I 
274 
argued  that  if  hereditary  transmission  does  occur,  all  the  flies  in  such 
a  restricted  area  should  have  been  infected.  As  is  known,  the  infection 
tends  to  die  out  of  the  intestine,  or  at  least  apparently  so,  though  the 
persistence  of  unrecognised  forms  of  the  parasites  cannot  yet  be 
excluded.  This,  however,  depends  on  the  lack  of  suitable  pabulum, 
and  it  is  within  the  bounds  of  possibility  that  when  a  ‘  negative  ’  fly 
does  get  a  meal  of  the  required  sort,  the  ordinary  forms  of  the  flagel¬ 
lates  may  be  evolved  from  the  unrecognised  forms.  In  the  flies  we 
examined,  traces  of  a  meal  (blood)  were  found  in  33' i  per  cent  and 
parasites  in  42  per  cent.,  but  as  the  blood  is  digested  in  a  shorter 
period  of  time  than  the  parasites  take  to  disappear  from  the  intestine, 
this  discrepancy  can  be  accounted  for.  To  some  degree,  the 
reinfection  of  the  intestine  may  be  due  to  ingestion  of  the  proboscidal 
forms,  provided  that  the  fly  feeds  a  second  time  before  these  have 
died  out. 
The  further  point  also  arises  as  to  the  identity  of  the  forms  seen 
in  Gl.  morsiians.  In  general  they  appear  to  be  similar  to  those  found 
in  Gl.  palpalis,  except  in  minor  points.  If  they  are  the  same  species, 
many  species  of  vertebrates  must  harbour  the  parasites,  if  we  accept 
the  view  that  they  are  derived  from  vertebrates,  since  Gl.  morsitam 
feeds  to  a  great  degree  on  a  different  fauna  from  that  on  which 
Gl.  palpalis  feeds.  There  is  no  reason  to  suppose  that  the  two 
species  are  identical  other  than  the  similarity  in  fonn,  and  this 
is  not  a  decisive  feature  for  the  determination  of  species  amongst  the 
trypanosomes. 
Diversity  of  opinion  also  exists  as  to  the  origin  of  the  forms  found 
in  the  proboscis.  Stuhlmann^  considers  that  they  are  derived  from 
the  intestinal  forms,  the  infection  of  the  proboscis  being  the  final 
stage  in  the  development  of  ingested  trypanosomes.  Roubaud,®  on 
the  other  hand,  thinks  that  the  infection  in  the  proboscis  is  an  entirely 
distinct  thing  and  that  it  is  to  be  accounted  for  by  the  fixation  and 
multiplication  of  pathogenic  trypanosomes,  in  situ,  which  have  been 
ingested  in  the  act  of  feeding. 
There  are  several  points  which  tend  to  invalidate  Roubaud's 
conclusions.  He  does  not  state  whether  he  examined  the  proboscides 
of  freshly-caught,  unfed  flies  for  the  presence  of  flagellates,  and  it  is 
possible  that  he  mistook  the  naturally-occurring  forms  for  those 
