﻿Brackett 
  and 
  Williams 
  — 
  Newtonite 
  and 
  Rector 
  ite. 
  19 
  

  

  analyses 
  so 
  that 
  the 
  discrepancy 
  between 
  them 
  and 
  the 
  cal- 
  

   culated 
  formula 
  may 
  well 
  be 
  ascribed 
  to 
  this 
  cause. 
  There 
  

   was, 
  however, 
  no 
  mineral 
  detected 
  which 
  would 
  account 
  for 
  

   the 
  relatively 
  large 
  amount 
  of 
  alkali 
  shown 
  by 
  the 
  analysis, 
  

   and 
  it 
  is 
  possible 
  that 
  the 
  soda 
  should 
  be 
  considered 
  as 
  replac- 
  

   ing 
  some 
  of 
  the 
  water 
  and 
  be 
  brought 
  into 
  the 
  formula. 
  Fur- 
  

   ther 
  investigation 
  will 
  probably 
  throw 
  some 
  light 
  on 
  this 
  point. 
  

   In 
  view 
  of 
  the 
  relatively 
  large 
  quantities 
  of 
  quartz 
  of 
  both 
  

   macroscopic 
  and 
  microscopic 
  dimensions, 
  which 
  have 
  been 
  

   observed 
  intermixed 
  with 
  the 
  rectorite, 
  it 
  may 
  be 
  allowable 
  to 
  

   consider 
  the 
  excess 
  of 
  silica 
  found 
  in 
  the 
  analyses 
  as 
  due 
  prin- 
  

   cipally 
  to 
  this 
  cause. 
  By 
  recalculating 
  the 
  analysis 
  after 
  de- 
  

   ducting 
  just 
  enough 
  silica 
  to 
  bring 
  that 
  constituent 
  down 
  to 
  

   the 
  theoretical 
  amount, 
  the 
  following 
  percentages 
  are 
  obtained 
  : 
  

  

  Theoretical 
  for 
  

   Vc. 
  YIc. 
  Alo0 
  3 
  . 
  2Si0 
  2 
  . 
  H 
  2 
  0. 
  

  

  Si0 
  2 
  __. 
  49-99 
  49-99.. 
  _..49'99 
  

  

  Al 
  2 
  O 
  s 
  41-26 
  41-08 
  42-52 
  

  

  HO 
  8-75 
  8*93 
  7'49 
  

  

  Total 
  100-00 
  100-00 
  100-00 
  

  

  In 
  order 
  to 
  determine 
  whether 
  or 
  not 
  the 
  soda 
  found 
  in 
  the 
  

   analyses 
  really 
  belonged 
  to 
  the 
  rectorite, 
  the 
  following 
  experi- 
  

   ment 
  was 
  made. 
  The 
  mineral, 
  in 
  small 
  flakes, 
  was 
  digested 
  

   with 
  concentrated 
  hydrochloric 
  acid 
  for 
  two 
  hours 
  on 
  a 
  sand 
  

   bath. 
  It 
  was 
  then 
  washed 
  and 
  filtered, 
  and 
  the 
  residue 
  was 
  

   boiled 
  with 
  sodic 
  carbonate 
  in 
  order 
  to 
  remove 
  any 
  separated 
  

   silica. 
  The 
  remaining 
  substance 
  was 
  then 
  washed 
  with 
  water, 
  

   hydrochloric 
  acid, 
  and 
  again 
  with 
  water, 
  and 
  was 
  finally 
  

   heated 
  before 
  the 
  blast 
  lamp. 
  A 
  portion 
  of 
  this 
  dried 
  and 
  

   purified 
  material 
  was 
  then 
  analyzed 
  with 
  the 
  following 
  result 
  : 
  

  

  VII. 
  

  

  Si0 
  2 
  _.. 
  57-10 
  

  

  AlO 
  4053 
  

  

  Sum 
  ._ 
  97*63 
  

  

  Impurities 
  (undetermined) 
  2*37 
  

  

  Total 
  .100-00 
  

  

  It 
  appears 
  from 
  this 
  that 
  about 
  half 
  of 
  the 
  alkaline 
  impuri- 
  

   ties 
  were 
  removed, 
  but 
  that 
  the 
  silica 
  and 
  alumina 
  had 
  approx- 
  

   imately 
  the 
  same 
  relative 
  values 
  as 
  before. 
  If 
  the 
  theoretical 
  

   amount 
  of 
  water 
  be 
  introduced 
  into 
  this 
  analysis, 
  and 
  the 
  

   silica 
  be 
  diminished 
  as 
  in 
  the 
  preceding 
  case, 
  the 
  analysis 
  then 
  

   expresses 
  very 
  nearly 
  the 
  theoretical 
  composition. 
  

  

  