i|8 
represent  altered  red  cells.  Another  difficulty  arises  when  the 
attempt  is  made  to  explain  suppression  of  urine  on  the  hypothesis  of 
glomerular  haemorrhcige.  Here  it  becomes  necessary  to  assume 
that  the  glomerular  haemorrhage  is  general  throughout  the  kidney, 
for  otherwise  a  certain  number  of  uriniferous  tubules  would  remain 
to  carry  on  the  work  of  the  kidney.  It  is  easy  to  imagine  that 
malarial  parasites  may  be  present  in  the  glomeruli  and  damage  the 
capillary  wall  thus  leading  to  haemorrhage,  but  the  necessity  of  sup¬ 
posing  that  this  is  general  throughout  the  glomeruli  of  both  kidneys 
and  does  not  occur  elsewhere  in  the  blood  vessels  of  the  kidneys  (for 
as  IS  pointed  out  below,  interstitial  haemorrhage  is  not  met  with) 
seems  to  be  a  fatal  objection  to  the  hypothesis  in  question.  It  may 
be  observed  in  passing  that  a  high  percentage,  namely  ri  per  cent 
to  27  per  cent,  of  haemoglobin  in  the  urine  was  observed  in  Cases  /a 
and  II  in  which  suppression  occurred,  but  examination  of  the  urine 
voided  at  intervals  of  several  hours  does  not  afford  a  comprehensive 
idea  of  what  is  taking  place  111  the  kidney.  This  will  be  made  clear 
at  once  if  the  varying  amounts  of  urine  collected  from  the  bladder  at 
short  intervals  in  Rabbit  10,  Table  35  (Fig.  9,  p.  84),  are  considered. 
It  is  obvious  that  only  when  the  urine  is  collected  continuously,  and 
its  haemoglobin  content  determined  at  short  intervals,  can  a  clear 
idea  of  the  limits  of  the  variation  of  the  rate  of  elimination  of  haemo¬ 
globin  be  obtained. 
3.  As  far  as  the  occurrence  of  relapses  is  concerned  the 
hypothesis  of  a  glomerular  haemorrhage  being  the  source  ol  the 
haemoglobiniiria  of  blackwater  fever  must  be  equally  applicable  to  the 
relapses  and  the  original  attack.  This  hypothesis  if  applicable  to  the 
original  attack  should  be  sufficient  to  explain  also  relapses  of  slight 
extent,  such  as  those  seen  in  Case  10  (Table  50). 
4.  The  appearance  of  red  blood  cells  in  the  urine  would  be 
expected  as  an  occasional,  if  not  a  frequent,  event  in  glomerular 
haemorrhage.  It  might  be  exjiected  that  partially  decolourised  red 
cells  and  red  cell  stromata  would  be  often  met  with,  and  that  red 
cells  which  were  perfectly  healthy  in  aspect  would  be  unusual,  b 
most  of  the  cases  (Table  43)  in  which  red  cells  were  found  in  the 
urine  in  blackwater  fever,  the  majority  of  these  cells  were  normal  m 
aspect.  » 
I 
