﻿INVESTIGATIONS 
  INTO 
  THE 
  BIONOMICS 
  OF 
  GLOSSINA 
  PALPALIS. 
  453 
  

  

  Bwendi 
  district 
  (fig. 
  6, 
  p. 
  411) 
  or 
  at 
  Mujuzi 
  Creek, 
  and 
  there 
  would 
  be 
  dispersion 
  from 
  

   them 
  as 
  from 
  the 
  points 
  in 
  the 
  Bwendi 
  district 
  or 
  from 
  the 
  point 
  at 
  the 
  mouth 
  of 
  

   Mujuzi 
  Creek 
  where 
  crocodiles 
  harbour 
  (p. 
  388), 
  or 
  in 
  Bugaba 
  Island 
  where 
  breeding 
  

   ground 
  marks 
  the 
  colony 
  centres 
  (p. 
  423). 
  There 
  can 
  be 
  no 
  super-density 
  of 
  fly 
  

   except 
  under 
  exactly 
  such 
  conditions 
  as 
  prevailed 
  at 
  Kimmi 
  when 
  there 
  is 
  food 
  

   generally 
  distributed 
  throughout 
  the 
  island, 
  shelter 
  and 
  breeding 
  ground 
  within 
  

   easy 
  reach 
  of 
  food, 
  and 
  no 
  opportunity 
  for 
  dispersion. 
  There 
  may 
  be 
  starvation 
  

   and 
  famine, 
  and 
  a 
  greatly 
  reduced 
  reproductivity, 
  but 
  it 
  cannot 
  be 
  induced 
  by 
  the 
  

   fly's 
  own 
  activities, 
  unless 
  (1) 
  conditions 
  of 
  life 
  are 
  locally 
  favourable 
  to 
  increase, 
  

   and 
  (2) 
  dispersion 
  from 
  the 
  locality 
  is 
  arbitrarily 
  restricted. 
  

  

  X. 
  General 
  Conclusions. 
  

  

  There 
  are 
  three 
  specific 
  requisites 
  to 
  life 
  of 
  Glossina 
  palpalis, 
  one 
  sustentative 
  and 
  

   two 
  protective 
  : 
  — 
  good 
  shelter 
  and 
  breeding 
  ground. 
  One 
  kind 
  of 
  food, 
  the 
  blood 
  

   of 
  vertebrates, 
  suffices 
  for 
  the 
  insect 
  in 
  all 
  its 
  stages. 
  The 
  two 
  kinds 
  of 
  protection 
  

   serve 
  adults 
  and 
  pupae 
  respectively. 
  

  

  All 
  these 
  appear 
  equally 
  requisite. 
  There 
  is 
  some 
  slight 
  evidence 
  that 
  other 
  food 
  

   than 
  blood 
  may 
  be 
  taken, 
  but 
  none 
  concerning 
  its 
  nature, 
  and 
  no 
  evidence 
  that 
  any 
  

   other 
  is 
  required. 
  There 
  is 
  much 
  evidence 
  that 
  blood 
  is 
  absolutely 
  required 
  — 
  as 
  

   supplied 
  by 
  the 
  experiment 
  cited 
  in 
  Sect. 
  II 
  (6), 
  and 
  the 
  effect 
  upon 
  behaviour 
  of 
  flies 
  

   of 
  depriving 
  them 
  of 
  their 
  normal 
  food 
  supply 
  ; 
  by 
  the 
  apparent 
  reduction 
  in 
  rate 
  

   of 
  reproduction 
  and 
  in 
  density 
  of 
  fly 
  on 
  Kimmi 
  Island 
  associated 
  with 
  obvious 
  

   shortage 
  of 
  food 
  (Sect. 
  IX 
  (h)) 
  ; 
  by 
  the 
  long-shore 
  distribution 
  of 
  fly 
  in 
  the 
  Bwendi 
  

   district 
  as 
  described 
  in 
  Sect. 
  VI 
  (g) 
  ; 
  by 
  the 
  distribution 
  of 
  fly 
  in 
  the 
  Mujuzi 
  Creek 
  

   district 
  (note 
  to 
  Sect. 
  Ill 
  (e)), 
  and 
  in 
  various 
  other 
  cases. 
  No 
  one 
  of 
  these 
  cases, 
  

   of 
  itself, 
  would 
  establish 
  the 
  fact, 
  but 
  the 
  cumulative 
  evidence 
  is 
  all 
  to 
  one 
  effect, 
  

   and 
  the 
  mass 
  of 
  it 
  convincing. 
  

  

  The 
  absolute 
  necessity 
  for 
  the 
  kind 
  of 
  protection 
  that 
  is 
  provided 
  by 
  breeding 
  

   grounds 
  is 
  particularly 
  well 
  shown 
  by 
  the 
  fly 
  survey 
  of 
  Bugaba 
  Island 
  (Sect. 
  VI 
  (g) 
  and 
  

   fig. 
  7). 
  This 
  was 
  confirmed 
  by 
  similar 
  surveys 
  of 
  Bubembe 
  and 
  Bunyama 
  Islands, 
  

   which 
  have 
  not 
  been 
  included 
  herein 
  for 
  the 
  sake 
  of 
  brevity, 
  and 
  by 
  observations 
  

   many 
  times 
  repeated. 
  It 
  is 
  not 
  requisite 
  that 
  the 
  breeding 
  grounds 
  should 
  be 
  in 
  sand 
  

   or 
  gravel 
  — 
  fine, 
  dry 
  vegetable 
  debris 
  will 
  serve 
  as 
  well 
  ; 
  but 
  it 
  happens 
  that 
  the 
  sand 
  

   and 
  gravel 
  breeding 
  places 
  are 
  most 
  extensive 
  in 
  this 
  particular 
  region, 
  and 
  more 
  attrac- 
  

   tive 
  than 
  any 
  other, 
  so 
  that 
  very 
  much 
  the 
  larger 
  number 
  of 
  flies 
  breed 
  in 
  them. 
  

  

  A 
  surprise 
  was 
  the 
  discovery 
  that 
  " 
  massive 
  shelter 
  " 
  is 
  requisite 
  to 
  the 
  life 
  of 
  this 
  

   fly. 
  This 
  discovery 
  was 
  entirely 
  unexpected, 
  and 
  for 
  a 
  long 
  time 
  it 
  seemed 
  incredible 
  

   that 
  absence 
  of 
  massive 
  shelter 
  should 
  be 
  the 
  explanation 
  for 
  absence 
  or 
  relative 
  

   scarcity 
  of 
  fly 
  on 
  the 
  islands 
  cited 
  in 
  Table 
  XXXIII 
  (p. 
  428). 
  Cumulative 
  evidence 
  

   finally 
  dispelled 
  all 
  doubts 
  — 
  notably 
  that 
  supplied 
  by 
  the 
  study 
  of 
  the 
  Lutoboka 
  

   Bay 
  colony 
  (fig. 
  9, 
  p. 
  431) 
  ; 
  by 
  conditions 
  along 
  the 
  southern 
  shore 
  of 
  Buganga 
  

   Peninsula 
  (Table 
  XXXIV, 
  p. 
  432) 
  ; 
  and 
  various 
  other 
  less 
  striking 
  cases 
  not 
  cited 
  

   in 
  these 
  pages. 
  There 
  could 
  be 
  no 
  doubt, 
  in 
  the 
  end, 
  that 
  the 
  light 
  shelter 
  that 
  is 
  

   most 
  attractive 
  to 
  the 
  active 
  flies, 
  and 
  which 
  also 
  serves 
  in 
  the 
  majority 
  of 
  cases 
  

   to 
  shade 
  breeding 
  grounds, 
  is 
  insufficient 
  to 
  protect 
  the 
  inactive 
  flies 
  from 
  destructors, 
  

   animate 
  or 
  inanimate, 
  as 
  the 
  case 
  may 
  be. 
  

  

  