﻿426 
  W. 
  F. 
  FISKE. 
  

  

  reproduction, 
  leads 
  to 
  increased 
  density 
  of 
  infestation 
  at 
  and 
  near 
  the 
  source 
  of 
  it 
  

   and 
  to 
  wider 
  dispersion 
  into 
  the 
  unprotected 
  territory 
  surrounding, 
  until 
  increasing 
  

   losses 
  through 
  dispersion 
  compensate 
  for 
  increased 
  rate 
  of 
  reproduction. 
  

  

  But 
  if 
  the 
  dispersion 
  of 
  water 
  at 
  the 
  desert 
  spring 
  were 
  to 
  be 
  completely 
  restricted 
  

   to 
  a 
  very 
  much 
  smaller 
  area 
  than 
  it 
  naturally 
  irrigates, 
  it 
  would 
  tend 
  to 
  accumulate, 
  

   and 
  the 
  " 
  balance 
  " 
  would 
  be 
  destroyed. 
  And 
  so 
  with 
  fly 
  ; 
  if 
  dispersion 
  of 
  it 
  from 
  

   protected 
  precincts 
  is 
  arbitrarily 
  restricted 
  — 
  as 
  actually 
  occurs 
  on 
  small 
  islets 
  — 
  the 
  

   " 
  balance 
  " 
  is 
  destroyed, 
  and 
  almost 
  anything 
  may 
  happen. 
  And, 
  indeed, 
  almost 
  

   everything 
  likely 
  to 
  happen 
  in 
  such 
  circumstances 
  actually 
  occurs 
  on 
  the 
  small 
  

   islets 
  in 
  Victoria 
  Nyanza. 
  The 
  balance 
  in 
  the 
  case 
  of 
  this 
  species 
  of 
  insect 
  is 
  

   indubitably 
  stabilised 
  by 
  dispersion 
  of 
  it 
  from 
  protected 
  precincts, 
  where 
  conditions 
  

   of 
  life 
  are 
  favourable 
  to 
  numerical 
  increase 
  from 
  generation 
  to 
  generation, 
  into 
  a 
  

   surrounding 
  unprotected 
  ' 
  zone 
  of 
  dispersion," 
  wherein 
  conditions 
  of 
  life 
  are 
  

   unfavourable 
  to 
  numerical 
  increase. 
  

  

  The 
  mechanism 
  of 
  the 
  balance 
  rarely 
  fails 
  to 
  function 
  smoothly, 
  except 
  when 
  

   dispersion 
  is 
  in 
  some 
  manner 
  arbitrarily 
  restricted. 
  The 
  islands 
  afford 
  the 
  

   commonest 
  exceptions. 
  

  

  The 
  islets 
  devoid 
  of 
  sand 
  or 
  gravel 
  breeding 
  places, 
  but 
  yet 
  infested 
  by 
  Glossina, 
  

   afford 
  excellent 
  proof 
  of 
  this 
  conception 
  of 
  the 
  " 
  balance 
  M 
  with 
  Glossina 
  palpalis, 
  

   and 
  this 
  conception 
  provides 
  the 
  explanation 
  for 
  the 
  occurrence 
  of 
  fly 
  under 
  these 
  

   conditions 
  of 
  extreme 
  insularity. 
  

  

  The 
  very 
  best 
  example 
  is 
  Lula 
  islet, 
  which 
  is 
  almost, 
  or 
  quite, 
  the 
  smallest 
  in 
  the 
  

   lake 
  that 
  is 
  infested. 
  It 
  contains 
  very 
  good, 
  but 
  very 
  small, 
  breeding 
  grounds 
  in 
  

   vegetable 
  debris. 
  They 
  are 
  so 
  small 
  that 
  they 
  would 
  certainly 
  not 
  serve 
  as 
  colony 
  

   centres 
  if 
  located 
  on 
  the 
  shore 
  of 
  a 
  large 
  island, 
  with 
  no 
  restriction 
  upon 
  dispersion 
  

   of 
  fly 
  from 
  them. 
  But 
  on 
  Lula 
  dispersion 
  is 
  arbitrarily 
  restricted 
  to 
  a 
  space 
  of 
  less 
  

   than 
  5 
  acres, 
  and 
  to 
  an 
  extreme 
  distance 
  of, 
  probably, 
  less 
  than 
  200 
  yards. 
  The 
  

   flies 
  cannot 
  get 
  so 
  far 
  from 
  them 
  that 
  they 
  cannot 
  readily 
  return 
  to 
  them, 
  and 
  in 
  

   whichsoever 
  direction 
  the 
  flies 
  seek 
  breeding 
  places 
  for 
  their 
  young 
  they 
  are 
  quickly 
  

   led 
  to 
  this 
  particular 
  spot. 
  

  

  If 
  Lula 
  were 
  a 
  promontory 
  on 
  Kome 
  Island, 
  where 
  conditions 
  along 
  shore 
  are 
  much 
  

   the 
  same, 
  the 
  infestation 
  would 
  be, 
  as 
  on 
  Kome, 
  extremely 
  light 
  or 
  nil 
  — 
  the 
  flies 
  

   would 
  naturally 
  disperse 
  from 
  so 
  restricted 
  a 
  bit 
  of 
  protected 
  area 
  if 
  they 
  could. 
  

   But 
  they 
  are 
  confined, 
  as 
  in 
  a 
  great 
  breeding-cage, 
  in 
  a 
  corner 
  of 
  which 
  their 
  young, 
  

   if 
  deposited 
  there, 
  will 
  find 
  protection 
  ; 
  and 
  they 
  live 
  as 
  flies 
  in 
  a 
  breeding-cage 
  

   would 
  live, 
  not 
  because 
  the 
  environment 
  is 
  pleasing 
  or 
  attractive, 
  but 
  because 
  it 
  

   happens 
  to 
  be 
  favourable 
  to 
  existence 
  and 
  because 
  they 
  cannot 
  well 
  escape 
  it. 
  

  

  The 
  same 
  conditions 
  prevail 
  on 
  Karambidi, 
  Mugogoya, 
  Lugazi, 
  Sari 
  and 
  Limnaiba 
  

   Islands. 
  All 
  these, 
  except 
  Limnaiba, 
  were 
  infested 
  to 
  a 
  higher 
  degree 
  than 
  the 
  

   extent 
  and 
  character 
  of 
  the 
  breeding 
  grounds 
  seemed 
  to 
  warrant, 
  until 
  this 
  question 
  

   of 
  insularity 
  and 
  restriction 
  of 
  dispersion 
  from 
  breeding 
  grounds 
  was 
  considered. 
  

  

  On 
  Limnaiba 
  Island 
  there 
  was 
  a 
  very 
  good, 
  but 
  very 
  small, 
  bit 
  of 
  breeding 
  ground 
  

   found, 
  and 
  a 
  very 
  much 
  less 
  dense 
  infestation 
  than 
  experience 
  on 
  Lula, 
  Mugogoya, 
  etc., 
  

   would 
  lead 
  one 
  to 
  expect. 
  Search 
  in 
  this 
  breeding 
  ground 
  disclosed 
  such 
  numbers 
  

  

  