273 
hours  later.  Two  of  these  flies  contained  intestinal  parasites.  In 
one  they  occurred  as  small  clumps  with  the  flagellar  ends  directed 
towards  the  periphery,  and  in  the  other  as  long  freely  motile  forms. 
Parasites  were  also  present  in  the  proboscis  of  this  latter  fly. 
IX.  ORIGIN  OF  PARASITES 
Some  conflict  of  opinion  exists  as  to  the  origin  of  these  ‘  wild  ’ 
parasites.  It  would  seem  that,  to  a  certain  degree,  this  may  be 
explained  by  the  views  held  by  various  writers  on  the  question 
as  to  the  food  of  tsetse  flies.  Novy,**  who  thinks  that  they  may 
possibly  feed  on  other  substances  than  blood,  is  of  the  opinion  that 
the  pai-asites  may  possibly  be  derived  from  stagnant  water,  or  some 
similar  fluid.  Reference  has  already  been  made  to  an  instance  in 
which  the  flies  may  have  been  imbibing  such  a  fluid.  Against  this, 
however,  is  the  experience  in  raising  and  maintaining  tsetse  flies  in 
captivity,  which,  so  far,  has  shown  that  blood  is  the  only  food  which 
they  will  ingest. 
On  the  other  hand,  Minchin’’  and  others  consider  that  blood  alone 
is  the  natural  food  of  the  tsetse  flies,  and  that  in  all  probability  the 
parasites  are  derived  from  a  vertebrate  host,  possibly  a  bird.  To  a 
certain  extent,  the  infection  may  be  due  to  the  ‘  contaminative  ’ 
transmission,  that  is  the  ingestion  of  the  cystic  forms  of  T.  grayi, 
though  the  probability  of  the  continuation  of  the  fly-infection  in  this 
way  is  rather  slight. 
So  far,  no  vertebrate  host  of  these  parasites  has  been  discovered. 
On  the  island  we  examined  the  blood  of  darters,  pigeons,  hornbills, 
several  species  of  small  birds,  snakes,  mice  and  crocodiles  (4),  but  in 
none  of  them  were  parasites  of  any  description  seen.  T.  sp.  (?) 
were  seen  in  frogs  caught  on  the  mainland. 
The  difference  of  opinion  on  the  question  of  hereditary  trans¬ 
mission  also  tends  to  complicate  the  solution.  Patton  and  Strickland,® 
who  incline  to  this  hypothesis,  quote  in  support  of  their  view  the 
instance  in  which  Minchin,  Gray  and  Tulloch'"^  observed  the  infection 
in  a  laboratory  raised  fly  which  had  been  fed  on  a  hen.  If  hereditary 
transmission  of  the  flagellates,  and  food  other  than  blood 
are  eliminated,  it  is  rather  difficult  to  account  for  the  infection  of  so 
large  a  percentage  of  Gl.  palpalis  on  the  island.  It  may  be 
s 
