Atomic  Hypothesis  and  Dissected  {Structural)  Formulae.     255 
similarly  the  true  formula  N2  O2,  the  gas  as  usually  met  with 
being  iu  this  quasi -dissociated  condition.  The  existence  of  the 
chlorides  and  oxychlorides  &c.  of  vanadium  and  wolfram,  recently 
investigated  by  Roscoe*,  however,  shows  that  dissociation  can- 
not be  considered  to  take  place  in  all  the  apparent  exceptions  to 
this  very  general  rule. 
23.  Observation  shows  that  the  following  rule  applies  to  all 
formulae  with  extremely  few  exceptions : — If  the  suffix  of  each 
symbol  in  the  formula  be  multiplied  by  the  valency  which  the 
symbol  exhibits  in  the  majority  of  its  compounds  (e.  g.  2  for 
oxygen,  4  for  carbon),  the  sum  of  the  products  thus  obtained  is  an 
even  number;  from  which  it  follows  that  the  sum  of  the  suffixes 
of  the  perissad  symbols  in  a  formula  is  always  an  even  number 
(Gerhardt's  (  Law  of  Even  lumbers'). 
24r.  Before  it  can  be  asserted  that  two  bodies  contain  a  com- 
mon radical,  a  reaction  must  be  found  whereby  one  of  them  is 
convertible  into  the  other ;  thus  water,  caustic  soda,  and  acetic 
acid  are  all  said  to  coutain  the  radical  hydroxyl  (OH),  because 
of  the  reactions 
2H20        +2Na  =  H2     +  2NaOH, 
NaOH        +C2  H3  0  Cl=NaCl  +  C2  H3  0  .  OH, 
C2H30C1  +  H20  =HC1  +  C2H3O.OH, 
and  not  because  the  formulae  H2  0,  NaOH,  C2  H4  O2  all  contain 
the  symbols  0  and  H.  Conversely  C2  H4  0  (aldehyde)  is  said 
not  to  contain  hydroxyl  (although  the  symbols  0  and  H  are  pre- 
sent in  the  formula),  because  no  reaction  of  this  kind  is  known. 
Reactions  such  as  the  above  are  alluded  to  and  expressed  by 
dividing  the  formula  of  each  generating  substance  into  two  por- 
tions or  radicals,  which  are  separated  from  one  another  by  dots, 
lines,  parentheses,  &c.  Thus  the  action  of  acetyl  chloride  on 
alcohol,  producing  hydrogen  chloride  and  acetic  ether,  is  repre- 
sented equationally  thus : — 
C2H3O.Cl  +  C2H\OH      =H.C1  +  C2H30  . O.C2H\ 
(C2H30)C1  +  (C2H5)OH       =HC1     +  (C2H30)0(C2H5), 
C2  h3  0— CI  +  C2H5— 0— II  =  H— CI  4-  C2H30— 0— C2H5, 
C2H301  K\Q         _HH >      C2H301Q. 
CI  J      +C2H5JU         ~C1J+C2H5     Ju> 
and  these  reactions  (and  others  of  similar  character)  are  referred 
to  when  it  is  said  that  acetic  ether  has  the  Dissected  Formula 
C2H30— 0— C2H5. 
AVhen  several  reactions  of  the  same  substance  are  compared, 
*  Phil.  Trans.  1868  and  1869;  Chem.  News,  vol.  xxv.  p.  61. 
