﻿Pratt 
  — 
  Sejparation 
  of 
  Alumina 
  from 
  Molten 
  Magmas, 
  227 
  

  

  Akt. 
  XX 
  Y. 
  — 
  On 
  the 
  Separation 
  of 
  Alumina 
  from 
  Molten 
  

   Magmas^ 
  and 
  the 
  formation 
  of 
  Corundum 
  ; 
  by 
  J. 
  H. 
  

   Pkatt. 
  

  

  Some 
  interesting 
  problems 
  have 
  arisen, 
  in 
  recent 
  years, 
  re- 
  

   garding 
  tlie 
  differentiation 
  of 
  igneous 
  magmas 
  upon 
  cooling. 
  

   It 
  has 
  been 
  the 
  author's 
  privilege 
  during 
  the 
  last 
  three 
  years 
  

   to 
  study 
  in 
  the 
  field 
  various 
  occurrences 
  of 
  corundum, 
  spinel 
  

   and 
  chromite 
  in 
  peridotite 
  rocks, 
  and 
  in 
  some 
  recent 
  papers* 
  

   the 
  igneous 
  origin 
  and 
  the 
  chemical 
  composition 
  of 
  these 
  

   minerals 
  have 
  been 
  discussed 
  by 
  him. 
  It 
  is 
  the 
  intention 
  in 
  

   this 
  paper 
  to 
  give 
  those 
  field 
  observations 
  that 
  have 
  a 
  theoreti- 
  

   cal 
  bearing 
  upon 
  the 
  separation 
  of 
  these 
  minerals 
  from 
  a 
  molten 
  

   magma 
  and 
  more 
  especially 
  of 
  the 
  alumina. 
  

  

  The 
  results 
  of 
  the 
  elaborate 
  studies 
  that 
  have 
  been 
  made 
  by 
  

   Yogt,f 
  Morozewicz,:}: 
  Lagorio§ 
  and 
  others 
  upon 
  the 
  artificial 
  

   formation 
  of 
  minerals 
  in 
  molten 
  magmas, 
  have 
  been 
  verified 
  

   by 
  observations 
  upon 
  the 
  occurrence 
  of 
  the 
  same 
  minerals 
  in 
  

   nature 
  that 
  have 
  also 
  separated 
  out 
  from 
  molten 
  magmas. 
  

  

  The 
  separation 
  of 
  alumina 
  is 
  well 
  illustrated 
  in 
  nature 
  in 
  the 
  

   occurrence 
  of 
  corundum, 
  spinel 
  and 
  chromite 
  in 
  the 
  rocks 
  of 
  

   the 
  peridotite 
  group. 
  Of 
  these 
  rocks, 
  the 
  most 
  common 
  one 
  

   containing 
  these 
  three 
  minerals 
  is 
  dunite, 
  consisting 
  almost 
  

   entirely 
  of 
  chrysolite 
  or 
  its 
  alteration 
  product, 
  serpentine. 
  In 
  

   the 
  analyses 
  of 
  the 
  dunite 
  rocks, 
  they 
  show 
  either 
  a 
  trace 
  or 
  

   very 
  small 
  percentage 
  of 
  alumina 
  or 
  chromic 
  oxide 
  or 
  even 
  

   none 
  ; 
  and 
  in 
  the 
  analyses 
  of 
  other 
  peridotites 
  than 
  dunite, 
  the 
  

   same 
  usually 
  holds 
  true 
  for 
  amounts 
  beyond 
  those 
  required 
  in 
  

   the 
  composition 
  of 
  the 
  normal 
  minerals 
  that 
  compose 
  these 
  

   rocks. 
  

  

  The 
  peridotite 
  localities 
  that 
  have 
  been 
  examined 
  by 
  the 
  

   author 
  are 
  all 
  located 
  in 
  the 
  western 
  part 
  of 
  JS'orth 
  Carolina, 
  

   and 
  at 
  these 
  various 
  localities 
  the 
  following 
  phenomena 
  have 
  

   been 
  observed 
  : 
  

  

  1. 
  Peridotite 
  rocks 
  containing 
  small 
  particles 
  and 
  grains 
  of 
  

   chromite 
  but 
  no 
  corundum 
  or 
  spinel. 
  

  

  2. 
  Peridotite 
  rocks 
  containing 
  deposits 
  of 
  chromite 
  (which 
  

   upon 
  analysis 
  showed 
  a 
  variable 
  per 
  cent 
  of 
  alumina 
  and 
  mag- 
  

   nesia), 
  but 
  no 
  corundum 
  or 
  spinel. 
  

  

  3. 
  Peridotite 
  rocks 
  containing 
  chromite, 
  corundum 
  and 
  

   spinel. 
  

  

  * 
  This 
  Journal, 
  vol. 
  vi, 
  July, 
  1898, 
  p. 
  49; 
  ibid, 
  vol. 
  vii, 
  April, 
  1899, 
  p. 
  281, 
  and 
  

   Trans. 
  Am. 
  Inst. 
  Mining 
  Bug., 
  vol. 
  xxix, 
  1899. 
  

  

  f 
  Zeitschr. 
  fur 
  Prakt. 
  Geol 
  , 
  Nos. 
  1, 
  4 
  and 
  7, 
  1893. 
  

  

  X 
  Zeitschr. 
  fiir 
  Kryst., 
  vol. 
  xxiv, 
  p, 
  281, 
  1895, 
  and 
  Tschermak's 
  Min. 
  u. 
  Petrog. 
  

   Mittheilungen, 
  Bd. 
  xviii. 
  H. 
  1-2-3, 
  pp. 
  1-90 
  and 
  105-240, 
  1898. 
  

  

  § 
  Zeitschr. 
  fur 
  Kryst., 
  vol. 
  xxiv, 
  p. 
  285, 
  1895. 
  

  

  