﻿Western 
  Pennsylvania. 
  209 
  

  

  through 
  which 
  it 
  flowed 
  is 
  an 
  open 
  question, 
  but 
  that 
  sub- 
  

   glacial 
  waters 
  exerted 
  a 
  peculiar 
  eroding 
  power 
  in 
  certain 
  places 
  

   near 
  the 
  ice 
  margin 
  is 
  conclusively 
  shown 
  in 
  various 
  parts 
  of 
  

   the 
  glaciated 
  district 
  by 
  the 
  presence 
  of 
  large 
  channels 
  made 
  

   by 
  them. 
  Some 
  of 
  these 
  are 
  remote 
  from 
  present 
  streams 
  and 
  

   have 
  been 
  little 
  aifected 
  by 
  post-glacial 
  erosion 
  Their 
  trend 
  

   is 
  in 
  line 
  with 
  the 
  striation 
  and 
  approximately 
  at 
  right 
  angles 
  

   with 
  the 
  moraine. 
  They 
  are 
  often 
  occupied 
  by 
  osars 
  and 
  

   hence 
  are 
  called 
  " 
  osar 
  troughs." 
  These 
  troughs 
  or 
  channels 
  

   sometimes 
  rise 
  toward 
  the 
  moraine 
  at 
  the 
  rate 
  of 
  several 
  feet 
  

   per 
  mile, 
  and 
  yet 
  the 
  material 
  in 
  the 
  osars 
  lying 
  in 
  them 
  shows 
  

   conclusively 
  that 
  the 
  flow 
  was 
  in 
  that 
  direction. 
  The 
  water 
  

   seems 
  to 
  have 
  been 
  forced 
  upward 
  toward 
  the 
  ice 
  margin 
  by 
  

   the 
  weight 
  of 
  the 
  ice 
  sheet 
  and 
  by 
  hydrostatic 
  pressure. 
  These 
  

   osar 
  troughs 
  were 
  formed 
  just 
  before 
  the 
  ice 
  made 
  its 
  final 
  

   retreat, 
  but 
  the 
  eroding 
  power, 
  of 
  which 
  they 
  are 
  the 
  product, 
  

   was 
  probably 
  in 
  operation 
  in 
  earlier 
  stages 
  of 
  the 
  ice 
  invasion. 
  

   The 
  outer 
  moraine 
  in 
  the 
  district 
  under 
  discussion 
  is 
  a 
  complex 
  

   one, 
  the 
  equivalent 
  of 
  several 
  moraines 
  farther 
  west 
  that 
  indi- 
  

   cate 
  a 
  succession 
  of 
  advances 
  and 
  retreats 
  of 
  the 
  ice 
  front. 
  In 
  

   the 
  early 
  stages 
  the 
  rock 
  floors 
  of 
  these 
  valleys 
  may 
  have 
  been 
  

   deepened 
  in 
  places 
  by 
  the 
  subglacial 
  streams 
  in 
  the 
  same 
  man- 
  

   ner 
  as 
  the 
  osar 
  troughs 
  were 
  produced. 
  By 
  reference 
  to 
  the 
  

   accompanying 
  map 
  it 
  will 
  be 
  seen 
  that 
  every 
  valley 
  in 
  which 
  

   a 
  very 
  low 
  rock 
  floor 
  has 
  been 
  reported 
  has 
  a 
  trend 
  approxi- 
  

   mately 
  at 
  right 
  angles 
  with 
  the 
  moraine 
  and 
  in 
  line 
  with 
  the 
  

   ice 
  movement; 
  that 
  is, 
  such 
  a 
  trend 
  as 
  to 
  invite 
  the 
  flow 
  of 
  

   ice 
  and 
  of 
  subglacial 
  waters. 
  Furthermore 
  it 
  appears 
  that 
  in 
  

   every 
  case 
  the 
  lowest 
  known 
  point 
  of 
  the 
  rock 
  channel 
  in 
  these 
  

   several 
  valleys 
  is 
  near 
  the 
  inner 
  border 
  of 
  the 
  moraine. 
  In 
  

   case 
  it 
  is 
  found 
  that 
  no 
  northward 
  outlets 
  exist 
  the 
  most 
  plausi- 
  

   ble 
  explanation 
  for 
  the 
  low 
  altitude 
  at 
  these 
  points 
  would 
  seem 
  

   to 
  be 
  a 
  deepening 
  of 
  the 
  channels 
  here 
  below 
  their 
  main 
  out- 
  

   lets 
  by 
  subglacial 
  waters 
  assisted, 
  perhaps, 
  by 
  the 
  ice 
  itself. 
  

  

  Summing 
  up 
  all 
  the 
  available 
  evidence, 
  it 
  appears 
  that 
  no 
  

   northward 
  outlets 
  have 
  been 
  found 
  for 
  the 
  low 
  channels 
  just 
  

   within 
  the 
  moraine 
  on 
  these 
  several 
  streams 
  which 
  are 
  not 
  

   embarrassed, 
  either 
  by 
  a 
  rise 
  in 
  the 
  rock 
  floor 
  or 
  an 
  extraordi- 
  

   nary 
  amount 
  of 
  drift. 
  In 
  the 
  streams 
  under 
  special 
  considera- 
  

   tion, 
  the 
  Shenango, 
  Mahoning 
  and 
  Beaver, 
  it 
  appears 
  that 
  the 
  

   rock 
  floor 
  rises 
  in 
  all 
  directions 
  from 
  Edenburg, 
  unless 
  there 
  

   be 
  a 
  descent 
  down 
  the 
  Beaver. 
  The 
  obstacles 
  to 
  a 
  northward 
  

   discharge 
  of 
  these 
  streams 
  seem, 
  on 
  the 
  whole, 
  greater 
  than 
  

   those 
  in 
  the 
  way 
  of 
  a 
  southward 
  discharge. 
  In 
  the 
  Mononga- 
  

   hela, 
  lower 
  Allegheny 
  and 
  the 
  Ohio 
  valleys, 
  the 
  available 
  

   evidence 
  all 
  indicates 
  southward 
  discharge 
  along 
  the 
  present 
  

   course 
  of 
  the 
  Ohio 
  from 
  the 
  interglacial 
  period 
  to 
  the 
  present 
  

   time. 
  

  

  