287 
It  is  hardly  necessary  to  consider  at  any  length  the  question  as  to 
whether  crocodiles  form  the  staple  article  of  diet  of  Glossina  falfalis, 
as  has  been  suggested  by  Koch  *  In  captivity  the  fly  will  feed  on 
various  animals — we  have  used  dogs,  monkeys,  sheep,  goats,  hens, 
a  hornbill,  and  some  small  birds — and  in  its  wild  state  the  source  of 
food  will  be  determined  altogether  by  the  local  fauna.  In  some 
places  where  Gl.  pal  falls  exists,  there  are  no  crocodiles ;  so  that  the 
association  of  the  two  is  more  or  less  accidental. 
For  instance,  at  Madona  we  were  informed  by  residents  of  many  years’ 
standing  that  it  was  an  extremely  rare  event  to  see  a  crocodile,  yet  the  fly  was 
plentiful.  Again,  some  of  the  streams  flowing  into  the  Lovu  river  were  too  small 
to  harbour  crocodiles  and  yet  the  flies  existed  on  the  streams. 
One  very  common  source  of  food  is  afforded  by  several  species  of 
water  birds,  especially  darters,  which  are  accustomed  to  sit  for  hours 
on  dead  trees  at  the  water’s  edge,  with  their  wings  widely  outspread. 
On  the  island  in  Lake  Tanganyika,  of  which  we  have  already  spoken,  the  only 
living  things  were  several  species  of  birds,  chiefly  water-fowl,  some  snakes,  mice, 
and  a  few  crocodiles.  This  island  lies  at  least  two  miles  from  the  mainland,  and 
has  been  uninhabited  for,  certainly,  the  pa.st  twenty  y^ears.  Although  Gl.  falfalis 
was  extremely  abundant,  and  was  found  around  the  whole  of  the  shore,  the 
crocodiles  w'ere  practically  confined  to  one  particular  bay%  so  that  all  the  fly  did 
not  have  the  opportunity  of  feeding  on  them.  We  attempted  to  feed  captive  flies 
on  freshly-shot  crocodiles,  and  while  they  made  vigorous  efforts  to  do  so,  they  were 
unable  to  pierce  even  the  thinnest  portions  of  the  skin  of  the  particular  crocodiles 
we  were  using. 
About  33  per  cent,  of  the  flies  we  examined  {61  out  of  185),  on  this  island, 
showed  evidence  of  having  fed  on  blood.  In  every  case,  with  the  exception  of  one. 
the  blood  cells,  when  recognisable,  were  oval  and  nucleated,  but  we  were  not  able 
to  determine,  with  any  degree  of  certainty,  whether  they  were  from  birds  or 
reptiles.  None  of  the  birds  or  crocodiles  we  shot  had  blood  parasites,  e.g.,  haemogre- 
garines,  and  the  changes  produced  in  the  contours  of  the  cells  and  nuclei  by  the 
process  of  digestion  rendered  measurements  useless.  In  the  odd  fly,  the  blood  was 
human,  and  quite  fresh,  and  had  evidently  come  from  one  of  the  boys  engaged  in 
catching  specimens. 
417  per  cent.  01  the  Gl.  falpalis  were  infected  with  intestinal  parasites 
of  the  herpetamonad  type.  The  discussion  of  these  forms  is  reserved  for  a  future 
paper. 
V.  HABITS  OF  GLOSSJNA  MORSITANS 
So  far  as  the  time  of  day  at  which  they  are  most  active,  their 
catholicity  of  taste  in  the  matter  of  food,  and  their  preference  for 
dark  colours  are  concerned,  the  habits  of  this  species  are  much  similar 
to  those  of  Gl.  palpalis. 
*  Koch,  1907.  Deutsche  med.  Wochenschrift,  Jahrgang  33,  No.  46. 
