166     Ferric  Citro- Phosphate  and  Citro-Pyrophosphate.  { ^^ipru^iggl.'''^'' 
oxide  must  be  an  anhydrate.  Nevertheless  the  statement  is  unceas- 
ingly made  that  these  substances  are  anhydrides.  Properly  speaking, 
anhydrides  would  be  the  residues  remaining  after  the  removal  of  the 
typic  hydrogen  of  the  hydrates.  But  these  residues,  at  the  instant  of 
their  liberation,  break  up  into  the  corresponding  basic  or  acid  oxides 
and  free  oxygen.  The  only  exceptions,  and  but  apparent  at  that,  are 
the  halogen  radicles  which,  also  freed  in  doublets,  immediately  combine 
to  generate  their  respective  elementary  molecules. 
The  category  of  hydrates  just  discussed  is  that  embraced  within  the 
scope  of  chemism  j^roper  and,  therefore,  representable  by  the  full  atom- 
icity of  the  elements  involved.  There  is,  however,  a  widening  series 
of  undoubted  chemical  compounds  containing  water  whose  molecular 
formulas  can  only  be  expressed  by  the  assumption  of  the  unlimited 
subdivision  of  the  atomic  bonds,  or  the  principle  of  ultravalence. 
Consistently  these  are  also  hydrates,  and  yet  their  anhydrates  have 
nothing  in  common  with  those  of  the  preceding  order.  These  second- 
ary hydrates  are  formed  by  the  superposition  of  apparently  arbitrary 
numbers  of  water  molecules  upon  the  hydrates  proper  or  primary 
hydrates,  salts,  etc.  The  separation  or  increase  of  this  water  eifects 
less  radical  changes  in  the  residues,  since  its  amount  is  readily  varied 
by  slight  alterations  in  the  attendant  physical  conditions ;  as,  for 
instance,  degree  of  concentration,  temperature,  and  tension  of  the 
solution  from  which  the  hydrate  formed,  and  the  effects  of  the  atmos- 
phere on  exposure. 
It  is  very  essential  that  a  direct  and  clear  distinction  should  be  made 
in  the  nomenclature  of  these  compounds.  The  first  class  may,  there- 
fore, be  styled  primary  hydrates  or  simply  hydrates,  and  the  corre- 
sponding anhydrates  be  designated  as  anhydric  oxides,  etc.  The  second 
class  may  be  termed  secondary  hydrates  or  hydrites,  and  the  corre- 
sponding anhydrites  be  called  anhydrous  salts,  etc. 
Another  false  impression  prevails  in  assuming  so-called  water  of 
crystallization.  This  term  has  no  meaning  in  relation  to  the  act  of 
crystallizing,  since  innumerable  salts  crystallize  without  combining 
with  water.  The  absorption  of  water  in  this  connection  is  not  an 
effect  of  crystallization  or  vice  versa,  but  merely  a  concomitant  phe- 
nomenon. 
In  physics  it  has  been  found  convenient  to  assume  a  crystallic  force, 
but  it  is  questionable  if  such  a  mode  of  motion  exists  as  a  distinct 
dynamic  form.    The  crystalline  condition  is  doubtless  due  to  the  fact 
