﻿484 
  F. 
  W. 
  Clarke 
  — 
  Constitution 
  of 
  Melilite 
  and 
  Gehlenite. 
  

  

  location 
  in 
  the 
  following 
  paragraphs. 
  The 
  acids 
  themselves 
  

   are 
  not 
  definitely 
  known 
  ; 
  but 
  the 
  first 
  of 
  the 
  series 
  may 
  be 
  

   represented 
  by 
  kryptotile, 
  although 
  that 
  is 
  not 
  certain. 
  

  

  In 
  the 
  present 
  discussion 
  we 
  have 
  three 
  fundamental 
  com- 
  

   pounds, 
  corresponding 
  with 
  the 
  three 
  acids. 
  

  

  Anorthite 
  is 
  normal 
  calcium 
  trialate, 
  and 
  from 
  it 
  the 
  basic 
  

   silicate 
  meionite 
  is 
  derived. 
  

  

  Garnet 
  is 
  normal 
  calcium 
  dialate, 
  with 
  sarcolite 
  as 
  an 
  iso- 
  

   meric 
  form. 
  Prehnite 
  is 
  an 
  acid, 
  and 
  zoisite 
  a 
  basic 
  derivative 
  

   of 
  garnet. 
  

  

  The 
  melilite 
  silicate 
  is 
  normal 
  calcium 
  unalate, 
  with 
  vesuvi- 
  

   anite 
  and 
  the 
  gehlenite 
  silicate 
  as 
  basic 
  derivatives. 
  

  

  These 
  three 
  fundamental 
  silicates 
  seem 
  to 
  undergo 
  altera- 
  

   tion 
  along 
  parallel 
  lines. 
  Anorthite 
  is 
  altered 
  to 
  muscovite, 
  a 
  

   trialate, 
  garnet 
  into 
  other 
  dialates, 
  such 
  as 
  biotite 
  and 
  the 
  cor- 
  

   responding 
  chlorites 
  ; 
  and 
  melilite 
  probably 
  follows 
  the 
  same 
  

   rule. 
  In 
  short, 
  nearly 
  all 
  of 
  the 
  orthosilicates 
  which 
  are 
  char- 
  

   acterized 
  as 
  aluminous 
  can 
  be 
  represented 
  as 
  normal, 
  acid, 
  or 
  

   basic 
  salts 
  of 
  the 
  three 
  alumo-silicic 
  acids. 
  This 
  rule 
  has 
  been 
  

   sufficiently 
  discussed 
  in 
  my 
  previous 
  publications, 
  and 
  needs 
  no 
  

   farther 
  development 
  here. 
  

  

  It 
  is 
  interesting 
  to 
  note 
  that 
  in 
  abundance 
  and 
  stability 
  the 
  

   trialates 
  outrank 
  the 
  other 
  two 
  series. 
  Anorthite 
  is 
  the 
  most 
  

   abundant 
  of 
  the 
  three 
  calcium 
  salts, 
  and 
  is 
  the 
  only 
  one 
  which 
  

   can 
  be 
  recrystallized 
  from 
  fusion. 
  Its 
  corresponding 
  mica, 
  

   muscovite, 
  is 
  also 
  the 
  most 
  abundant 
  and 
  the 
  least 
  alterable. 
  

   Garnet, 
  upon 
  fusion, 
  breaks 
  down 
  into 
  a 
  mixture 
  of 
  silicates, 
  

   and 
  anorthite 
  is 
  commonly 
  one 
  of 
  them. 
  The 
  dialate 
  mica, 
  

   moreover, 
  is 
  easily 
  alterable. 
  As 
  for 
  the 
  pure 
  melilite 
  silicate, 
  

   its 
  behavior 
  as 
  regards 
  stability 
  is 
  not 
  known, 
  but 
  it 
  is 
  the 
  

   least 
  abundant 
  of 
  the 
  three 
  calcium 
  compounds. 
  Nephelite, 
  

   the 
  sodium 
  trialate, 
  is 
  very 
  stable, 
  easily 
  synthesized, 
  and 
  alter- 
  

   able 
  into 
  muscovite. 
  The 
  normal 
  sodium 
  salts 
  of 
  the 
  other 
  

   alumosilicic 
  acids 
  are 
  not 
  known. 
  Their 
  high 
  proportion 
  of 
  

   sodium 
  would 
  probably 
  render 
  them 
  too 
  easily 
  alterable 
  for 
  

   permanence 
  under 
  geochemical 
  conditions. 
  The 
  commonest 
  

   unalates 
  seem 
  to 
  be 
  among 
  the 
  chloritic 
  minerals, 
  which 
  are 
  

   basic 
  salts 
  and 
  easily 
  decomposable. 
  

  

  In 
  conclusion 
  a 
  word 
  of 
  caution 
  is 
  necessary. 
  Formulae 
  such 
  

   as 
  have 
  been 
  developed 
  in 
  this 
  paper 
  are 
  not 
  to 
  be 
  taken 
  as 
  

   absolute 
  representations 
  of 
  molecular 
  structure. 
  They 
  are 
  

   merely 
  graphic 
  expressions 
  of 
  known 
  relations, 
  and 
  as 
  such 
  are 
  

   legitimate 
  and 
  useful. 
  They 
  show 
  the 
  elementary 
  atoms, 
  more- 
  

   over, 
  as 
  arranged 
  upon 
  a 
  plane 
  surface, 
  but 
  the 
  true 
  arrange- 
  

   ment, 
  the 
  real 
  molecular 
  structure, 
  should 
  be 
  tridimensional. 
  

   Such 
  an 
  arrangement, 
  which 
  might 
  correlate 
  chemical 
  relations 
  

   with 
  crystalline 
  form, 
  does 
  not 
  seem 
  to 
  be 
  as 
  yet 
  possible. 
  It 
  

   may 
  be 
  developed 
  in 
  the 
  future. 
  

  

  