1 
272 
intestinal  contents  of  the  flies  were  examined  at  various  periods  after¬ 
wards,  from  a  few  hours  to  several  days.  The  results  of  this  examina¬ 
tion  were  rather  surprising,  for  whereas  42  per  cent,  of  unfed 
(artificially)  flies  contained  flagellates,  only  a  very  small  proportion  of 
the  fed  ones  were  found  to  harbour  parasites. 
The  discrepancies  shown  by  these  figures  are  rather  interesting, 
and  rather  difficult  to  explain,  unless  it  be  due  to  the  fact  that 
mammalian  blood  exerts  some  unfavourable  action  on  the  parasites  in 
these  flies.  As  we  have  already  said,  mammalian  blood  does  not 
form  any  part  of  the  ordinary  food-supply  of  the  flies  on  the  island, 
The  difference  in  the  results  between  the  flies  fed  on  the  hen  and 
those  fed  on  birds  inhabiting  the  island  is  also  suggestive.  However, 
the  results  are  so  scanty  that  no  great  weight  can  be  laid  on  the 
suggestion  we  have  offered. 
In  Uganda,  Minchin,  Gray  and  TulloclH  found  that  goats’  serum 
acted  deleteriously  on  the  ‘  wild  ’  parasites,  and  found  that  by  this 
means  they  could  be  distinguished  from  J.  gambiense.  Other  sorts 
of  serum  had,  however,  no  such  action. 
The  following  experiment  was  made  to  ascertain  whether  crocodile 
serum  was  more  favourable  to  the  parasites  than  that  of  some  of  the 
local  birds. 
Hanging  drop  preparations  were  made  from  the  intestinal  contents 
of  a  fly  containing  very  many  parasites,  and  were  diluted  in  the 
proportion  of  i  :  i  with  salt-citrate  solution  and  serum  obtained  from 
a  crocodile,  a  hen  and  a  darter.  The  preparations  were  examined  at 
intervals  of  fifteen  minutes  for  a  period  of  four  hours.  The  changes 
which  occurred,  consisted  of  a  clumping  of  the  parasites  with  the 
flagella  towards  the  periphery,  and  a  gradual  cessation  of  the  move¬ 
ment  of  the  clumped  forms.  Simultaneously  they  started  to  break 
down,  until  finally  there  was  left  a  granular  and  refractile  mass.  The 
clumping  occurred  rather  more  slowly  in  the  preparation  containing 
crocodile  serum  than  in  the  other  two,  but  otherwise  precisely  similar 
changes  occurred,  and  at  the  end  of  four  hours  there  were  only  a  fe«' 
free  trypanosomes  left. 
A  few  Glossinn  morsitans  were  fed  on  a  dog.  In  one  series  ten 
flies  were  fed,  and  examined  at  intervals  of  from  one  and  a  quarter  to 
three  hours  later.  None  of  them  contained  parasites.  In  another 
series  eleven  were  fed  on  the  same  dog,  and  examined  forty-eight 
I  t 
