286 
the  flies  found,  and,  curiously,  the  one  or  two  instances  of  this  which 
we  have  seen,  have  been  the  reverse  of  the  usual  statement  that  they 
are  more  abundant  in  the  wet  season. 
The  Lovu  river  was  visited  by  Dr.  Spillane,  the  P.M.O.  of  N.E 
Rhodesia,  in  November,  1907,  when  the  rains  were  on,  and  at  that 
time  no  flies  were  seen  by  him ;  yet  when  we  got  there  in  June,  igo8, 
two  months  after  the  close  of  the  wet  season,  they  were  particularly 
plentiful.  It  is  not  clear  why  this  should  be  so,  for  along  this 
particular  stretch  of  the  river  there  are  many  villages  and  an  assured 
supply  of  food.  On  an  island  in  Lake  Tanganyika,  which  was  visited 
at  the  height  of  the  dry  season  (July),  and  again  at  the  commencement 
of  the  rains  (November),  there  was  no  apparent  difference  in  their 
numbers,  but  these  were  so  great  that  it  would  be  difflcult  to  draw 
any  definite  conclusions.  On  the  lake  shore,  however,  fly  were  caught 
during  the  rains  at  several  places  where  they  were  not  seen  in  the 
dry  season. 
There  is  apparently  a  seasonal  variation  in  the  proportion  of  the 
sexes.  On  the  island  referred  to,  of  1,409  specimens  of  Gl.  palfdu 
caught  in  July  (height  of  dry^  season),  1,282  were  males  and  127 
females,  a  proportion  of  lO  :  i.  In  November,  just  after  the  rains 
had  set  in,  of  401  specimens  caught  at  the  same  place,  396  were  males 
and  only  five  females,  a  proportion  of  79  :  i.  This  may  indicate  that 
it  is  at  this  time  of  the  year  that  the  females  are  depositing  larvae, 
and  that  during  this  process  they  become  less  active  than  in  the  dry 
season,  when  the  humidity  and  the  temperature  are  lower. 
In  spite  of  prolonged  and  careful  search,  we  were  never 
successful  in  finding  pupae,  and  in  this  our  experience  coincides  with 
that  of  Ensor*  in  the  Soudan.  It  may  have  been  due  to  the  fact  that 
the  island  was  composed  of  very  rough,  broken  stone,  with  very  little 
soil,  so  that  the  pupae  might  easily  have  dropped  into  the  crevices  of 
the  rocks  where  they  could  not  be  found.  ' 
Our  search  was  made  along  the  shores  of  an  island,  just  above  the  edge  of  the 
water  and  two  or  three  yards  inland.  We  were  assisted  by  five  or  six  ‘fly-boys'  ^ 
trained  in  the  work,  and  although  they  often  brought  us  pupa  cases  they  were  never  1 
those  of  Gl.  falfalis.  On  each  occasion,  several  hours  were  devoted  to  the  work, 
and  during  our  stay  on  the  island,  in  the  months  of  July  and  November,  we  J 
searched  at  least  half  a  dozen  times.  We  always  selected  the  places  where  the  [ 
shade  was  most  abundant. 
*  Knsor,  rgo8.  Reports  of  the  Wellcome  Research  Laboratory,  No.  3.  I 
I 
‘1 
I 
I 
