of  the  recent  attack  on  the  Atomic  Theory."  509 
cumstance  that  experimental  generalizations  can  only  express 
observed  facts  with  approximate  accuracy  is  to  be  taken  as  evi- 
dence of  their  worthlessness,  the  sooner  we  cease  to  waste  time 
by  attempting  to  systematize  phenomena  and  show  their  corre- 
lation by  the  employment  of  experimental  generalizations  the 
better.  Even  astronomical  observations  do  not  always  exhibit  a 
mathematical  agreement  with  the  generalizations  that  express 
them  with  a  great  degree  of  approximation ;  and  if  the  genera- 
lizations are  "  worthless,"  what  becomes  of  the  theory  of  gravi- 
tation (t.  e.  the  developed  hypothesis)  which  rests  on  them  ? 
The  remarks  of  Mr.  Atkinson  on  the  applicability  of  the  term 
"atom"  to  express  the  observed  facts  that  16  parts  by  weight 
of  oxygen  or  12  of  carbon  is  the  minimum  relative  weight  or 
smallest  proportion  in  which  it  exists  in  compounds  have  been 
already  discussed.  The  term  atom  is  not  used  in  the  sense  of  a 
proportion  or  relative  number  anywhere  throughout  my  paper ; 
it  is  used  in  reference  to  a  particular  hypothesis,  which  is  not 
identical  with  the  facts  from  which  it  takes  its  origin.  Apropos, 
Mr.  Atkinson  considers  that  "  Dr.  Williamson  would  probably 
be  astonished  to  learn  that  he  looked  upon  an  atom  as  a  pure 
number."  Now  it  is  not  asserted  that  Dr.  Williamson  always 
employs  the  term  in  this  particular  defined  sense;  but  when  he 
distinctly  states  that  the  notion  of  an  atom  being  indivisible  is  a 
question  not  raised  in  chemistry  by  any  evidence  whatever,  and 
that  he  knows  not  whether  atoms  are  really  ultimate  atoms  at 
all,  whether  spherical,  regular,  or  irregular,  or  whether  they  are 
not  vortices,  it  is  somewhat  difficult  to  see  in  what  connexion 
Dr.  Williamson  views  them  other  than  as  simple  numerical 
quantities.  Mr.  Atkinson  does  not  seem  to  be  quite  clear  in  his 
own  mind  as  to  the  meaning  of  the  phrase  "  being  possessed  of 
dimensions  in  space,  mass,  and  time."  Perhaps  he  is  not  aware 
that  these  terms  are  fundamental  conceptions  involved  in  any 
proposition  which  relates  to  energy  or  to  force,  and  that  therefore 
they  are  necessarily  connected  with  any  question  in  which  weight 
is  involved. 
Mr.  Atkinson  inquires  what  is  the  meaning  of  an  equivalent 
quantity  of  a  radical,  which  radical  is  defined  to  be  one  or  more 
symbols  and  suffixes  transferable  from  one  formula  to  another; 
and  he  appears  to  be  amused  at  the  idea  of  valency  being  a  func- 
tion of  an  assemblage  of  symbols.  Nevertheless  this  is  precisely 
what  valency  is ;  it  is  a  term  expressing  certain  differences  be- 
tween symbols,  such  differences  being  deduced  from  the  symbolic 
representation  of  observed  reactions.  Thus,  to  take  the  question 
of  so-called  direct  and  indirect  combination  referred  to  by  Mr. 
Atkinson,  experiment  shows  that  caustic  potash  is  denoted  by  the 
formula  OKH   (KOH,   KHO,   &c.)  ;    if  this  substance,  when 
