at  first  in  the  absence  of  quinine,  and  then  still  continuing  when,  after  ' 
a  few  days,  the  administration  of  quinine  was  recommenced,  without, 
however,  presenting  any  definite  relation  to  the  time  of  administration  i 
of  quinine.  In  this  case  the  administration  of  quinine  did  not  appear 
to  call  forth  the  relapses,  nor  did  any  marked  improvement  occur  j 
during  the  period  when  quinine  was  discontinued. 
It  has  already  been  pointed  out  that  a  possible  explanation  of  the  I* 
occurrence  of  haemoglobinuria  after  quinine  in  malarial  subjects  may  || 
be  the  destruction  of  the  red  cells,  which,  as  already  shown,  are 
capable  of  taking  up  quinine  from  .solution  (Table  21,  p.  41),  by  the  | 
double  action  of  the  malarial  parasite  and  quinine,  but  that  it  is 
doubtful  if  the  number  of  parasites  in  the  red  cells  of  the  peripheral 
blood  is  large  enough  to  admit  of  the  assumption  that  the  haeino-  | 
globin  passing  into  solution  in  the  blood  plasma  is  derived  from  the 
haemolysis  of  these  affected  red  cells,  while  we  are  without  data 
pointing  conclusively  to  the  occurrence  of  haemolysis  in  the  spleen 
or  elsewhere.  It  must,  however,  be  observed  that  we  are  not  on  safe  | 
ground  in  assuming  that  because  quinine  is  effective  in  producing  ^ 
blackwater  in  malarial  subjects  it  would  fail  to  do  so  in  non-malarial  I 
patients.  Such  an  assumption  is  vitiated  by  the  fact  that  the 
administration  of  quinine  in  large  doses  is  in  general  use  only  in  '| 
the  treatment  of  malaria.  In  no  other  disease  is  the  administration  | 
of  this  drug  an  habitual  and  generally  recognised  mode  of  treatment 
Moreover,  on  the  relatively  few  occasions  on  which  quinine  is  j 
administered  in  other  diseases  it  is  usually  gi\'en  in  much  smaller  1 
doses  gr.  to  2  gr.).  Whether  haemoglobinuria  would  occasionally 
occur  if  quinine  were  employed  generally  in  pyrexial  affections  in  ! 
doses  similar  to  those  employed  in  malaria,  is  purely  a  matter  of 
conjecture. 
The  above  remark  is  prompted  by  a  point  of  resemblance  between 
many  cases  of  blackwater  fever  and  the  haemoglobinuric  attacks  ' 
produced  by  haemolytic  serum,  potassium  chlorate  and  other  drugs 
(p.  128).  In  the  latter  cases  haemoglobinuria  is  induced  in  an  animal 
which  has  for  some  time  previously  been  suffering  from  more  or  less  ■ 
severe  constitutional  disturbance.  Frequently  this  is  also  precisely  , 
the  condition  in  which  blackwater  fever  patients  exist  at  or  for  some  |p 
days  before,  the  date  of  appearance  of  haemoglobinuria.  In  each  I* 
case  there  are  present  two  factors  :  more  or  less  severe  constitutional  * 
