﻿380 
  ANNUAL 
  EEPOKT 
  SMITHSONIAN" 
  INSTITUTION, 
  1927 
  

  

  In 
  our 
  fresh-water 
  fishes 
  each 
  species 
  on 
  an 
  average 
  has 
  been 
  

   described 
  as 
  new 
  some 
  three 
  or 
  four 
  times, 
  on 
  account 
  of 
  minor 
  varia- 
  

   tions, 
  real 
  or 
  supposed. 
  In 
  Europe, 
  where 
  the 
  fishes 
  have 
  been 
  

   studied 
  longer 
  and 
  by 
  more 
  different 
  men, 
  upwards 
  of 
  six 
  or 
  eight 
  

   nominal 
  species 
  have 
  been 
  described 
  for 
  each 
  one 
  that 
  is 
  now 
  con- 
  

   sidered 
  tenable. 
  

  

  It 
  is 
  evident 
  that 
  the 
  idea 
  of 
  a 
  separate 
  creation 
  for 
  each 
  species 
  

   of 
  fishes 
  in 
  each 
  river 
  basin, 
  as 
  entertained 
  by 
  Agassiz, 
  is 
  wholly 
  in- 
  

   compatible 
  with 
  our 
  present 
  knowledge 
  of 
  the 
  specific 
  distinctions 
  or 
  

   of 
  the 
  geographical 
  distribution 
  of 
  fishes. 
  There 
  is 
  an 
  unbroken 
  gra- 
  

   dation 
  in 
  the 
  variations 
  from 
  the 
  least 
  to 
  the 
  greatest 
  — 
  from 
  the 
  pecu- 
  

   liarities 
  of 
  the 
  individual, 
  through 
  local 
  varieties, 
  geographical 
  sub- 
  

   species, 
  species, 
  subgenera, 
  genera, 
  families, 
  superfamilies, 
  and 
  so 
  on, 
  

   until 
  all 
  fishlike 
  vertebrates 
  are 
  included 
  in 
  a 
  single 
  bond 
  of 
  union. 
  

  

  It 
  is, 
  however, 
  evident 
  that 
  not 
  all 
  American 
  types 
  of 
  fishes 
  had 
  

   their 
  origin 
  in 
  America, 
  or 
  even 
  first 
  assumed 
  in 
  America 
  their 
  

   present 
  forms. 
  Some 
  of 
  these 
  are 
  perhaps 
  immigrants 
  from 
  north- 
  

   ern 
  Asia, 
  where 
  they 
  still 
  have 
  their 
  nearest 
  relatives. 
  Still 
  others 
  

   are 
  evidently 
  modified 
  importations 
  from 
  the 
  sea; 
  and 
  of 
  these 
  some 
  

   are 
  very 
  recent 
  immigrants, 
  landlocked 
  species 
  which 
  have 
  changed 
  

   very 
  little 
  from 
  the 
  parent 
  stock. 
  

  

  The 
  character 
  and 
  possible 
  origin 
  of 
  each 
  of 
  the 
  34 
  families 
  of 
  

   North 
  American 
  fresh 
  -water 
  fishes 
  may 
  be 
  summarized 
  briefly 
  as 
  

   follows 
  : 
  

  

  The 
  lampreys 
  are 
  evidently 
  of 
  marine 
  origin, 
  as 
  the 
  marine 
  species 
  

   are 
  still 
  anadromous. 
  The 
  fresh-water 
  species, 
  compared 
  with 
  the 
  

   marine 
  ones, 
  are 
  smaller 
  in 
  size 
  and 
  weaker 
  in 
  organization, 
  and 
  rep- 
  

   resent 
  larval 
  conditions 
  or 
  often 
  arrests 
  of 
  development 
  of 
  the 
  latter 
  

   form. 
  In 
  most 
  river 
  forms, 
  the 
  loss 
  of 
  the 
  upper 
  part 
  of 
  the 
  intestine 
  

   is 
  a 
  sort 
  of 
  senile 
  degradation. 
  

  

  The 
  paddlefish 
  is 
  allied 
  to 
  extinct 
  ganoid 
  types. 
  The 
  group 
  is 
  now 
  

   represented 
  by 
  one 
  species 
  in 
  America 
  and 
  another 
  in 
  central 
  Asia. 
  

  

  The 
  sturgeons, 
  like 
  the 
  lampreys, 
  are 
  anadromous. 
  But 
  three 
  of 
  

   the 
  American 
  species 
  are 
  now 
  confined 
  to 
  the 
  fresh 
  waters, 
  and 
  two 
  

   of 
  these 
  belong 
  to 
  peculiar 
  genera 
  {Sca'phirhyncJi'iis 
  and 
  ParascapM- 
  

   rhynchus), 
  which 
  (like 
  Polyodon) 
  have 
  representatives 
  also 
  in 
  central 
  

   Asia. 
  As 
  to 
  whether 
  the 
  parent 
  stock 
  in 
  either 
  case 
  is 
  American 
  or 
  

   Asiatic, 
  I 
  know 
  of 
  no 
  positive 
  evidence. 
  

  

  The 
  gar 
  pikes 
  and 
  the 
  howpns 
  are 
  strictly 
  American 
  types 
  allied 
  

   to 
  extinct 
  ganoid 
  forms, 
  and 
  doubtless 
  developed 
  from 
  such 
  in 
  the 
  

   waters 
  they 
  now 
  inhabit. 
  

  

  The 
  cat-fishes 
  of 
  America 
  are 
  all 
  probably 
  descendants 
  of 
  a 
  com- 
  

   mon 
  stock 
  not 
  closely 
  allied 
  to 
  South 
  American 
  forms, 
  but 
  not 
  find- 
  

   ing 
  any 
  nearer 
  relatives 
  in 
  Asia. 
  The 
  single 
  species 
  of 
  this 
  type 
  

  

  