﻿114 
  NEW 
  YORK 
  STATE 
  MUSEUM 
  

  

  what 
  lie 
  has 
  said 
  in 
  his 
  report 
  to 
  the 
  United 
  States 
  Board 
  of 
  

   Engineers 
  on 
  Deep 
  Waterways, 
  viz 
  : 
  

  

  The 
  data 
  for 
  estimating 
  the 
  water 
  supply 
  of 
  a 
  large 
  canal, 
  

   specially 
  when 
  on 
  a 
  large 
  scale, 
  should 
  be 
  based, 
  when 
  such 
  data 
  

   are 
  available, 
  upon 
  actual 
  gagings 
  of 
  streams, 
  rather 
  than 
  on 
  

   general 
  considerations 
  derived 
  from 
  study 
  of 
  the 
  rainfall 
  alone. 
  

   An 
  examination 
  of 
  a 
  large 
  number 
  of 
  estimates 
  of 
  canal 
  water 
  

   supplies, 
  based 
  on 
  the 
  usual 
  method, 
  shows 
  that 
  rainfall 
  data 
  

   alone 
  are 
  in 
  close 
  cases 
  inadequate 
  for 
  solving 
  a 
  water-supply 
  

   problem 
  of 
  the 
  magnitude 
  of 
  the 
  one 
  now 
  under 
  consideration. 
  

   When, 
  however, 
  actual 
  gagings 
  of 
  the 
  streams, 
  extending 
  over 
  a 
  

   sufficient 
  number 
  of 
  years, 
  are 
  available, 
  there 
  is 
  no 
  reason 
  why 
  a 
  

   water-supply 
  problem 
  on 
  a 
  large 
  scale 
  may 
  not 
  be 
  worked 
  out 
  with 
  

   the 
  precision 
  of 
  a 
  proposition 
  in 
  mathematics. 
  

  

  What 
  is 
  here 
  said 
  in 
  regard 
  to 
  water 
  supplies 
  for 
  canals 
  is 
  

   equally 
  true 
  as 
  regards 
  all 
  other 
  water 
  supplies, 
  either 
  municipal 
  

   or 
  for 
  water 
  power, 
  etc. 
  Farther 
  on 
  in 
  the 
  same 
  chapter 
  it 
  is 
  

   stated 
  : 
  

  

  It 
  is 
  not 
  intended 
  to 
  say, 
  however, 
  that 
  rainfall 
  data 
  are 
  not 
  of 
  

   use 
  in 
  a 
  hydro 
  logic 
  discussion. 
  When, 
  as 
  in 
  the 
  present 
  case, 
  in 
  

   addition 
  to 
  stream 
  gagings 
  an 
  extended 
  series 
  of 
  such 
  data 
  are 
  

   available, 
  the 
  argument 
  is 
  made 
  doubly 
  good 
  and 
  the 
  demonstra- 
  

   tion 
  strengthened. 
  

  

  When 
  records 
  of 
  gagings 
  are 
  available 
  the 
  computation 
  becomes 
  

   very 
  simple. 
  It 
  is 
  merely 
  a 
  matter 
  of 
  simple 
  addition 
  and 
  sub- 
  

   traction. 
  

  

  The 
  complete 
  data 
  required 
  in 
  order 
  to 
  compute 
  the 
  safe 
  pos- 
  

   sible 
  yield 
  of 
  a 
  stream 
  are 
  as 
  follows 
  : 
  

  

  1) 
  The 
  catchment 
  area. 
  

  

  2) 
  The 
  rainfall 
  of 
  the 
  minimum 
  year, 
  as 
  well 
  as 
  for 
  a 
  series 
  of 
  

   years. 
  

  

  3) 
  A 
  ground- 
  water 
  diagram 
  of 
  the 
  stream 
  or, 
  lacking 
  such, 
  a 
  

   diagram 
  for 
  a 
  neighboring 
  stream 
  lying 
  in 
  the 
  same 
  or 
  a 
  similar 
  

   geologic 
  formation, 
  and, 
  so 
  far 
  as 
  possible, 
  with 
  similar 
  condi- 
  

   tions 
  of 
  forestation. 
  

  

  4) 
  The 
  available 
  storage 
  capacity 
  of 
  the 
  stream. 
  

  

  5) 
  The 
  loss 
  by 
  water 
  surface 
  evaporation 
  from 
  the 
  reservoirs, 
  

   together 
  with 
  an 
  estimate 
  of 
  the 
  loss 
  by 
  percolation. 
  

  

  