In  I  able  43  llic  condition  of  the  urine  in  respect  of  haeinof^lobin 
is  given  ni  all  cxcejDl  three  (4,  y,  13)  of  the  cases  of  blackwatcr  fe\cr 
coming  under  our  observation.  It  will  be  .seen  (Column  3)  that  in 
thirteen  out  of  sixteen  cases  haemoglobin  was  pre.sent  in  solution, 
in  the  form  of  oxyhaemoglobin,  in  amounts  \ar\'ing  from  o' 2  |)cr  cent, 
to  3  8  per  cent.,  while  in  addition  a  further  amount  of  haemoglobin 
was  changed  into  the  brown  substance,  exhibiting  no  bands  in  the 
solar  spectnim,  already  referred  to  (p.  51),  w'hich  gives  rise  to  the 
porter-colour  of  the  urine.  The  total  amount  of  dissolved  haemo 
global  passing  into  the  urine  consists  of  these  tw^o  fractions,  which 
arc  determined  by  the  methods  gii^en  below.  Although  in  three  cases 
(2,  Oil,  /)  haemoglobin  111  solution  could  not  be  recognised  liy  spectro¬ 
scopic  e.xamination  of  the  brown  or  porter-coloured  urine  passed, 
nevertheless  dissolved  haemoglobin  was  in  all  probability  originally 
present,  but  had  been  broken  up  by  the  action  of  the  urine  (cp.  Table 
25-  P-  53);  m  Case  2  the  urine  was  retained  for  several  hours  before 
being  voided,  and  micro-organisms  were  present  in  tlie  bladder  (sec 
cluneal  notes,  p.  178);  m  Case  6a  the  percentage  of  haemoglobin  in 
the  urine  was  o'S  per  cent.  ;  and  m  Case  7  the  percentage  of  haemo¬ 
globin  present  was  also  very  small,  and  the  urine  had  .stood  for  some 
time  before  being  examined.  In  a  subsequent  attack  (7a)  the  last 
patient  passed  urine  containing  dissolved  haemoglobin. 
When  the  urinary  deposit  obtained  after  centrifugalisation  was 
examined  (Column  4,  Table  43),  red  blood  .cells  were  found  to  be 
present  m  ten  out  of  sixteen  attacks,  but  in  three  of  the.se  (7a,  10,  i  i) 
red  cells  were  at  first  absent  from  the  urine  and  made  their  appcarmicc 
subsequently  (clinical  notes,  pp.  195,  207  and  217).  As  a  rule  the  red 
cells  were  not  present  in  large  numbers.  They  were  in  fact  usually 
much  less  numerous  than  the  casts,  but  an  exception  occurred  in 
ase  2,  in  which  the)'  were  present  in  very  considerable  numbers  so 
as  to  form  at  the  bottom  of  the  urine  glass  a  brownish  red  deposit 
visible  to  the  naked  eye.  In  this  case  there  were  a  number  of 
decolourised  red  cells  and  also  red  cell  stromata.  The  red  cells  were 
sometimes  crenated,  but  beyond  this  and  the  partial  decolouration 
just  referred  to  they  did  not  present  any  other  abnormal  appearance. 
In  order  to  estimate  haemoglobin  in  urine,  I  he  following  three 
methods  were  employed  ;  the  first  two  being  employed  for  the 
estimation  of  unaltered  dissolved  haemoglobin,  and  the  third  for  the 
