606  'The  Constitution  of  the  Atom. 
electron.  I  suggest  that  the  objection  applies  equally  to  the 
"  orbital  motion  "  interpretation. 
Lord  Rayleigh's  objection  to  the  statical  interpretation, 
that  the  theoretically  calculated  spectrum  series  would  be  for 
p2  and  not  for  p,  and  that  they  would  apparently  be  too 
complex  to  agree  with  the  observed  facts,  remains,  and  seems, 
it  is  true,  to  be  almost  insuperable;  but  is  the  case  any  worse 
than  that  which  mig lit  have  been  made  out  against  the  atomic 
theory  a  century  ago  from  the  fact  of  the  ratios  of  specific 
heats  of  gases  forming  the  simple  series  1+  ^— — ,  in  spite 
of  the  enormous  complexity  of  the  atomic  conception  of 
matter  ? 
Evidence,  which  seems  to  me  to  have  great  weight,  can  be 
derived  from  considerations  of  the  partition  of  energy.  The 
value  of  Dulong  and  Petit' s  constant  shows  that  those  parts 
of  the  energy  which  vary  with  the  temperature  are  fully 
accounted  for  by  the  potential  and  kinetic  energies  of  the 
atom  regarded  as  a  point;  and  the  same  is  shown  by  the  fact 
that  7  for  monatomic  gases  has  the  value  If-.  Thus  the 
energies  of  the  degrees  of  freedom  which  represent  rotation 
or  internal  motions  of  the  atom,  are  either  infinitesimal  or  are 
incapable  of  variation  under  the  ordinary  interatomic  forces. 
Thus  if  electrons  are  describing  orbits,  the  planes  of  these 
orbits  must  remain  almost  or  completely  fixed  in  space.  If 
we  rotate  a  body  in  the  hand  the  atoms  must  maintain  their 
directions  in  space,  as  indeed,  on  this  view  of  the  atom,  would 
be  shown  also  by  the  fact  of  our  not  experiencing  any  gyro- 
static  couple.  But  if  the  body  is  a  crystal,  its  optical  constants 
and  axes  do  not  change  by  rotation — as  they  surely  would  if 
all  the  atoms  rotated  their  planes  relatively  to  the  body  as  a 
whole — and  if  the  body  is  a  permanent  magnet,  its  magnetic 
properties  and  constants  do  not  change. 
A  similar  dilemma  is  soon  reached  with  reference  to  the 
velocity  of  rotation.  The  Zeeman  effect  shows  that  this  can 
be  altered  even  by  the  fields  available  in  the  laboratory,  while 
the  value  of  y  would  show  that  it  is  not  altered  by  the  much 
more  effective  magnetic  fields  which  ought  (on  the  "  orbital 
motion  "  theory)  to  be  found  at  collisions  of  molecules. 
Finally,  may  I  mention  that  in  the  Phil.  Mag.  for  Nov. 
1901,  I  attacked  a  problem  similar  in  many  respects  to  that 
which  forms  the  main  substance  of  Lord  Rayleigh's  paper  ? 
The  actual  premises  upon  which  I  worked  were  different  from 
those  of  Lord  Rayleigh — his  positive  sphere  being  repre- 
sented, in  my  work,  by  a  crowd  of  positive  electrons,  and 
the  definiteness  of  structure,  which  he  obtains  bv  regarding 
