of  the  recent  attack  on  the  Atomic  Theory."  511 
sition,  inasmuch  as  the  motion  which  appears  or  disappears 
during  a  reaction  can  only  be  considered  to  be  derived  from,  or 
imparted  to,  the  atoms  themselves;  i.e.  intramolecular  motion  of 
variable  amount  must  exist.  Since,  then,  the  atoms  must  have 
a  limited  amount  of  motion  (vibratory,  rotatory,  &c.)  among 
themselves,  what  are  the  differences  in  the  rate  or  kind  of  mo- 
tion corresponding  to  the  chemical  differences  between  (for  ex- 
ample) the  three  kinds  of  hydrogen  existing  in  ethylic  alcohol 
in  the  groups  CH3,  CH2,  and  OH  ?  Before  the  atomic  hypothesis 
can  be  said  to  explain  isomerism,  some  connexion  must  be  made 
out  between  these  different  rates  or  kinds  of  intramolecular  mo- 
tion and  the  chemical  and  other  properties  of  the  isomerides; 
for  isomerides  differ  not  only  in  mechanical  properties,  such  as 
specific  gravity,  melting-point,  &c,  and  in  chemical  reactions, 
but  also  in  the  amount  of  heat  given  out  on  combustion  of  equal 
weights  of  the  substances  respectively;  i.  e.  each  isomeride  repre- 
sents a  different  amount  of  potential  energy ;  and  the  same  holds 
in  the  case  of  allotropic  modifications,  e.  y.  the  carbon  and  phos- 
phorus allotropes.  The  mere  hypothesis  of  the  existence  of  atoms 
is  insufficient  to  explain  these  observed  facts ;  whilst,  even  if 
supplemented  by  a  number  of  further  hypotheses  as  to  the  rate 
and  character  of  intramolecular  motions  in  given  instances,  a 
satisfactory  explanation  of  chemical  phenomena  by  these  means 
does  not  seem  probable.  Thus,  granting  that  the  substitution 
of  a  chlorine  atom  for  a  hydrogen  atom  in  one  molecule  is  ac- 
companied by  a  given  change  in  the  motion,  either  of  that  atom 
alone,  or  in  the  motion  of  the  other  atoms  in  the  molecule  as 
well,  it  does  not  appear  that  it  will  be  possible  to  predict  what 
will  be  the  changes  of  motion  that  take  place  when  the  substi- 
tution of  the  chlorine  atom  for  the  hydrogen  atom  occurs  in  an- 
other molecule,  or  even  in  a  different  portion  of  the  same  mole- 
cule. For  instance,  granting  that  the  substitution  of  chlorine  for 
hydrogen  is  attended  by  a  given  change  of  motion,  the  inverse 
substitution  must  be  accompanied  by  the  opposite  change. 
Now,  when  a  molecule  of  hydrogen  is  acted  on  by  a  molecule  of 
chlorine,  in  one  molecule  an  atom  of  hydrogen  is  replaced  by 
one  of  chlorine,  and  in  the  other  an  atom  of  chlorine  is  replaced 
by  one  of  hydrogen.  These  complementary  changes,  therefore, 
should  cause,  on  the  whole,  no  difference  as  to  the  ultimate 
amount  of  motion  in  the  two  original  and  the  two  resulting  mo- 
lecules; but  as  a  matter  of  fact  there  is  such  a  difference  ob- 
served (t.  e.  potential  energy  becomes  actual)  when  hydrogen  and 
chlorine  act  on  one  another  in  accordance  with  the  equation 
H2  +  C12=HC1  +  C1H. 
Again,  to  take  a  simple  case,  when  chlorine  acts  on  hexyl- 
