in  sufficiently  supersaturated  damp  air  a cloud  is  deposited  on  these  charged  particles, and  they  are  thus  rendered  visible.  This  is  the first  step  toward  counting  them.  The  drops  are, however,  far  too  small  and  too  numerous  to  be counted  directly.  We  can,  however,  get  their number  indirectly  as  follows :  suppose  we  have  a number  of  these  particles  in  dust-free  air  in  a closed  vessel,  the  air  being  saturated  with  water vapor,  suppose  now  that  we  produce  a  sudden
 
 
76  ELECTRICITY  AND  MATTER
 
expansion  of  the  air  in  the  vessel;  this  will  cool the  air,  it  will  be  supersaturated  with  vapor,  and drops  will  be  deposited  round  the  charged  parti- cles. Now  if  we  know  the  amount  of  expansion produced  we  can  calculate  the  cooling  of  the  gas, and  therefore  the  amount  of  water  deposited. Thus,  we  know  the  volume  of  water  in  the  form  of drops,  so  that  if  we  know  the  volume  of  one  drop we  can  deduce  the  number  of  drops.  To  find  the size  of  a  drop  we  make  use  of  an  investigation by  Sir  George  Stokes  on  the  rate  at  which  small spheres  fall  through  the  air.  In  consequence  of the  viscosity  of  the  air  small  bodies  fall  exceedingly slowly,  and  the  smaller  they  are  the  slower  they fall.  Stokes  showed  that  if  a  is  the  radius  of  a drop  of  water,  the  velocity  v  with  which  it  falls through  the  air  is  given  by  the  equation
 
 
when  g  is  the  acceleration  due  to  gravity  =  981 and  p.  the  coefficient  of  viscosity  of  air  =  .00018; thus
 
v  =  1.21  X  106a2;
 
hence  if  we  can  determine  v  we  can  determine the  radius  and  hence  the  volume  of  the  drop.
 
 
THE  ATOMIC  STRUCTURE  OF  ELECTRICITY  77
 
But  v  is  evidently  the  velocity  with  which  the cloud  round  the  charged  particle  settles  down,  and can  easily  be  measured  by  observing  the  move- ment of  the  top  of  the  cloud.  In  this  way  I found  the  volume  of  the  drops,  and  thence  n  the number  of  particles.  As  n  e  had  been  determined by  electrical  measurements,  the  value  of  e  could be  deduced  when  n  was  known ;  in  this  way  I found  that  its  value  is
 
3.4  X  10-10  Electrostatic  C.  G.  S.  units.
 
Experiments  were  made  with  air,  hydrogen, and  carbonic  acid,  and  it  was  found  that  the ions  had  the  same  charge  in  all  these  gases;  a strong  argument  in  favor  of  the  atomic  character of  electricity.
 
We  can  compare  the  charge  on  the  gaseous  ion with  that  carried  by  the  hydrogen  ion  in  the  elec- trolysis of  solutions  in  the  following  way:  We know  that  the  passage  of  one  electro-magnetic unit  of  electric  charge,  or  3  X  1010  electrostatic units,  through  acidulated  water  liberates  1.23  c.c. of  hydrogen  at  the  temperature  15°C.  and  pressure of  one  atmosphere ;  if  there  are  N  molecules  in  a c.c.  of  a  gas  at  this  temperature  and  pressure  the number  of  hydrogen  ions  in  1.23  c.c.  is  2.46  N,
 
 
7g  ELECTRICITY  AND  MATTER
 
so  that  if  E  is  the  charge  on  the  hydrogen  ion  in the  electrolyis  of  solution,
 
2.46  NE=  3X  1010, or    E=  1.22  X  1010-f-JV:
 
Now,  e,  the  charge  on  the  gas  ion  is  3.4  X10"10, hence  if  ^Y=3.6X1019  the  charge  on  the  gaseous ion  will  equal  the  charge  on  the  electrolytic  ion. Now,  in  the  kinetic  theory  of  gases  methods  are investigated  for  determining  this  quantity  N,  or Avogadro's  Constant,  as  it  is  sometimes  called ;  the values  obtained  by  this  theory  vary  somewhat with  the  assumptions  made  as  to  the  nature  of the  molecule  and  the  nature  of  the  forces  which  one molecule  exerts  on  another  in  its  near  neighbor- hood. The  value  3.6 X1019  is,  however,  in  good agreement  with  some  of  the  best  of  these  deter- minations, and  hence  we  conclude  that  the  charge on  the  gaseous  ion  is  equal  to  the  charge  on  the electrolytic  ion.
 
Dr.  H.  A.  Wilson,  of  the  Cavendish  Laboratory, by  quite  a  different  method,  obtained  practically the  same  value  for  e  as  that  given  above.  His method  was  founded  on  the  discovery  by  C.  T.  R. Wilson  that  it  requires  less  supersaturation  to deposit  clouds  from  moist  air  on  negative  ions
 
 
THE  ATOMIC    STRUCTURE  OF    ELECTRICITY        79
 
than  it  does  on  positive.  Thus,  by  suitably choosing  the  supersaturation,  we  can  get  the cloud  deposited  on  the  negative  ions  alone,  so that  each  drop  in  the  cloud  is  negatively  charged ; by  observing  the  rate  at  which  the  cloud  falls  we can,  as  explained  above,  determine  the  weight  of each  drop.  Now,  suppose  we  place  above  the cloud  a  positively  electrified  plate,  the  plate  will attract  the  cloud,  and  we  can  adjust  the  charge  on the  plate  until  the  electric  attraction  just  balances the  weight  of  a  drop,  and  the  drops,  like  Mahomet's coffin,  hang  stationary  in  the  air;  if  X  is  the electric  force  then  the  electric  attraction  on  the drop  is  Xe,  when  e  is  the  charge  on  the  drop.  As X  e  is  equal  to  the  weight  of  the  drop  which  is known,  and  as  we  can  measure  X,  e  can  be  at  once determined.
 
Townsend  showed  that  the  charge  on  the gaseous  ion  is  equal  to  that  on  the  ion  of  hydrogen in  ordinary  electrolysis,  by  measuring  the  coeffi- cient of  diffusion  of  the  gaseous  ions  and  com- paring it  with  the  velocity  acquired  by  the  ion under  a  given  electric  force.  Let  us  consider  the case  of  a  volume  of  ionized  gas  between  two horizontal  planes,  and  suppose  that  as  long  as  we keep  in  any  horizontal  layer  the  number  of  ions
 
 
8Q  ELECTRICITY    AND  MATTER
 
remains  the  same,  but  that  the  number  varies  as we  pass  from  one  layer  to  another;  let  x  be  the distance  of  a  layer  from  the  lower  plane,  n the  number  of  ions  of  one  sign  in  unit  volume of  this  layer,  then  if  D  be  the  coefficient  of diffusion  of  the  ions,  the  number  of  ions  which in  one  second  pass  downward  through  unit  area
 
of  the  layer  is
 
d  n
 
»ir*'
 
so  that  the  average  velocity  of  the  particles  down- ward is
 
l)dn
 
n  dx
 
The  force  which  sets  the  ions  in  motion  is  the variation  in  the  partial  pressure  due  to  the  ions  ;  if this  pressure  is  equal  top,  the  force  acting  on  the
 
ions  in  a  unit  volume  is  T->  and  the  average  force
 
per  ion  is  -  -r— .    Now  we  can  find  the  velocity ndx
 
which  an  ion  acquires  when  acted  upon  by  a known  force  by  measuring,  as  Rutherford  and Zeleny  have  done,  the  velocities  acquired  by  the ions  in  an  electric  field.  They  showed  that  this velocity  is  proportional  to  the  force  acting  on  the ion,  so  that  if  A.  is  the  velocity  when  the  electric
 
 
THE  ATOMIC  STRUCTURE  OF  ELECTRICITY  81
 
force  is  JTand  when  the  force  acting  on  the  ion  is therefore  X  e,  the  velocity  for  unit  force  will  be
 
-==r-,  and  the  velocity  when  the  force  is  1  ^  will .A  e  n  ax
 
therefore  be
 
i  dp  ^L.
 
n  dx  Xe'
 
this  velocity  we  have  seen,  however,  to  be  equal  to D  dnf n  dx' hence  we  have
 
dP    A  --T)  dn  (l\
 
~7  -        ~Tr~  -  •*/      "7  —    •  V1/
 
dx  Xe  dx
 
Now  if  the  ions  behave  like  a  perfect  gas,  the pressure  p  bears  a  constant  ratio  to  n,  the  number of  ions  per  unit  volume.  This  ratio  is  the  same for  all  gases,  at  the  same  temperature,  so  that  if jV  is  Avogadro's  constant,  i.e.,  the  number  of  mol- ecules in  a  cubic  centimetre  of  gas  at  the  atmos- pheric pressure  P
 
p  _    n P~W
 
and  equation  (1)  gives  us PA
 
 
Thus,  by  knowing  D  and  a  we  can  find  the  value of  Ne.  In  this  way  Townsend  found  that  Ne  was
 
 
g2  ELECTRICITY  AND  MATTER
 
the  same  in  air,  hydrogen,  oxygen,  and  carbonic acid,  and  the  mean  of  his  values  was  N  e  =  1.24 X  1010.  We  have  seen  that  if  E  is  the  charge  on the  hydrogen  ion
 
NE  =  1.22  X  1010-
 
Thus,  these  experiments  show  that  e—E,  or  that the  charge  on  the  gaseous  ion  is  equal  to  the charge  carried  by  the  hydrogen  ion  in  the  elec- trolysis of  solutions.
 
The  equality  of  these  charges  has  also  been proved  in  a  very  simple  way  by  H.  A.  Wilson, who  introduced  per  second  into  a  volume  of  air at  a  very  high  temperature,  a  measured  quan- tity of  the  vapor  of  metallic  salts.  This  vapor got  ionized  and  the  mixture  of  air  and  vapor  ac- quired very  considerable  conductivity.  The  cur- rent through  the  vapor  increased  at  first  with  the electromotive  force  used  to  drive  it  through  the gas,  but  this  increase  did  not  go  on  indefinitely, for  after  the  current  had  reached  a  certain  value no  further  increase  in  the  electromotive  force  pro- duced any  change  in  the  current.  The  current,  as in  all  cases  of  conduction  through  gases,  attained a  maximum  value  called  the  "  saturation  current," which  was  not  exceeded  until  the  electric  field  ap- plied to  the  gas  approached  the  intensity  at  which
 
 
THE    ATOMIC    STRUCTURE  OF  ELECTRICITY        33
 
sparks  began  to  pass  through  the  gas.  Wilson found  that  the  saturation  current  through  the  salt vapor  was  just  equal  to  the  current  which  if  it passed  through  an  aqueous  solution  of  the  salt would  electrolyse  in  one  second  the  same  amount of  salt  as  was  fed  per  second  into  the  hot  air.
 
It  is  worth  pointing  out  that  this  result  gives us  a  method  of  determining  Avogadro's  Constant which  is  independent  of  any  hypothesis  as  to  the shape  or  size  of  molecules,  or  of  the  way  in  which they  act  upon  each  other.  If  .N"  is  this  constant,  e the  charge  on  an  ion,  then  JV  e  —  1.22  X  1010  and we  have  seen  that  e  =  3.4 X  l^10,  so  that  jV=  3.fc X  1019.
 
Thus,  whether  we  study  the  conduction  of  elec- tricity through  liquids  or  through  gases,  we  are led  to  the  conception  of  a  natural  unit  or  atom  of electricity  of  which  all  charges  are  integral  mul- tiples, just  as  the  mass  of  a  quantity  of  hydro- gen is  an  integral  multiple  of  the  mass  of  a hydrogen  atom.
 
Mass  of  the  Carriers  of  Electricity
 
We  must  now  pass  on  to  consider  the  nature of  the  systems  which  carry  the  charges,  and  in order  to  have  the  conditions  as  simple  as  possible
 
 
g4  ELECTRICITY    AND    MATTER
 
let  us  begin  with  the  case  of  a  gas  at  a  very  low pressure,  where  the  motion  of  the  particles  is  not impeded  by  collisions  with  the  molecules  of  the gas.  Let  us  suppose  that  we  have  a  particle  of mass  m,  carrying  a  charge  e,  moving  in  the  plane of  the  paper,  and  that  it  is  acted  on  by  a  uniform magnetic  field  at  right  angles  to  this  plane.  We have  seen  that  under  these  circumstances  the  par- ticle will  be  acted  upon  by  a  mechanical  force equal  to  He  v,  where  His  the  magnetic  force  and v  the  velocity  of  the  particle.  The  direction  of this  force  is  in  the  plane  of  the  paper  at  right angles  to  the  path  of  the  particle.  Since  the force  is  always  at  right  angles  to  the  direction of  motion  of  the  particle,  the  velocity  of  the  par- ticle and  therefore  the  magnitude  of  the  force  act- ing upon  it  will  not  alter,  so  that  the  path  of  the particle  will  be  that  described  by  a  body  acted upon  by  a  constant  normal  force.  It  is  easy  to  show that  this  path  is  a  circle  whose  radius  a  is  given
 
by  the  equation
 
mv  ,tv
 
 
The  velocity  v  of  the  particle  may  be  deter- mined by  the  following  method  :  Suppose  the  par- ticle is  moving  horizontally  in  the  plane  of  the
 
 
THE    ATOMIC    STRUCTURE    OF    ELECTRICITY        35
 
paper,  through  a  uniform  magnetic  field  If,  at right  angles  to  this  plane,  the  particle  will  be acted  upon  by  a  vertical  force  equal  to  H  e  v.  Now, if  in  addition  to  the  magnetic  force  we  apply  a vertical  electric  force  X,  this  will  exert  a  verti- cal mechanical  force  X  e  on  the  moving  particle. Let  us  arrange  the  direction  of  JTso  that  this  force is  in  the  opposite  direction  to  that  due  to  the  mag- net, and  adjust  the  value  of  X  until  the  two  forces are  equal.  We  can  tell  when  this  adjustment  has been  made,  since  in  this  case  the  motion  of  the particle  under  the  action  of  the  electric  and  mag- netic forces  will  be  the  same  as  when  both  these forces  are  absent.  When  the  two  forces  are  equal we  have
 
X  e  —  H  e  v,      or
 
 
Hence  if  we  have  methods  of  tracing  the  motion of  the  particle,  we  can  measure  the  radius  a  of  the circle  into  which  it  is  bent  by  a  constant  magnetic force,  and  determine  the  value  of  the  electric  force required  to  counteract  the  effect  of  the  magnetic force.  Equations  (1)  and  (2)  then  give  us  the
 
means  of  finding  both  v  and  —  •
 
 
gg  ELECTRICITY  AND  MATTER
 
Values  of  —  for  Negatively  Electrified  Particles
 
in  Gases  at  Low  Pressures
 
The  value  of  e.  has  been  determined  in  this m
 
way  for  the  negatively  electrified  particles  which form  the  cathode  rays  which  are  so  conspicuous  a part  of  the  electric  discharge  through  a  gas  at  low pressures  ;  and  also  for  the  negatively  electrified particles  emitted  by  metals,  (1)  when  exposed  to ultra-violet  light,  (2)  when  raised  to  the  tempera- ture of  incandescence.  These  experiments  have led  to  the  very  remarkable  result  that  the  value  of
 
—  is  the  same  whatever  the  nature  of  the  gas  in
 
171
 
which  the  particle  may  be  found,  or  whatever  the nature  of  the  metal  from  which  it  may  be  sup- posed to  have  proceeded.  In  fact,  in  every  case
 
p
 
in  which  the  value  of  —  has  been  determined  for m
 
negatively  electrified  particles  moving  with  veloci- ties considerably  less  than  the  velocity  of  light,  it has  been  found  to  have  the  constant  value  about 10T,  the  units  being  the  centimetre,  gram,  and second,  and  the  charge  being  measured  in  electro- magnetic units.  As  the  value  of  —  for  the  hydro-
 
 
THE  ATOMIC  STRUCTURE  OF  ELECTRICITY  37
 
gen  ion  in  the  electrolysis  of  liquids  is  only  104, and  as  we  have  seen  the  charge  on  the  gaseous ion  is  equal  to  that  on  the  hydrogen  ion  in  ordi- nary electrolysis,  we  see  that  the  mass  of  a  carrier of  the  negative  charge  must  be  only  about  one- thousandth  part  of  the  mass  of  hydrogen  atom ; the  mass  was  for  a  long  time  regarded  as  the  small- est mass  able  to  have  an  independent  existence.
 
I  have  proposed  the  name  corpuscle  for  these units  of  negative  electricity.  These  corpuscles are  the  same  however  the  electrification  may  have arisen  or  wherever  they  may  be  found.  Negative electricity,  in  a  gas  at  a  low  pressure,  has  thus  a structure  analogous  to  that  of  a  gas,  the  corpuscles taking  the  place  of  the  molecules.  The  "  negative electric  fluid,"  to  use  the  old  notation,  resembles a  gaseous  fluid  with  a  corpuscular  instead  of  a molecular  structure.
 
Carriers  of  Positive  Electrification We  can  apply  the  same  methods  to  determine
 
the  values  of  —  for  the  carriers  of  positive  electri- fy
 
fication.  This  has  been  done  by  Wien  for  the positive  electrification  found  in  certain  parts  of the  discharge  in  a  vacuum  tube,  and  I  have
 
 
CQ  ELECTRICITY  AND  MATTER
 
oo
 
measured  —  for  the  positive  electrification  given
 
off  by  a  hot  wire.  The  results  of  these  measure- ments form  a  great  contrast  to  those  for  the  nega- tive electrification,  for  —  for  the  positive  charge,
 
instead  of  having,  as  it  has  for  the  negative,  the constant  high  value  107,  is  found  never  to  have  a value  greater  than  104,  the  value  it  would  have  if the  carrier  were  the  atom  of  hydrogen.  In  many
 
cases  the  value  of  —  is  very  much  less  than  104,  in- m
 
dicating  that  in  these  cases  the  positive  charge  is carried  by  atoms  having  a  greater  mass  than  that
 
^ of  the  hydrogen  atom.  The  value  of  —  varies  with
 
the  nature  of  the  electrodes  and  with  the  gas  in the  discharge  tube,  just  as  it  would  if  the  carriers of  the  positive  charge  were  the  atoms  of  the elements  which  happened  to  be  present  when  the positive  electrification  was  produced.
 
These  results  lead  us  to  a  view  of  electrification which  has  a  striking  resemblance  to  that  of Franklin's  "One  Fluid  Theory  of  Electricity." Instead  of  taking,  as  Franklin  did,  the  electric fluid  to  be  positive  electricity  we  take  it  to  be negative.  The  "electric  fluid"  of  Franklin  cor-
 
 
TI1K    ATOMIC    STRUCTURE    OF    ELECTRICITY        §9
 
responds  to  an  assemblage  of  corpuscles,  negative electrification  being  a  collection  of  these  corpus- cles. The  transference  of  electrification  from  one place  to  another  is  effected  by  the  motion  of  cor- puscles from  the  place  where  is  a  gain  of  positive electrification  to  the  place  where  there  is  a  gain of  negative.  A  positively  electrified  body  is  one that  has  lost  some  of  its  corpuscles.  We  have seen  that  the  mass  and  charge  of  the  corpuscles have  been  determined  directly  by  experiment.  We in  fact  know  more  about  the  "  electric  fluid  "  than we  know  about  such  fluids  as  air  or  water.
 


==CHAPTER  V==
==CHAPTER  V==

Revision as of 21:43, 11 August 2025

ELECTRICITY AND MATTER


"

J. J. THOMSON, D.Sc., LL.D., PH.D., F.R.S.

""FELLOW OF TRINITY COLLEGE, CAMBRIDGE; CAVENDISH PROFESSOR OF EXPERIMENTAL PHYSICS, CAMBRIDGE


WITH DIAGRAMS


NEW YORK

CHARLES SCRIBNER'S SONS 1904


COPYRIGHT, 1904 BY YALE UNIVERSITY

Published, March, 1904


THE SILLIMAN FOUNDATION.

In the year 1883 a legacy of eighty thousand dollars was left to the President and Fellows of Yale College in the city of New Haven, to be held in trust, as a gift from her children, in memory of their beloved and honored mother Mrs. Hepsa Ely Silliinan.

On this foundation Yale College was requested and directed to establish an annual course of lectures de- signed to illustrate the presence and providence, the wisdom and goodness of God, as manifested in the natural and moral world. These were to be designated as the Mrs. Hepsa Ely Silliinan Memorial Lectures. It was the belief of the testator that any orderly presenta- tion of the facts of nature or history contributed to the end of this foundation more effectively than any attempt to emphasize the elements of doctrine or of creed; and he therefore provided that lectures on dog- matic or polemical theology should be excluded from the scope of this foundation, and that the subjects should be selected rather from the domains of natural science and history, giving special prominence to astronomy, chemistry, geology, and anatomy.

It was further directed that each annual course should be made the basis of a volume to form part of a series constituting a memorial to Mrs. Sillimau. The memo- rial fund came into the possession of the Corporation of Yale University in the year 1902; and the present volume constitutes the first of the series of memorial lectures.


PREFACE

In these Lectures given at Yale University in May, 1903, I have attempted to discuss the bear- ing of the recent advances made in Electrical Science on our views of the Constitution of Matter and the Nature of Electricity; two questions which are probably so intimately connected, that the solution of the one would supply that of the other. A characteristic feature of recent Electri- cal Researches, such as the study and discovery of Cathode and Rontgen Rays and Radio-active Substances, has been the very especial degree in which they have involved the relation between Matter and Electricity.

In choosing a subject for the Silliman Lectures, it seemed to me that a consideration of the bear- ing of recent work on this relationship might be suitable, especially as such a discussion suggests multitudes of questions which would furnish ad- mirable subjects for further investigation by some of my hearers.

Cambridge, Aug., 1903.

J. J. THOMSON.


CONTENTS

CHAPTER- I

CHE ELECTRIC OF FORCE 1


PAGE

REPRESENTATION OF THE ELECTRIC FIELD BY LINES


CHAPTER II ELECTRICAL AND BOUND MASS 30

CHAPTER III

EFFECTS DUE TO THE ACCELERATION OF FARADAY TUBES 68

CHAPTER IV THE ATOMIC STRUCTURE OF ELECTRICITY ... 71

CHAPTER V

THK CONSTITUTION OF THE ATOM 90

CHAPTER VI

llAUIO-ACTIVITY AND RADIO-ACTIVE SUBSTANCES . . 140


ELECTRICITY AND MATTER

CHAPTER I

Thomson 1904/Chapter 1

CHAPTER II

ELECTRICAL AND BOUND MASS.

Thomson 1904/Chapter 2

CHAPTER III

EFFECTS DUE TO ACCELERATION OF THE FARADAY TUBES

Thomson 1904/Chapter 3

CHAPTER IV

THE ATOMIC STRUCTURE OF ELECTRICITY

Thomson 1904/Chapter 4

CHAPTER V

CONSTITUTION OF THE ATOM

WE have seen that whether we produce the corpuscles by cathode rays, by ultra-violet light, or from incandescent metals, and whatever may be the metals or gases present we always get the same kind of corpuscles. Since corpuscles similar in all respects may be obtained from very different agents and materials, and since the mass of the corpuscles is less than that of any known atom, we see that the corpuscle must be a constituent of the atom of many different substances. That in fact the atoms of these substances have something in common.

We are thus confronted with the idea that the atoms of the chemical elements are built up of sim- pler systems ; an idea which in various forms has been advanced by more than one chemist. Thus Prout, in 1815, put forward the view that the atoms of all the chemical elements are built up of atoms of hydrogen ; if this were so the combining weights of all the elements would, on the assump-


CONSTITUTION OF THE ATOM 91

tion that there was no loss of weight when the atoms of hydrogen combined to form the atom of some other element, be integers ; a result not in ac- cordance with observation. To avoid this discrep- ancy Dumas suggested that the primordial atom might not be the hydrogen atom, but a smaller atom having only one-half or one-quarter of the mass of the hydrogen atom. Further support was given to the idea of the complex nature of the atom by the discovery by Newlands and Mende- leeff of what is known as the periodic law, which shows that there is a periodicity in the properties of the elements when they are arranged in the or- der of increasing atomic weights. The simple rela- tions which exist between the combining weights of several of the elements having similar chemical properties, for example, the fact that the combin- ing weight of sodium is the arithmetic mean of those of lithium and potassium, all point to the conclusion that the atoms of the different elements have something in common. Further evidence in the same direction is afforded by the similarity in the structure of the spectra of elements in the same group in the periodic series, a similarity which re- cent work on the existence in spectra of series of lines whose frequencies are connected by definite


g2 ELECTRICITY AND MATTER

numerical relations has done much to emphasize and establish ; indeed spectroscopic evidence alone has led Sir Norman Lockyer for a long time to advocate the view that the elements are really compounds which can be dissociated when the circumstances are suitable. The phenomenon of radio-activity, of which I shall have to speak later, carries the argument still further, for there seems good reasons for believing that radio-activity is due to changes going on within the atoms of the radio-active substances. If this is so then we must face the problem of the constitution of the atom, and see if we can imagine a model which has in it the potentiality of explaining the re- markable properties shown by radio-active sub- stances. It may thus not be superfluous to con- sider the bearing of the existence of corpuscles on the problem of the constitution of the atom ; and although the model of the atom to which we are led by these considerations is very crude and im- perfect, it may perhaps be of service by suggesting lines of investigations likely to furnish us with fur- ther information about the constitution of the atom.


CONSTITUTION OF THE ATOM 93

The Nature of the Unit from which the Atoms are Built Up

Starting from the hypothesis that the atom is an aggregation of a number of simpler systems, let us consider what is the nature of one of these systems. We have seen that the cor- puscle, whose mass is so much less than that of the atom, is a constituent of the atom, it is natural to regard the corpuscle as a constituent of the primor- dial system. The corpuscle, however, carries a definite charge of negative electricity, and since with any charge of electricity we always associate an equal charge of the opposite kind, we should expect the negative charge on the corpuscle to be associated with an equal charge of positive electri- city. Let us then take as our primordial system an electrical doublet, with a negative corpuscle at one end and an equal positive charge at the other, the two ends being connected by lines of electric force which we suppose to have a material existence. For reasons which will appear later on, we shall suppose that the volume over which the positive electricity is spread is very much larger than the volume of the corpuscle. The lines of force will therefore be very much more condensed near the


94


ELECTRICITY AND MATTER


corpuscle than at any other part of the system, and therefore the quantity of ether bound by the lines of force, the mass of which we regard as the mass of the system, will be very much greater near the corpuscle than elsewhere. If, as we have sup- posed, the size of the corpuscle is very small com- pared with the size of the volume occupied by the positive electrification, the mass of the system will practically arise from the mass of bound ether close to the corpuscle ; thus the mass of the sys- tem will be practically independent of the position of its positive end, and will be very approximately the mass of the corpuscles if alone in the field. This mass (see page 21) is for each corpuscle

equal to — , where e is the charge on the corpuscle

and a its radius — a, as we have seen, being about 10-13 cm.

Now suppose we had a universe consisting of an immense number of these electrical doublets, which we regard as our primordial system  ; if these were at rest their mutual attraction would draw them together, just as the attractions of a lot of little magnets would draw them together if they were free to move, and aggregations of more than one system would be formed.


CONSTITUTION OF THE ATOM 95

If, however, the individual systems were orig- inally moving with considerable velocities, the rel- ative velocity of two systems, when they came near enough to exercise appreciable attraction on each other, might be sufficient to carry the sys- tems apart in spite of their mutual attraction. In this case the formation of aggregates would be postponed, until the kinetic energy of the units had fallen so low that when they came into collision, the tendency to separate due to their relative motion was not sufficient to prevent them remaining together under their mutual attraction.

Let us consider for a moment the way in which the kinetic energy of such an assemblage of units would diminish. We have seen (p. 68) that when- ever the velocity of a charged body is changing the body is losing energy, since it generates electrical waves which radiate through space, car- rying energy with them. Thus, whenever the units come into collision, i.e., whenever they come so close together that they sensibly acceler- ate or retard each other's motion, energy will be radiated away, the whole of which will not be absorbed by the surrounding units. There will thus be a steady loss of kinetic energy, and after a time, although it may be a very long time, the


gg ELECTRICITY AND MATTER

kinetic energy will fall to the value at which aggre- gation of the units into groups of two will begin ; these will later on be followed by the formation of aggregates containing a larger number of units. In considering the question of the further ag- gregation of these complex groups, we must re- member that the possibility of aggregation will depend not merely upon the velocity of the aggre- gate as a whole, i.e., upon the velocity of the centre of gravity, but also upon the relative ve- locities of the corpuscles within the aggregate.

Let us picture to ourselves the aggregate as, like the ^Epinus atom of Lord Kelvin, consisting of a sphere of uniform positive electrification, and ex- erting therefore a radial electric force proportional at an internal point to the distance from the centre, and that the very much smaller negatively electri- fied corpuscles are moving about inside it. The number of corpus- cles is the number of units which had gone to make up the aggre- gate, and the total negative elec- trification on the corpuscles is

Fio. 15

equal to the positive electrifica- tion on the sphere. To fix our ideas let us take the case shown in Fig. 15 of three corpuscles


CONSTITUTION OF THE ATOM 97

A, B, C, arranged within the sphere at the corners of an equilateral triangle, the centre of the triangle coinciding with the centre of the sphere. First suppose the corpuscles are at rest ; they will be in equilibrium when they are at such a distance from the centre of the sphere that the repulsion between the corpuscles, which will evidently be radial, just balances the radial attraction excited on the corpuscles by the positive electrification of the sphere. A simple calculation shows that this will be the case when the distance of the corpuscle from the centre is equal to .57 times the radius of the sphere. Next suppose that the corpuscles, in- stead of being at rest, are describing circular orbits round the centre of the sphere. Their centrifugal force will carry them farther away from the centre by an amount depending upon the speed with which they are rotating in their orbits. As we increase this speed the distance of the corpuscles from the centre of the sphere will increase until at a certain speed the corpuscles will reach the surface of the sphere ; further increases in speed will cause them first to rotate outside the sphere and finally leave the sphere altogether, when the atom will break up.

In this way we see that the constitution of the


QO ELECTRICITY AND MATTER

t7O

aggregate will not be permanent, if the kinetic energy due to the velocity of the corpuscles inside the sphere relative to the centre of the sphere ex- ceeds a certain value. We shall, for the sake of brevity, speak of this kinetic energy of the cor- puscles within the atom as the corpUBfatfar tem- perature of the atom, and we may express the preceding result by saying that the atom will not be stable unless its corpuscular temperature is below a certain value.

We must be careful to distinguish between cor- puscular temperature, which is the mean kinetic energy of the corpuscles inside the atom, and the molecular temperature, which is the mean kinetic energy due to the motion of the centre of gravity of the atom. These temperatures are probably not in any very close relationship with each other. They would be proportional to each other if the law known as the law of equipartition of energy among the various degrees of freedom of the atom were to apply. This law is, however, inconsistent with the physical properties of gases, and in the proof given of it in the kinetic theory of gases, no estimate is given of the time required to establish the state con- templated by the law ; it may be that this time is so long that gases are never able to get into this state.


CONSTITUTION OF THE ATOM 99

Let us now take the case of two aggregations, A and £, whose corpuscular temperatures are high, though not so high, of course, as to make A and B unstable when apart, and suppose, in order to give them the best possible chance of combining, that the centres of gravity of A and B when quite close to each other are at rest, will A and B unite to form a more complex aggregate as they would if the corpuscles in them were at rest  ? We can easily, I think, see that they will not necessa- rily do so. For as A and B approach each other, under their mutual attractions, the potential en- ergy due to the separation of A and B will dimin- ish and their kinetic energy will increase. This in- crease in the kinetic energy of the corpuscles in A and B will increase the tendency of the corpuscles to leave their atoms, and if the increase in the kin- etic energy is considerable A and B may each lose one or more corpuscles. The departure of a cor- puscle will leave A and B positively charged, and they will tend to separate under the repulsion of these charges. When separated they will have each a positive charge ; but as there are now free corpuscles with negative charges moving about in the region in which A and B are situated, these positive charges will ultimately be neutralized by


100 ELECTRICITY AND MATTER

corpuscles striking against A and B and remain- ing in combination with them.

We thus conclude that unless the corpuscular temperature after union is less than a certain limit- ing value, the union cannot be permanent, the complex formed being unstable, and incapable of a permanent existence. Now, the corpuscular temperature of the aggregate formed by A and B will depend upon the corpuscular temperatures of A and B before union, and also upon the diminu- tion in the potential energy of the system occa- sioned by the union of A and B. If the corpuscu- lar temperatures of A and B before union were very high, the corpuscular temperature after union would be high also; if they were above a cer- tain limit, the corpuscular temperature after union would be too high for stability, and the aggre- gate AB would not be formed. Thus, one con- dition for the formation of complex aggregate's is that the corpuscular temperature of their constitu- ents before combination should be sufficiently low.

If the molecula/t* temperature of the gas in which A and B are molecules is very high, com- bination may be prevented by the high relative velocity of A and B carrying them apart in spite of their mutual attraction. The point, however,


CONSTITUTION OF THE ATOM JQ1

which I wish to emphasize is, that we cannot se- cure the union merely by lowering the molecular temperature, i.e., by cooling the gas  ; union will be impossible unless the corpuscular temperature, i.e., the kinetic energy due to the motion of the corpuscles inside the atom, is reduced below a cer- tain value. We may prevent union by raising the molecular temperature of a gas, but we cannot en- sure union by lowering it.

Thus, to take a specific example, the reason, on this view, why the atoms of hydrogen present on the earth do not combine to form some other ele- ment, even at the exceedingly low temperature at which hydrogen becomes liquid, is that even at this temperature the kinetic energy of the corpus- cles inside the atom, i.e., the corpuscular tempera- ture, is too great. It may be useful to repeat here what we stated before, that there is no very inti- mate connection between the corpuscular and mo- lecular temperatures, and that we may reduce the latter almost to the absolute zero without greatly affecting the former.

We shall now proceed to discuss the bearing of these results on the theory that the different chemical elements have been gradually evolved by the aggregation of primordial units.


102 ELECTRICITY AND MATTER

Let us suppose that the first stage has been reached and that we have a number of systems formed by the union of two units. When first these binary systems, as we shall call them, were formed, the corpuscles in the system would have a considerable amount of kinetic energy. This would be so, because when the two units have come together there must be an amount of kinetic energy produced equal to the diminution in the potential energy consequent upon the coalescence of the two units. As these binary systems have or- iginally high corpuscular temperatures they will not be likely to combine with each other or with another unit ; before they can do so the kinetic energy of the corpuscles must get reduced.

We shall proceed immediately to discuss the way in which this reduction is effected, but we shall anticipate the result of the discussion by saying that it leads to the result that the rate of decay in the corpuscular temperature probably varies greatly from one binary system to another.

Some of the systems will therefore probably have reached a condition in which they are able to combine with each other or with a single unit long before others are able to do so. The systems of the first kind will combine, and thus we shall


CONSTITUTION OF THE ATOM 1Q3

have systems formed, some of which contain three, others four units, while at the same time there are many of the binary systems left. Thus, the appear- ance of the more complex systems need not be simultaneous with the disappearance of all the sim- pler ones.

The same principle will apply to the formation of further aggregations by the systems containing three or four units ; some of these will be ready to unite before the others, and we may have systems containing eight units formed before the more per- sistent of those containing four, three, two or even one unit have disappeared. With the further ad- vance of aggregation the number of different sys- tems present at one and the same time will in- crease.

Thus, if we regard the systems containing differ- ent numbers of units as corresponding to the different chemical elements, then as the universe gets older elements of higher and higher atomic weight may be expected to appear. Their appear- ance, however, will not involve the annihilation of the elements of lower atomic weight. The number of atoms of the latter will, of course, diminish, since the heavier elements are by hypothesis built up of material furnished by the lighter. The whole


104 ELECTRICITY AND MATTER

of the atoms of the latter would not, however, all be used up at once, and thus we may have a very large number of elements existing at one and the same time.

If, however, there is a continual fall in the cor- puscular temperature of the atoms through radia- tion, the lighter elements will disappear in time, and unless there is disintegration of the heavier atoms, the atomic weight of the lightest element surviving will continually increase. On this view, since hydrogen is the lightest known element and the atom of hydrogen contains about a thousand corpuscles, all aggregations of less than a thousand units have entered into combination and are no longer free.

Tfie way the Corpuscles in tlie Atom Lose or Gain Kinetic Energy

If the kinetic energy arising from the motion of the corpuscles relatively to the centre of gravity of the atom could by collisions be transformed into kinetic energy due to the motion of the atom as a whole, i.e., into molecular temperature, it would follow from the kinetic theory of gases, since the number of corpuscles in the atom is ex- ceedingly large, that the specific heat of a gas at


CONSTITUTION OF THE ATOM 105

constant pressure would be very nearly equal to the specific heat at constant volume ; whereas, as a matter of fact, in no gas is there any approach to equality in these specific heats. We conclude, therefore, that it is not by collisions that the kinetic energy of the corpuscles is diminished.

We have seen, however (page 68), that a mov- ing electrified particle radiates energy whenever its velocity is changing either in magnitude or direction. The corpuscles in the atom will thus emit electric waves, radiating energy and so losing kinetic energy.

The rate at which energy is lost in this way by the corpuscles varies very greatly with the num- ber of the corpuscles and the way in which they are moving. Thus, if we have a single corpuscle describing a circular orbit of radius a with uni- form velocity v, the loss of energy due to radia- tion per second is - =^-5, where e is the charge on 3 V or

the corpuscles and V the velocity of light. If instead of a single corpuscle we had two corpus- cles at opposite ends of a diameter moving round the same orbit with the same velocity as the sin- gle corpuscle, the loss of energy per second from the two would be very much less than from the single


106 ELECTRICITY AND MATTER

corpuscle, and the smaller the velocity of the cor- puscle the greater would be the diminution in the loss of energy produced by increasing the number of corpuscles. The effect produced by increasing the number of corpuscles is shown in the following table, which gives the rate of radia- tion for each corpuscle for various numbers of corpuscles arranged at equal angular intervals round the circular orbit.

The table applies to two cases  ; in one the veloc- ity of the corpuscles is taken as one-tenth that of light, and in the second as one-hundredth. The radiation from a single corpuscle is in each case taken as unity.

Number of corpuscles. Radiation from each corpuscle.

_F V

~~IQ ~ TOO

1 1 1

2 9.6 x 10~2 9.6 x ID"4

3 4.6 x l(r3 4.6 x 10-7

4 1.7x10-^ 1.7 x 10-10

5 5.6X10-5 5.6 x 10~13

6 1.6 x 10~7 1.6xlO~17

Thus, we see that the radiation from each of a group of six corpuscles moving with one-tenth the velocity of light is less than one-five-millionth part of the radiation from a single corpuscle, describ-


CONSTITUTION OF THE ATOM 1Q7

ing the same orbit with the same velocity, while, when the velocity of the corpuscles is only one- hundredth of that of light, the reduction in the radiation is very much greater.

If the corpuscles are displaced from the sym- metrical position in which they are situated at equal intervals round a circle whose centre is at rest, the rate of radiation will be very much in- creased. In the case of an atom containing a large number of corpuscles the variation in the rate at which energy is radiated will vary very rapidly with the way the corpuscles are moving about in the atom. Thus, for example, if we had a large number of corpuscles following close on one an- other's heels round a circular orbit the radiation would be exceedingly small ; it would vanish alto- gether if the corpuscles were so close together that they formed a continuous ring of negative electrification. If the same number of particles were moving about irregularly in the atom, then though the kinetic energy possessed by the cor- puscles in the second case might be no greater than in the first, the rate of radiation, i.e., of cor- puscular cooling, would be immensely greater.

Thus, we see that in the radiation of energy from corpuscles whose velocity is not uniform we


10g ELECTRICITY AND MATTER

have a process going on which will gradually cool the corpuscular temperature of the atom, and so, if the view we have been discussing is correct, enable the atom to form further aggregations and thus tend to the formation of new chemical ele- ments.

This cooling process must be an exceedingly slow one, for although the corpuscular tempera- ture when the atom of a new element is formed is likely to be exceedingly high, and the lowering in that temperature required before the atom can enter again into fresh aggregations very large, yet we have evidence that some of the elements must have existed unchanged for many thousands, nay, millions of years ; we have, indeed, no direct evidence of any change at all in the atom. I think, however, that some of the phenomena of radio-activity to which I shall have to allude later, afford, I will not say a proof of, but a very strong presumption in favor of some such secular changes taking place in the atom.

We must remember, too, that the corpuscles in any atom are receiving and absorbing radiation from other atoms. This will tend to raise the corpuscular temperature of the atom and thus help to lengthen the time required for that


CONSTITUTION OF THE ATOM \QQ

temperature to fall to the point where fresh aggregations of the atom may be formed.

The fact that the rate of radiation depends so much upon the way the corpuscles are moving about in the atom indicates that the lives of the different atoms of any particular element will not be equal ; some of these atoms will be ready to enter upon fresh changes long before the others. It is important to realize how large are the amounts of energy involved in the formation of a complex atom or in any rearrangement of the con- figuration of the corpuscles inside it. If we have an atom containing n corpuscles each with a charge e measured in electrostatic units, the total quantity of negative electricity in the atom is n e and there is an equal quantity of positive elec- tricity distributed through the sphere of positive electrification; hence, the work required to sep- arate the atom into its constituent units will be

comparable with ^ — '- , a being the radius of the

sphere containing the corpuscles. Thus, as the atom has been formed by the aggregation of these

units v16' will be of the same order of magni- a

tude as the kinetic energy imparted to those con-


HO ELECTRICITY AND MATTER

stituents during their whole history, from the time they started as separate units, down to the time they became members of the atom under consideration. They will in this period have radi- ated away a large quantity of this energy, but the following calculation will show what an enormous amount of kinetic energy the corpuscles in the atom must possess even if they have only retained an exceedingly small fraction of that communi-

cated to them. Let us calculate the value of * — '-

a

for all the atoms in a gram of the substance  ; let N be the number of these atoms in a gram, then

N^- — '- is the value of the energy acquired by these atoms. If Mis the mass of an atom NM= 1, thus  :


but if m is the mass of a corpuscle and therefore


_

a ma'


now when e is measured in electrostatic units

<L = 3 x 1017 and e = 3.4 X 1Q-10  ; m


CONSTITUTION OF THE ATOM m

and therefore

N&*y= 10.2X107X -. (1) «  a

Let us take the case of the hydrogen atom for which n = 1000, and take for a the value usually assumed in the kinetic theory of gases for the radius of the atom, i.e., 10"8 cm. then

jy (M-= 1.02 X 1019ergs;

this amount of energy would be sufficient to lift a million tons through a height considerably ex- ceeding one hundred yards. We see, too, from (1) that this energy is proportional to the num- ber of corpuscles, so that the greater the molecu- lar weight of an element, the greater will be the amount of energy stored up in the atoms in each gram.

We shall return to the subject of the internal changes in the atom when we discuss some of the phenomena of radio-activity, but before doing so it is desirable to consider more closely the way the corpuscles arrange themselves in the atom. We shall begin with the case where the corpuscles are at rest. The corpuscles are supposed to be in a sphere of uniform positive electrification which produces a radial attractive force on each cor-


112


ELECTRICITY AND MATTER


Flo. 16.


puscle proportional to its distance from the centre of the sphere, and the problem is to arrange the corpuscles in the sphere so that they are in equilibrium under this attraction and their mutual re- pulsions. If there are only two corpuscles, A B, we can see at once that they will be in equi- librium if placed so that A B and the centre of the sphere are in the same straight line and OA = OB = $ the radius of the sphere.

If there are three corpuscles, A B C, they will be in equilibrium of A B C as an equilateral tri- angle with its centre at O and OA= OB = OC = (\y, or .57 times the radius of the sphere.

If there are four corpuscles these will be in equilibrium if placed at the angular points of a regular tetrahedron with its cen- tre at the centre of the sphere. In these cases the corpuscles are all on the surface of a sphere con- centric with the sphere of positive electrification, and we might suppose that whatever the number of corpuscles the position of equilibrium would be


FIG. 15.


CONSTITUTION OF THE ATOM H3

one of symmetrical distribution over the surface of a sphere. Such a distribution would indeed technically be one of equilibrium, but a mathe- matical calculation shows that unless the number of corpuscles is quite small, say seven or eight at the most, this arrangement is unstable and so can never persist. When the number of corpuscles is greater than this limiting number, the corpuscles break up into two groups. One group containing the smaller number of corpuscles is on the surface of a small body concentric with the sphere ; the remainder are on the surface of a larger concen- tric body. When the number of corpuscles is still further increased there comes a stage when the equilibrium cannot be stable even with two groups, and the corpuscles now divide themselves into three groups, arranged on the surfaces of con- centric shells ; and as we go on increasing the number we pass through stages in which more and more groups are necessary for equilibrium. With any considerable number of corpuscles the prob- lem of finding the distribution when in equilibrium becomes too complex for calculation  ; and we have to turn to experiment and see if we can make a model in which the forces producing equilibrium are similar to those we have supposed to be at


114 ELECTRICITY AND MATTER

work in the corpuscle. Such a model is afforded by a very simple and beautiful experiment first made, I think, by Professor Mayer. In this experi- ment a number of little magnets are floated in a vessel of water. The magnets are steel needles magnetized to equal strengths and are floated by being thrust through small disks of cork. The magnets are placed so that the positive poles are either all above or all below the surface of the water. These positive poles, like the corpuscles, repel each other with forces varying inversely as the distance between them. The attractive force is provided by a negative pole (if the little mag- nets have their positive poles above the water) sus- pended some distance above the surface of the water. This pole will exert on the positive poles of the little floating magnets an attractive force the component of which, parallel to the surface of the water, will be radial, directed to 0, the projection of the negative pole on the surface of the water, and if the negative pole is some dis- tance above the surface the component of the force to O will be very approximately proportional to the distance from O. Thus the forces on the poles of the floating magnets will be very similar to those acting on the corpuscle in our hypothetical atom ;


CONSTITUTION OF THE ATOM


H5


the chief difference being that the corpuscles are free to move about in all directions in space, while the poles of the floating magnets are constrained to move in a plane parallel to the surface of the water.

The configurations which the floating magnets assume as the number of magnets increases from two up to nineteen is shown in Fig. 17, which was given by Mayer.


FIG. 17.

The configuration taken up when the magnets are more numerous can be found from the follow- ing table, which is also due to Mayer. From this table it will be seen that when the number of floating magnets does not exceed five the magnets


116


ELECTRICITY AND MATTER


arrange themselves at the corners of a regular polygon, five at the corners of a pentagon, four at the corners of a square and so on. When the number is greater than five this arrangement no longer holds. Thus, six magnets do not arrange themselves at the corners of a hexagon, but divide into two systems, one magnet being at the centre and five outside it at the corners of a regular penta- gon. This arrangement in two groups lasts until there are fifteen magnets, when we have three groups ; with twenty-seven magnets we get four groups and so on.


Arrangement of Magi

tets (Mayer)

1. 2. 3.

4. 6.

1-5 (2.6 (3.7 •1-6 (2.7 (3.8

I4-8 I5-9

(4.9 (

1 . 7

1.6.9 C2 . 7 . 10 3.7 10

(4 . 8 . 12 J5 . 9 . 12

1.6.9 J 2 . 8 . 10

3 . 7 . 11

1 4 . 8 . 13 1 5 . 9 . 13

1 . 6 . 10 (2 . 7 . 11 1 . 6 . 11

3 . 8 . 10 3 . 8 . 11

»4 . 9 . 12 4 . 9 . 13

3 . 8 . 12

3 . 8 . 13

CONSTITUTION OF THE ATOM


f • 6

. 5

. 9 . . 9 .

12 ( 2 . 7 . 10 . 15

13 \2 . 7 . 12 . 14

3 . 7 . 12 . 13 f4 . 9 . 13 . 14 3 . 7 . 12 . 14 J 4 .9 . 13 . 15

. 6

. 9 .

12

3.7. 13 . 14 (4 . 9 . 14 . 15

. 6

. 10 .

12

3.7 13 . 15

. C

. 10 .

13

. 6

. 11 .

2

. 6

. 11 .

13

. 6

. 11 .

14

. • 6

. 11 .

15

Where, for example, 3. 7. 12. 13 means that thirty-five magnets arrange themselves so that there is a ring of three magnets inside, then a ring of seven, then one of twelve, and one of thirteen outside.

I think this table affords many suggestions tow- ard the explanation of some of the properties possessed by atoms. Let us take, for example, the chemical law called the Periodic Law ; according to this law if we arrange the elements in order of increasing atomic weights, then taking an element of low atomic weight, say lithium, we find certain properties associated with it. These properties are not possessed by the elements immediately following it in the series of increasing atomic weight ; but they appear again when we come to sodium, then they disappear again for a time,


Ug ELECTRICITY AND MATTER

but reappear when we reach potassium, and so on. Let us now consider the arrangements of the floating magnets, and suppose that the number of magnets is proportional to the combining weight of an element. Then, if any property were asso- ciated with the triangular arrangement of magnets, it would be possessed by the elements whose com- bining weight was on this scale three, but would not appear again until we reached the combining weight ten, when it reappears, as for ten magnets we have the triangular arrangement in the middle and a ring of seven magnets outside. When the number of magnets is increased the triangular arrangement disappears for a time, but reappears with twenty magnets, and again with thirty-five, the triangular arrangement appearing and dis- appearing in a way analogous to the behavior of the properties of the elements in the Periodic Law. As an example of a property that might very well be associated with a particular grouping of the corpuscles, let us take the times of vibra- tion of the system, as shown by the position of the lines in the spectrum of the element. First let us take the case of three corpuscles by them- selves in the positively electrified sphere. The three corpuscles have nine degrees of freedom, so


CONSTITUTION OF THE ATOM HQ

that there are nine possible periods. Some of these periods in this case would be infinitely long, and several of the possible periods would be equal to each other, so that we should not get nine dif- ferent periods.

Suppose that the lines in the spectrum of the three corpuscles are as represented in Fig. 18 a,


A


a c D c


e 3 / 2


A B c a £


FIG. 18.

where the figures under the lines represent the number of periods which coalesce at that line ; i.e., regarding the periods as given by an equation with nine roots, we suppose that there is only one root giving the period corresponding to the line -4, while corresponding to I> there are two equal roots, three equal roots corresponding to <7, one


120 ELECTRICITY AND MATTER

root, to O, and two to E. These periods would have certain numerical relations to each other, in- dependent of the charge on the corpuscle, the size of the sphere in which they are placed, or their distance from the centre of the sphere. Each of these quantities, although it does not affect the ratio of the periods, will have a great effect upon the absolute value of any one of them. Now, suppose that these three corpuscles, instead of being alone in the sphere, form but one out of several groups in it, just as the triangle of mag- nets forms a constituent of the grouping of 3, 10, 20, and 35 magnets. Let us consider how the presence of the other groups would affect the periods of vibration of the three corpuscles. The absolute values of the periods would generally be entirely different, but the relationship existing be- tween the various periods would be much more persistent, and although it might be modified it would not be destroyed. Using the phraseology of the Planetary Theory, we may regard the motion of the three corpuscles as " disturbed " by the other groups.

When the group of three corpuscles was by it- self there were several displacements which gave the same period of vibration ; for example, corre-


CONSTITUTION OF THE ATOM 121

spending to the line C there were three displace- ments, all giving the same period. When, how^ ever, there are other groups present, then these different displacements will no longer be sym- metrical with respect to these groups, so that the three periods will no longer be quite equal. They would, however, be very nearly equal unless the effect of the other groups is very large. Thus, in the spectrum, <7, instead of being a single line, would become a triplet, while B and E would be- come doublets. A D would remain single lines.

Thus, the spectrum would now resemble Fig. 18 b', the more groups there are surrounding the group of three the more will the motion of the latter be disturbed and the greater the separation of the constituents of the triplets and doublets. The appearance as the number of groups increases is shown in Fig. 18 £, c. Thus, if we regarded the element which contain this particular group- ing of corpuscles as being in the same group in the classification of elements according to the Periodic Law, we should get in the spectra of these ele- ments homologous series of lines, the distances be- tween the components of the doublets and triplets increasing with the atomic weight of the elements. The investigations of Rydberg, Runge and Pas-


122 ELECTRICITY AXD MATTER

chen and Keyser have shown the existence in the spectra of elements of the same group series of lines having properties in many respects analogous to those we have described.

Another point of interest given by Mayer's ex- periments is that there is more than one stable configuration for the same number of magnets; these configurations correspond to different amounts of potential energy, so that the passage from the configuration of greater potential energy to that of less would give kinetic energy to the corpuscle. From the values of the potential energy stored in the atom, of which we gave an estimate on page 111, we infer that a change by even a small fraction in that potential energy would develop an amount of kinetic energy which if converted into heat would greatly transcend the amount of heat developed when the atoms undergo any known chemical combination.

An inspection of the table shows that there are certain places in it where the nature of the con- figuration changes very rapidly with the number of magnets  ; thus, five magnets form one group, while six magnets form two; fourteen magnets form two groups, fifteen three ; twenty - seven magnets form three groups, twenty-eight four,


CONSTITUTION OF THE ATOM 123

and so on. If we arrange the chemical elements in the order of their atomic weights we find there are certain places where the difference in proper- ties of consecutive elements is exceptionally great ; thus, for example, we have extreme differences in properties between fluorine and sodium. Then there is more or less continuity in the properties until we get to chlorine, which is followed by potassium; the next break occurs at bromine and rubidium and so on. This effect seems analogous to that due to the regrouping of the magnets.

So far we have supposed the corpuscles to be at rest ; if, however, they are in a state of steady motion and describing circular orbits round the centre of the sphere, the effect of the centrifugal force arising from this motion will be to drive the corpuscles farther away from the centre of the sphere, without, in many cases, destroying the character of the configuration. Thus, for example, if we have three corpuscles in the sphere, they will, in the state of steady motion, as when they are at rest, be situated at the corners of an equi- angular triangle ; this triangle will, however, be rotating round the centre of the sphere, and the distance of the corpuscles from the centre will be


124 ELECTRICITY AND MATTER

greater than when they are at rest and will in- crease with the velocity of the corpuscles.

There are, however, many cases in which rota- tion is essential for the stability of the configura- tion. Thus, take the case of four corpuscles. These, if rotating rapidly, are in stable steady motion when at the corners of a square, the plane of the square being at right angles to the axis of rotation ; when, however, the velocity of rotation of the corpuscles falls below a certain value, the arrangement of four corpuscles in one plane be- comes unstable, and the corpuscles tend to place themselves at the corners of a regular tetrahedron, which is the stable arrangement when the cor- puscles are at rest. The system of four corpuscles at the corners of a square may be compared with a spinning top, the top like the corpuscles being unstable unless its velocity of rotation exceeds a certain critical value. Let us suppose that initially the velocity of the corpuscles exceeds this value, but that in some way or another the corpuscles gradually lose their kinetic energy; the square arrangement will persist until the ve- locity of the corpuscles is reduced to the critical value. The arrangement will then become un- stable, and there will be a convulsion in the sys-


CONSTITUTION OF THE ATOM 125

tern accompanied by a great evolution of kinetic energy.

Similar considerations will apply to many as- semblages of corpuscles. In such cases the con- figuration when the corpuscles are rotating with great rapidity will (as in the case of the four cor- puscles) be essentially different from the configu- ration of the same number of corpuscles when at rest. Hence there must be some critical velocity of the corpuscles, such that, for velocities greater than the critical one, a configuration is stable, which becomes unstable when the velocity is reduced below the critical value. When the ve- locity sinks below the critical value, instability sets in, and there is a kind of convulsion or ex- plosion, accompanied by a great diminution in the potential energy and a corresponding increase in the kinetic energy of the corpuscles. This increase in the kinetic energy of the corpuscles may be sufficient to detach considerable numbers of them from the original assemblage.

These considerations have a very direct bearing on the view of the constitution of the atoms which we have taken in this chapter, for they show that with atoms of a special kind, i.e., with special atomic weights, the corpuscular cooling caused by


126 ELECTRICITY AND MATTER

the radiation from the moving corpuscles which we have supposed to be slowly going on, might, when it reached a certain stage, produce instabil- ity inside the atom, and produce such an in- crease in the kinetic energy of the corpuscles as to give rise to greatly increased radiation, and it might be detachment of a portion of the atom. It would cause the atom to emit energy ; this energy being derived from the potential energy due to the arrangement of the corpuscles in the atom. We shall see when we consider the phe- nomenon of radio-activity that there is a class of bodies which show phenomena analogous to those just described.

On the view that the lighter elements are formed first by the aggregation of the unit doublet, the negative element of which is the cor- puscle, and that it is by the combination of the atoms of the lighter elements that the atoms of the heavier elements are produced, we should ex- pect the corpuscles in the heavy atoms to be ar- ranged as it were in bundles, the arrangement of the corpuscles in each bundle being similar to the arrangement in the atom of some lighter element. In the heavier atom these bundles would act as subsidiary units, each bundle corresponding to


CONSTITUTION OF THE ATOM J27

one of the magnets in the model formed by the floating magnets, while inside the bundle them- selves the corpuscle would be the analogue of the magnet.

We must now go on to see whether an atom built up in the way we have supposed could pos- sess any of the properties of the real atom. Is there, for example, in this model of an atom any scope for the electro-chemical properties of the real atom ; such properties, for example, as those illustrated by the division of the chemical ele- ments into two classes, electro-positive and electro- negative. Why, for example, if this is the con- stitution of the atom, does an atom of sodium or potassium tend to acquire a positive, the atom of chlorine a negative charge of electricity  ? Again, is there anything in the model of the atom to suggest the possession of such a property as that called by the chemists valency  ; i.e., the property which enables us to divide the elements into groups, called monads, dyads, triads, such that in a compound formed by any two elements of the first group the molecule of the compound will contain the same number of atoms of each element, while in a compound formed by an element A in the first group with one B in the second, the mole-


12g ELECTRICITY AND MATTER

cule of the compound contains twice as many atoms of A as of B, and so on  ?

Let us now turn to the properties of the model atom. It contains a very large number of corpus- cles in rapid motion. We have evidence from the phenomena connected with the conduction of electricity through gases that one or more of these corpuscles can be detached from the atom. These may escape owing to their high veloc- ity enabling them to travel beyond the attrac- tion of the atom. They may be detached also by collision of the atom with other rapidly moving atoms or free corpuscles. When once a corpuscle has escaped from an atom the latter will have a pos- itive charge. This will make it more difficult for a second negatively electrified corpuscle to escape, for in consequence of the positive charge on the atom the latter will attract the second corpuscle more strongly than it did the first. Now we can readily conceive that the ease with which a par- ticle will escape from, or be knocked out of, an atom may vary very much in the atoms of the dif- ferent elements. In some atoms the velocities of the corpuscles may be so great that a corpuscle escapes at once from the atom. It may even be that after one has escaped, the attraction of the


CONSTITUTION OF THE ATOM 129

positive electrification thus left on the atom is not sufficient to restrain a second, or even a third, corpuscle from escaping. Such atoms would ac- quire positive charges of one, two, or three units, according as they lost one, two, or three corpus- cles. On the other hand, there may be atoms in which the velocities of the corpuscles are so small that few, if any, corpuscles escape of their own accord, nay, they may even be able to receive one or even more than one corpuscle before the repulsion exerted by the negative electrification on these foreign corpuscles forces any of the original corpuscles out. Atoms of this kind if placed in a region where corpuscles were present would by aggregation with these corpuscles re. ceive a negative charge. The magnitude of the negative charge would depend upon the firmness with which the atom held its corpuscles. If a negative charge of one corpuscle were not suf- ficient to expel a corpuscle while the negative charge of two corpuscles could do so, the maxi- mum negative charge on the atom would be one unit. If two corpuscles were not sufficient to expel a corpuscle, but three were, the maximum nega- tive charge would be two units, and so on. Thus, the atoms of this class tend to get charged with


L30 ELECTRICITY AND MATTER

negative electricity and correspond to the electro- negative chemical elements, while the atoms of the class we first considered, and which readily lose corpuscles, acquire a positive charge and corre- spond to the atoms of the electro-positive elements. We might conceive atoms in which the equilib- rium of the corpuscles was so nicely balanced that though they do not of themselves lose a cor- puscle, and so do not acquire a positive charge, the repulsion exerted by a foreign corpuscle coming on to the atom would be sufficient to drive out a corpuscle. Such an atom would be incapable of receiving a charge either of positive or negative electricity.

Suppose we have a number of the atoms that readily lose their corpuscles mixed with a num- ber of those that can retain a foreign corpuscle. Let us call an atom of the first class A, one of the second £, and suppose that the A atoms are of the kind that lose one corpuscle while the IB atoms are of the kind that can retain one, but not more than one  ; then the corpuscles which escape from the A atoms will ultimately find a home on the B atoms, and if there are an equal number of the two kinds of atoms present we shall get ultimate- ly all the A atoms with the unit positive charge,


CONSTITUTION OF THE ATOM 131

all the B atoms with the unit negative charge. These oppositely electrified atoms will attract each other, and we shall get the compound A B formed. If the A atoms had been of the kind that lost two corpuscles, and the B atoms the same as before, then the A atoms would get the charge of two positive units, the B atoms a charge of one unit of negative electricity. Thus, to form a neutral system two of the B atoms must com- bine with one of the A's and thus the compound A HI would be formed.

Thus, from this point of view a univalent elec- tro-positive atom is one which, under the circum- stances prevailing when combination is taking place, has to lose one and only one corpuscle be- fore stability is attained ; a univalent electro-neg- ative atom is one which can receive one but not more than one corpuscle without driving off other corpuscles from the atom; a divalent electro- positive atom is one that loses two corpuscles and no more, and so on. The valency of the atom thus depends upon the ease with which corpus- cles can escape from or be received by the atom ; this may be influenced by the circumstances existing when combination is taking place. Thus, it would be easier for a corpuscle, when once it


132 ELECTRICITY AND MATTER

had got outside the atom, to escape being pulled back again into it by the attraction of its positive electrification, if the atom were surrounded by good conductors than if it were isolated in space. We can understand, then, why the valency of an atom may in some degree be influenced by the physical conditions under which combination is taking place. On the view that the attraction between the atoms in a chemical compound is electrical in its origin, the ability of an element to enter into chemical combination depends upon its atom hav- ing the power of acquiring a charge of electricity. This, on the preceding view, implies either that the uncharged atom is unstable and has to lose one or more corpuscles before it can get into a steady state, or else that it is so stable that it can retain one or more additional corpuscles without any of the original corpuscles being driven out. If the range of stability is such that the atom, though stable when uncharged, becomes unstable when it receives an additional corpuscle, the atom will not be able to receive a charge either of positive or negative electricity, and will therefore not be able to enter into chemical combination. Such an atom would have the properties of the atoms of such elements as argon or helium.


CONSTITUTION OF THE ATOM 133

The view that the forces which bind together the atoms in the molecules of chemical compounds are electrical in their origin, was first proposed by Berzelius ; it was also the view of Davy and of Faraday. Helmholtz, too, declared that the mightiest of the chemical forces are electrical in their origin. Chemists in general seem, however, to have made but little use of this idea, having apparently found the conception of "bonds of affinity" more fruitful. This doctrine of bonds is, however, when regarded in one aspect almost identical with the electrical theory. The theory of bonds when represented graphically supposes that from each univalent atom a straight line (the symbol of a bond) proceeds; a divalent atom is at the end of two such lines, a trivalent atom at the end of three, and so on ; and that when the chemical compound is represented by a graphic formula in this way, each atom must be at the end of the proper number of the lines which represent the bonds. Now, on the electrical view of chemical combination, a univalent atom has one unit charge, if we take as our unit of charge the charge on the corpuscle ; the atom is therefore the beginning or end of one unit Fara- day tube : the beginning if the charge on the


134 ELECTRICITY AND MATTEB

atom is positive, the end if the charge is nega- tive. A divalent atom has two units of charge and therefore it is the origin or termination of two unit Faraday tubes. Thus, if we interpret the "bond" of the chemist as indicating a unit Fara- day tube, connecting charged atoms in the mole- cule, the structural formulae of the chemist can be at once translated into the electrical theory. There is, however, one point of difference which deserves a little consideration : the symbol indi- cating a bond on the chemical theory is not re- garded as having direction  ; no difference is made on this theory between one end of a bond and the other. On the electrical theory, however, there is a difference between the ends, as one end cor- responds to a positive, the other to a negative charge. An example or two may perhaps be the easiest way of indicating the effect of this consid- eration. Let us take the gas ethane whose structu- ral formula is written


According to the chemical view there is no differ-


CONSTITUTION OF THE ATOM 135

ence between the two carbon atoms in this com- pound ; there would, however, be a difference on the electrical view. For let us suppose that the hydrogen atoms are all negatively electrified; the three Faraday tubes going from the hydrogen atoms to each carbon atom give a positive charge of three units on each carbon atom. But in addition to the Faraday tubes coming from the hydrogen atoms, there is one tube which goes from one car- bon atom to the other. This means an additional positive charge on one carbon atom and a nega- tive charge on the other. Thus, one of the carbon atoms will have a charge of four positive units, while the other will have a charge of three positive and one negative unit, i.e., two positive units ; so that on this view the two carbon atoms are not in the same state. A still greater difference must exist between the atoms when we have what is called double linking, i.e., when the carbon atoms are supposed to be connected by two bonds, as in the compound


136 ELECTRICITY AND MATT K II

Here, if one carbon atom had a charge of four posi- tive units, the other would have a charge of two positive and two negative units.

We might expect to discover such differences as are indicated by these considerations by the in- vestigation of which are known as additive prop- erties, i.e, properties which can be calculated when the chemical constitution of the molecule is known. Thus, let A B C represent the atoms of three chemical elements, then if p is the value of some physical constant for the molecule of . I ,, q the value for 13 %, and r for 6'2, then if this con- stant obeys the additive law, its value for a mole- cule of the substance whose chemical composition is represented by the formula A& J2y Cz is

\px-\r\qy-\-\rz.

We can only expect relations like this to hold when the atoms which occur in the different compounds corresponding to different values of x y z are the same. If the atom A occurs in different states in different compounds we should have to use different values of p for these compounds.

A well-known instance of the additive prop- erty is the refractive power of different substances for light, and in this case chemists find it neces-


CONSTITUTION OP THE ATOM 137

sary to use different values for the refraction due a carbon atom according as the atom is doubly or singly linked. They use, however, the same value for the refraction of the carbon atom when singly linked with another atom as when, as in the com- pound G HI, it is not linked with another carbon atom at all.

It may be urged that although we can conceive that one atom in a compound should be positively and the other negatively electrified when the atoms are of different kinds, it is not easy to do so when the atoms are of the same kind, as they are in the molecules of the elementary gases HI, OK JVj and so on. With reference to this point we may remark that the electrical state of an atom, depending as it does on the power of the atom to emit or retain corpuscles, may be very largely influenced by circumstances external to the atom. Thus, for an example, an atom in a gas when surrounded by rapidly moving atoms or corpuscles which keep striking against it may have corpuscles driven out of it by these collisions and thus become positively electrified. On the other hand, we should expect that, ceteris paribus, the atom would be less likely to lose a corpuscle when it is in a gas than when in a solid or a


138 ELECTRICITY AND MATTER

liquid. For when in a gas after a corpuscle has just left the atom it has nothing beyond its own velocity to rely upon to escape from the attraction of the positively electrified atom, since the other atoms are too far away to exert any forces upon it. When, however, the atom is in a liquid or a solid, the attractions of the other atoms which crowd round this atom may, when once a corpus- cle has left its atom, help it to avoid falling back again into atom. As an instance of this effect we may take the case of mercury in the liquid and gaseous states. In the liquid state mercury is a good conductor of electricity. One way of regarding this electrical conductivity is to suppose that corpuscles leave the atoms of the mercury and wander about through the inter- stices between the atoms. These charged cor- puscles when acted upon by an electric force are set in motion and constitute an electric cur- rent, the conductivity of the liquid mercury in- dicating the presence of a large number of cor- puscles. When, however, mercury is in the gaseous state, its electrical conductivity has been shown by Strutt to be an exceedingly small fraction of the conductivity possessed by the same number of molecules when gaseous. We have thus indications


CONSTITUTION OF THE ATOM 139

that the atoms even of an electro-positive sub- stance like mercury may only lose comparatively few corpuscles when in the gaseous state. Sup- pose then that we had a great number of atoms all of one kind in the gaseous state and thus mov- ing about and coming into collision with each other; the more rapidly moving ones, since they would make the most violent collisions, would be more likely to lose corpuscles than the slower ones. The faster ones would thus by the loss of their corpuscles become positively electrified, while the corpuscles driven off would, if the atoms were not too electro-positive to be able to retain a negative charge even when in the gase- ous state, tend to find a home on the more slowly moving atoms. Thus, some of the atoms would get positively, others negatively electrified, and those with changes of opposite signs would com- bine to form a diatomic molecule. This argu- ment would not apply to very electro-positive gases. These we should not expect to form mole- cules, but since there would be many free cor- puscles in the gas we should expect them to possess considerable electrical conductivity.


CHAPTER VI

RADIO-ACTIVITY AND RADIO-ACTIVE SUB- STANCES

IN 1896 Becquerel discovered that uranium and its salts possess the power of giving out rays which, like Rontgen and cathode rays, affect a photographic plate, and make a gas through which they pass a conductor of electricity. In 1898 Schmidt discovered that thorium possesses similar properties. This power of emitting rays is called radio-activity, and substances which possess the power are said to be radio-active.

This property of uranium led to a careful ex- amination of a large number of minerals contain- ing this substance, and M. and Mme. Curie found that some of these, and notably some specimens of pitch-blende, were more radio-active than equal volumes of pure uranium, although only a fraction of these minerals consisted of uranium. This in- dicated that these minerals contained a substance or substances much more radio-active than uran- ium itself, and a systematic attempt was made to


RADIO-ACTIVE SUBSTANCES 141

isolate these substances. After a long investigation, conducted with marvellous skill and perseverance, M. and Mme. Curie, with the collaboration of MM. Bemont and Debierne, succeeded in establishing the existence of three new radio-active substances in pitch-blende : radium associated with the ba- rium in the mineral, and closely resembling it in its chemical properties ; polonium associated with the bismuth, and actinium with the thorium. They succeeded in isolating the first of these and deter- mined its combining weight, which was found to be 225. Its spectrum has been discovered and exam- ined by Demarcay. Neither polonium nor actinium has yet been isolated, nor have their spectra been observed. The activity of polonium has been found to be fugitive, dying away in some months after its preparation.

These radio-active substances are not confined to rare minerals. I have lately found that many specimens of water from deep wells contain a radio-active gas, and Elster and Geitel have found that a similar gas is contained in the soil.

These radio-active substances may be expected to be of the greatest possible assistance in the task of investigating problems dealing with the nature of the atom, and with the changes that go on in


142 ELECTRICITY AND MATTER

the atom from time to time. For the properties possessed by these substances are so marked as to make the detection of exceedingly minute quanti- ties of them a matter of comparative ease. The quantity of these substances which can be detected is to the corresponding amount of the other ele- ments which have to be detected by the ordinary methods of chemical analysis, in the proportion of a second to thousands of years. Thus, changes which would have to go on for almost geological epochs with the non-radio-active substances, be- fore they became large enough to be detected, could with radio-active substances prove appreci- able effects in the course of a few hours.

Character of the Radiation

Rutherford found that the radiation from uran ium, and it has subsequently been found that the same is true for thorium and radium, is made up of three distinct types which he calls the a, /3, and y radiations.

The a radiation is very easily absorbed, being unable to penetrate more than a few millimetres of air at atmospheric pressure, the /8 radiation is much more penetrating, while the y radiation is the most penetrating of all. Investigations of the


RADIO-ACTIVE SUBSTANCES 143

effects of magnetic and electric forces on these three types of radiation have shown that they are of entirely different characters. Becquerel showed that the ft rays were deflected by electric and mag- netic forces, the direction of the deflection show- ing that the rays carried a charge of negative elec- tricity. He determined, using the method described

in Chapter IV, the value of —, the ratio of the

m

charge to the mass of the carriers of the negative electricity ; he found that it was about 10T, and that the velocity for some of the rays was more than two- third s that of light. He thus proved that the ft rays consisted of corpuscles travelling at prodig- ious speeds.

The a rays are not nearly so easily deflected as the ft rays, but Rutherford has recently shown that they can be deflected, and the direction of deflection shows that they carry a, positive charge. He finds, and his measurements have been con

firmed by Des Coudres, that the ratio of — is 6 X

m

103, and the velocity of these particles is 2 X 10* centimetres per second. The value of — shows that the carriers of the positive electrification have


!44 ELECTRICITY AND MATTER

masses comparable with those of ordinary atoms ;

thus — for hydrogen is 104 and for helium 2.5 X m

103. The very high velocity with which these are shot out involves an enormous expenditure of en- ergy, a point to which we shall return later. One of the most interesting things about this result is

that the value of — shows that the atoms shot off m

are not the atoms of radium, indicating either that radium is a compound containing lighter elements or else that the atom of radium is disintegrating

into such elements. The value of — for the a

m

rays obtained by Rutherford and Des Coudres suggests the existence of a gas heavier than hy- drogen but lighter than helium. The y rays, as far as we know, are not deflected either by mag- netic or electric forces.

There is considerable resemblance between a radio-active substance and a substance emitting secondary radiation under the influence of Ront- gen rays : the secondary radiation is known to contain radiation of the ($ and y types ; and as part of the radiation is exceedingly easily absorbed, being unable to penetrate more than a millimetre or so of air at atmospheric pressure, it is possible


RADIO-ACTIVE SUBSTANCES 145

that closer investigation may show that a rays, i.e.j positively electrified particles, are present also. This analogy raises the question as to whether there may not, in the case of the body struck by the Rbntgen rays, be a liberation of energy such as we shall see occurs in the case of the radio-active substances, the energy emitted by the radiating substances being greater than the energy in the Rontgen rays falling upon it ; this excess of energy being derived from changes taking place in the atoms of the body exposed to the Rontgen rays. This point seems worthy of investigation, for it might lead to a way of doing by external agency what radio-active bodies can do spontane- ously, i.e., liberate the energy locked up in the atom.

Emanation from Radio-Active Substances

Rutherford proved that thorium emits some- thing which is radio-active and which is wafted about by currents of air as if it were a gas ; in order to avoid prejudging the question as to the physical state in which the substance given off by radium exists, Rutherford called it the " emana- tion." The emanation can pass through water or the strongest acid and can be raised to tempera-


146 ELECTRICITY AND MATTER

tures at which platinum is incandescent without suffering any loss of radio-activity. In this inertness it resembles the gases argon and helium, the latter of which is almost always found associated with thorium. The radio-activity of the thorium emana- tion is very transient, sinking to half its value in about one minute.

The Curies found that radium also gives off a radio-active emanation which is much more persistent than that given off by thorium, taking about four days to sink to half its activity.

There seems every reason for thinking that those emanations are radio-active matter in the gaseous form ; they can be wafted from one place to another by currents of air  ; like a gas they dif- fuse through a porous plug at a rate which shows that their density is very high. They diffuse gradually through air and other gases. The coeffi- cient of diffusion of the radium emanation through air has been measured by Rutherford and Miss Brooks and they concluded that the density of the emanation was about eighty. The emanation of radium has been liquefied by Rutherford and Soddy ; and I have, by the kindness of Professor Dewar, been able to liquefy the radio-active gas found in water from deep wells, which very


RADIO-ACTIVE SUBSTANCES 147

closely resembles the emanation and is quite possibly identical with it. In short the emana- tions seem to satisfy every test of the gaseous state that can be applied to them. It is true that they are not capable of detection by any chemical tests of the ordinary type, nor can they be detected by spectrum analysis, but this is only because they are present in very minute quantities — quantities far too small to be detected even by spectrum analysis, a method of detection which is exceedingly rough when compared with the elec- trical methods which we are able to employ for radio-active substances. It is not, I think, an ex- aggeration to say that it is possible to detect with certainty by the electrical method a quantity of a radio-active substance less than one-hundred-thou- sandth part of the least quantity which could be detected by spectrum analysis.

Each portion of a salt of radium or thorium is giving off the emanation, whether that portion be on the inside or the outside of the salt; the emanation coming from the interior of a salt, how- ever, does not escape into the air, but gets entangled in the salt and accumulates. If such a radio- active salt is dissolved in water, there is at first a great evolution of the emanation which has been


148 ELECTRICITY AND MATTER

stored up in the solid salt. The emanation can be extracted from the water either by boiling the water or bubbling air through it. The stored up emanation can also be driven off from salts in the solid state by raising them to a very high tem- perature.

Induced Radio-Activity

Kutherford discovered that substances exposed to the emanation from thorium become radio-active, and the Curies discovered almost simultaneously that the same property is possessed by the emana- tion from radium. This phenomenon is called in- duced radio-activity. The amount of induced radio-activity does not depend upon the nature of the substance on which it is induced ; thus, paper becomes as radio-active as metal when placed in contact with the emanations of thorium or radium.

The induced radio-activity is especially de- veloped on substances which are negatively elec- trified. Thus, if the emanation is contained in a closed vessel, in which a negatively electrified wire is placed, the induced radio-activity is concentrated on the negatively electrified wire, and this induced activity can be detected on negatively electrified


RADIO-ACTIVE SUBSTANCES ^49

bodies when it is too weak to be detected on un- electrified surfaces. The fact that the nature of the induced radio-activity does not depend on the substance in which it is induced points to its being due to a radio-active substance which is deposited from the emanation on substances with which it comes in contact.

Further evidence of this is afforded by an ex- periment made by Miss Gates, in which the in- duced radio-activity on a fine wire was, by raising it to incandescence, driven off the wire and de- posited on the surrounding surfaces. The induced radio-activity due to the thorium emanation is very different from that due to the radium emana- tion, for whereas the activity of the thorium ema- nation is so transient that it drops to half its value in one minute, the induced radio-activity due to it takes about eleven hours to fall in the same proportion. The emanation due to radium, which is much more lasting than the thorium emanation, taking about four days instead of one minute to fall to half its value, gives rise to a very much less durable induced radio-activity, one fall, ing to half its value in about forty minutes instead of, as in the case of thorium, eleven hours. The emanation due to actinium is said only to be active


150 ELECTRICITY AND MATTER

for a few seconds, but the induced radio-activity due to it seems to be nearly as permanent as that due to radium.

Separation of the Active Constituent from Thorium

Rutherford and Soddy, in a most interesting and important investigation, have shown that the radio-activity of thorium is due to the passage of the thorium into a form which they call T h X, which they showed could be separated from the rest of the thorium by chemical means. When this separation has been effected the thorium left be- hind is for a time deprived of most of its radio-activ- ity, which is now to be found in the T h X. The radio-activity of the thorium X slowly decays while that of the rest of the thorium increases until it has recovered its original activity. While this has been going on, the radio-activity of the Th JThas vanished. The time taken for the radio-activity of the T h X to die away to half its original value has been shown by Rutherford and Soddy to be equal to the time taken by the thorium from which the T 7i X has been separated to recover half its original activity. All these results support the view that the radio-active part of the thorium, the thorium X, is continually being produced from the


RADIO-ACTIVE SUBSTANCES 151

thorium itself ; so that if the activity of thorium X were permanent, the radio-activity of the tho- rium would continually increase. The radio-activ. ity of the thorium X, however, steadily dies away. This prevents the unlimited increase of the radio- activity of the mixture, which will reach a steady value when the increase in the radio-activity due to the production of fresh T h X is balanced by the decay in the activity of that already produced. The question arises as to what becomes of the Tli X and the emanation when they have lost their radio-activity. This dead ThX, as we may call it, is accumulating all the time in the thorium; but inasmuch as it has lost its radio-activity, we have only the ordinary methods of chemical analy- sis to rely upon, and as these are almost infinitely less delicate than the tests we can apply to radio- active substances, it might take almost geological epochs to accumulate enough of the dead TJiX to make detection possible by chemical analysis. It seems possible that a careful examination of the minerals in which thorium and radium occur might yield important information. It is remark- able that helium is almost invariably a constitu- ent of these minerals.

You will have noticed how closely, as pointed


152 ELECTRICITY AND MATTER

out by Rutherford and Soddy, the production of radio-activity seems connected with changes tak- ing place in the radio-active substance. Thus, to take the case of thorium, which is the one on which we have the fullest information,we have first the change of thorium into thorium X, then the change of the thorium X into the emanation and the substance forming the a rays. The radio- activity of the emanation is accompanied by a fur- ther transformation, one of the products being the substance which produces induced radio-activity.

On this view the substance while radio-active is continually being transformed from one state to another. These transformations may be ac- companied by the liberation of sufficient energy to supply that carried off by the rays it emits while radio-active. The very large amount of energy emitted by radio-active substances is strik- ingly shown by some recent experiments of the Curies on the salts of radium. They find that those salts give out so much energy that the absorption of this by the salt itself is sufficent to keep the temperature of the salt permanently above that of the air by a very appreciable amount — in one of their experiments as much as 1.5° C. It appears from their measurements that a gram of radium


RADIO-ACTIVE SUBSTANCES 153

gives out enough energy per hour to raise the temperature of its own weight of water from the freezing to the boiling point. This evolution of energy goes on uninterruptedly and apparently without diminution. If, however, the views we have just explained are true, this energy arises from the transformation of radium into other forms of matter, and its evolution must cease when the stock of radium is exhausted ; unless, indeed, this stock is continually being replenished by the transformation of other chemical elements into radium.

We may make a rough guess as to the probable duration of a sample of radium by combining the result that a gram of radium gives out 100 calories per hour with Rutherford's result that the a rays are particles having masses comparable with the mass of an atom of hydrogen projected with a velocity of about 2 X 109 centimetres per second  ; for let us suppose that the heat measured by the Curies is due to the bombardment of the radium salt by these particles, and to get a superior limit to the time the radium will last, let us make the assumption that the whole of the mass of radium gets transformed into the a par- ticles (as a matter of fact we know that the emana-


154 ELECTRICITY AND MATTER

tion is produced as well as the a particles). Let a? be the life in hours of a grain of radium  ; then since the gram emits per hour 100 calories, or 4.2 X 109 ergs, the amount of energy emitted by the radium during its life is x X 4.2 X 109 ergs. If jVis the number of a particles emitted in this time, m the mass of one of them in grams, v the velocity, then the energy in the a particles is J Nmv*, but this is to be equal to a? X 4.2 X 10* ergs, hence £ Nm v* = x X 4.2 X 109 ; but if the gram of radium is converted into the a particles, Nm = 1, and by Rutherford's experiments v = 2

4 V 1018 10' X 10', hence we have «  = | Jg * ^ = g

hours, or about 50,000 years.

From this estimate we should expect the life of a piece of radium to be of the order of 50,000 years. This result shows that we could not expect to detect any measurable changes in the space of a few months. In the course of its life the gram of radium will have given out about 5 X 1010 calories, a result which shows that if this energy is derived from transformations in the state of the radium, the energy developed in these transformations must be on a very much greater scale than that developed in any known chemical


RADIO-ACTIVE SUBSTANCES 155

reactions. On the view we have taken the differ- ence between the case of radium and that of or- dinary chemical reactions is that in the latter the changes are molecular, while in the case of ra- dium the changes are atomic, being of the nature of a decomposition of the elements. The example given on page (111) shows how large an amount of energy may be stored up in the atom if we re- gard it as built up of a number of corpuscles.

We may, I think, get some light on the processes going on in radium by considering the behavior of a model atom of the kind described on page 124, and which may be typified by the case of the corpuscles which when rotating with a high velocity are stable when arranged in a certain way, which arrangement becomes unstable when the energy sinks below a certain value and is succeeded by another configuration. A top spin- ning about a vertical axis is another model of the same type. This is stable when in a vertical position if the kinetic energy due to its rotation exceeds a certain value. If this energy were gradually to decrease, then, when it reached the critical value, the top would become unstable and would fall down, and in so doing would give a considerable amount of kinetic energy.


156 ELECTRICITY AND MATTER

Let us follow, then, the behavior of an atom of this type, i.e., one which is stable in one configura- tion of steady motion when the kinetic energy of the corpuscles exceeds a certain value, but be- comes unstable and passes into a different config- uration when the kinetic energy sinks below that value. Suppose now that the atom starts with an amount of kinetic energy well above the critical value, the kinetic energy will decrease in conse- quence of the radiation from the rapidly moving corpuscles ; but as long as the motion remains steady the rate of decrease will be exceedingly slow, and it may be thousands of years before the energy approaches the critical value. When it gets close to this value, the motion will be very easily disturbed and there will probably be considerable departure from the configuration for steady motion accompanied by a great increase in the rate at which kinetic energy is loss by radiation. The atom now emits a much greater number of rays and the kinetic energy rapidly approaches the critical value ; when it reaches this value the crash comes, the original configuration is broken up, there is a great decrease in the potential energy of the sys- tem accompanied by an equal increase in the kinetic energy of the corpuscles. The increase in


RADIO-ACTIVE SUBSTANCES 157

the velocity of the corpuscles may cause the dis- ruption of the atom into two or more systems, cor- responding to the emission of the a rays and the emanation.

If the emanation is an atom of the same type as the original atom, i.e., one whose configuration for steady motion depends on its kinetic energy, the process is repeated for the emanation, but in a very much shorter time, and is repeated again for the various radio-active substances, such as the induced radio-active substance formed out of the emanation.

We have regarded the energy emitted by radium and other radio-active substances as de- rived from an internal source, i.e., changes in the constitution of the atom ; as changes of this kind have not hitherto been recognized, it is desirable to discuss the question of other possible sources of this energy. One source which at once sug- gests itself is external to the radium. We might suppose that the radium obtained its energy by absorbing some form o£ radiation which is passing through all bodies on the surface of the earth, but which is not absorbed to any extent by any but those which are radio-active. This radiation must be of a very penetrating character, for radium


158 ELECTRICITY AND MATTER

retains its activity when surrounded by thick lead or when placed in a deep cellar. We are familiar with forms of Rontgen rays, and of rays given out by radium itself, which can produce appreci- able effects after passing through several inches of lead, so that the idea of the existence of very pene- trating radiation does not seem so improbable as it would have done a few years ago. It is interest- ing to remember that very penetrating radiation was introduced by Le Sage more than a century ago to explain gravitation. Le Sage supposed that the universe was thronged with exceedingly small particles moving with very high velocities. He called these ultra-mundane corpuscles and as- sumed that they were so penetrating that they could pass through masses as large as the sun or the planets without suffering more than a very slight absorption. They were, however, absorbed to a slight extent and gave up to the bodies through which they passed a small fraction of their momentum. If the direction of the ultra- mundane corpuscles passing through a body were uniformly distributed, the momentum communi- cated by them to the body would not tend to move it in one direction rather than another, so that a body A alone in the universe and exposed to


RADIO-ACTIVE SUBSTANCES 159

bombardment by Le Sage's corpuscles would re- main at rest ; if, however, there is a second body B in the neighborhood of A, B will shield off from A some of the corpuscles moving in the direction B A  ; thus, A will not receive as much momentum in this direction as it did when it was alone in the field, but in the latter case it only re- ceived enough momentum in this direction to keep it in equilibrium ; hence, when B is present, the momentum in the opposite direction will get the upper hand so that A will move in the direction, A B, i.e., will be attracted to B. Maxwell pointed out that this transference of momentum from Le Sage's corpuscles to the body through which they were passing involved the loss of kinetic energy by the corpuscles  ; and that if the loss of momen- tum were sufficient to account for gravitation, the kinetic energy lost by the ultra-mundane cor- puscles would be sufficient, if converted into heat, to keep the gravitating body white hot. The fact that all bodies are not white hot was urged by Maxwell as an argument against Le Sage's theory. It is not necessary, however, to suppose that the energy of the corpuscles is transformed into heat ; we might imagine it transformed into a very penetrating radiation which might escape


1(50 ELECTRICITY AND MATTER

from the gravitating body. A simple calculation will show that the amount of kinetic energy transformed per second in each gram of the gravitating body must be enormously greater than that given out in the same time by one gram of radium.

"We have seen in the first chapter that waves of electric and magnetic force possess momentum in their direction of propagation; we might there- fore replace Le Sage's corpuscles by very pene- trating Rontgen rays. Those, if absorbed, would give up momentum to the bodies through which they pass, and similar consideration to those given by Le Sage would show that two bodies would attract each other inversely as the square of the distance between them. If the absorption of these waves per unit volume depended only upon, and was proportional to, the density, the attraction between the bodies would be directly proportional to the product of their masses. It ought to be mentioned that on this view any changes in gravitation would be propagated with the velocity of light; whereas, astronomers be- lieve they have established that it travels with a very much greater velocity.

As in the case of Le Sage's corpuscles, the loss


RADIO-ACTIVE SUBSTANCES IQ±

of momentum by the Rontgen rays would be ac- companied by a loss of energy; for each unit of momentum lost v units of energy would be lost, v being the velocity of light. If this energy were transformed into that of rays of the same type as the incident rays, a little reflection will show that he absorption of the rays would not produce gravitational attraction. To get such attraction the transformed rays must be of a more pene- trating type than the original rays. Again, as in the case of Le Sage's corpuscles, the absorp- tion of energy from these rays, if they are the cause of gravitation, must be enormous — so great that the energy emitted by radium would be but an exceedingly small fraction of the energy being transformed within it. From these considerations I think that the magnitude of the energy radiated from radium is not a valid argument against the energy being derived from radiation. The reason which induces me to think that the source of the energy is in the atom of radium itself and not ex- ternal to it is that the radio-activity of substances is, in all cases in which we have been able to local- ize it, a transient property. No substance goes on being radio-active for very long. It may be asked how can this statement be reconciled with the fact


162 ELECTRICITY AND MATTER

that thorium and radium keep up their activity without any appreciable falling off with time. The answer to this is that, as Rutherford and Soddy have shown in the case of thorium, it is only an exceedingly small fraction of the mass which is at any one time radio-active, and that this radio-active portion loses its activity in a few hours, and has to be replaced by a fresh supply from the non-radio-active thorium. Take any of the radio-active substances we have described, the ThX, the emanations from thorium or radium, the substance which produces induced radio- activity, all these are active for at the most a few days and then lose this property. This is what we should expect on the view that the source of the radio-activity is a change in the atom ; it is not what we should expect if the source were ex- ternal radiation.

Retrieved from "https://msterms.org/wiki/index.php?title=Thomson_1904&oldid=21975"