3U 
may  not  perceive  when  he  is  bitten.  This  is  by  no  means  uncom- 
mon, especially  when  people  are  asleep,  and  I have  known  cases 
where  two  people  were  sleeping  close  together,  and  one  complained 
that  he  could  not  sleep  on  account  of  the  mosquitoes,  while  the  other 
said  he  did^not  notice  them.  Secondly  mosquitoes  may  be  absent 
or  very  few  in  number,  and  their  absence  may  be  attributed  to  the 
presence  of  the  basil  plants.  It  was  necessary,  therefore  to  devise 
some  means  of  excluding  the  personal  element,  and  it  appeared  to 
me  that  if  two  specially  constructed  cases  were  placed  at  some  dis- 
tance from  each  other,  with  free  communication  between  them, 
these  requirements  would  be  met.  Accordingly  I had  two  cases 
made,  three  sides  of  which  were  covered  with  mosquito  net,  and  the 
fourth  was  solid.  In  this  side  a large  circular  opening  was  made. 
The  cases  were  placed  with  the  solid  sides  facing  each  other,  and 
the  openings  were  connected  with  a short  tube  of  mosquito  netting. 
Along  this  tube  it  was  found  the  mosquitoes  would  fly  easily.  I 
attach  an  account  of  the  experiments  indetiil  for  reference,  but 
here  I content  myself  with  epitomising  the  details. 
The  mosquitoes  used  were  bred  out  from  larvae  and  consisted 
mainly  of  Stegomyia,  the  most  common  mosqu'to  in  Freetow.i  but 
there  were  also  a few  Anopheles  and  Culices. 
In  one  case  (called  No.  i)  tvyo  large  basil  plants  (about  i8  inches 
high)  were  placed,  along  with  some  ripe  bananas  as  food  for  the 
mosquitoes.  In  the  other  case  (No.  2)  only  bananas  were  placed 
so  as  to  equalise  the  conditions  as  regards  food.  On  the  3rd  June, 
eleven  mosquitoes  were. liberated  in  case  No.  I.  They  showed  no' 
discomfort  and  an  hour  afterwards  were  found  perched  on  the  roof 
and  sides  of  the  case.  On  the  morning  of  the  4th  sevem  were 
counted  in  case  I,  one  was  in  the  tube  and  none  in  c is  i No.  2. 
Two  free  mosquitoes  were  found  perched  on  the  outside  of  the 
netting  of  case  No.  i (containing  the  plants).  This  is  a very  im- 
portant observation  as  showing  that,  although  there  was  a large 
room  to  perch  in,  the  attraction  of  the  food  overcame  the  antipathy 
to  the  basil-plant.  On  the  5th  one  only  was  found  in  case  No.  2, 
the  rest  in  case  No.  i.  On  the  6th  the  same  condition  was  found 
and  the  experiment  was  concluded.  It  may  be  mentioned  here  that 
in  all  the  experiments  the  number  of  mosquitoes  gradually  dimin- 
ished owing  to  their  dying.  In  some  cases  the  dead  bodies  were 
found,  in  others  they  were  eaten  by  small  ants.  Here,  therefore, 
it  was  clearly  shown  that  mosquitoes  would  remain  in  close  proxi- 
mity to  basil  plants  although  there  was  a free  way  of  escape,  and  a 
supply  of  food  accessible  elsewhere.  I repeated  the  experiments 
a second  time  with  similar  results.  On  this  occasion  one  mos- 
quito settled  on  a leaf  and  remained  there  for  some  minutes,  and 
on  the  following  day  two  were  perched  on  the  stem  of  the  plant  for 
over  an  hour.  A free  mosquito  was  found  on  the  netting  outside 
the  case.  Several  mosquitoes  were  observed  perched  on  the  wall 
of  rooms  at, distances  varying  from  six  to  twelve  feet  from  the  basil 
plants. 
In  the  third  experiment  the  conditions  were  altered.  Three  pots 
