and  also  for  the  marked  anaemia  which  usually  attends  blackwater 
fever. 
2.  The  hypothesis  of  a  primary  haemoglobinaemia*'  is  in 
harmony  with  another  characteristic  phenomenon  of  blackwater 
fever,  namely,  the  appearance  of  granular  casts  in  the  urine.  The 
experimental  work  on  rabbits  shows  that  haemoglobinaemia  is 
necessarily  attended  with  the  formation  of  these  casts.  The  exact 
mode  in  which  the  granular  material  is  formed  remains  in  both  cases 
undetermined.  The  experiments  made  by  different  obseiwers  as  to 
the  site  of  elimination  of  haemoglobin  are  still  regarded  as 
indecisive.!  It  is  uncertain  whether  dissolved  haemoglobin  isi 
eliminated  by  the  glomerular  or  the  tubular  epithelium:  in  the  former' 
case  the  granules,  which  contain  iron  and  are  evidently  derived  from 
haemoglobin,  would  be  deposited  or  precipitated  in  the  urine  in  the 
tubules  and  would  increase  in  number,  perhaps  also  in  size,  as  the 
renal  pelvis  was  approached  ;  in  the  second  case  the  granules  would 
presmnably  be  formed  in  the  renal  epithelium  and  would  then  be- 
discharged  into  the  renal  tubules,  without  any  subsequent  increase  uj 
size,  though  they  would  become  more  closely  packed  if  water  were| 
removed  from  the  lumen  of  the  tubules  by  the  action  of  the  renal] 
epithelium.  Our  observations  on  Cases  /a,  ii  and  i6  (Figs.  33-4J. 
and  53-58)  do  not  enable  us  to  assert  that  any  increase  in  the  sizeo(| 
the  granules  takes  place  as  the  collecting  tubules  are  approached.  I»| 
the  latter,  coarse  masses  of  stainable  material  were  encountered  more| 
often  than  definite  granules.  Possibly  some  change  in  the  solid] 
material  contained  within  the  tubules  occurred  during  the  tinie| 
elapsing  between  the  cessation  of  haemoglobinuria  and  death,  hi 
the  kidney  of  the  rabbit,  in  experimental  haemoglobinuria| 
(Figs.  45-51  and  59-60),  no  unmistakable  increase  in  size  of  tlif| 
granules  could  be  recognised  as  the  collecting  tubes  were  approached| 
but  only  a  more  dense  packing  of  the  granules  was  observable! 
Several  authors  have,  however,  described  an  increase  in  size  of  tli(, 
*  By  this  it  is  not  meant  to  imply  that  laking  necessarily  occurs  in  the  l>enpl»i 
circulating  blood.  What  is  meant  is  a  haemoglobinaemia  not  secondary  to  reM| 
haemorrhage. 
t  Ponfick,  Experimentelle  Beitrage  zur  Lehre  von  der  Transfusion,  Virch.  Aich.,  iM 
B.  62,  S.  273  (c]5.  Plate),  concluded  that  dissolved  haemoglolrin  was  eliminated  by  ‘“'"‘I 
glomeruli  and  red  cell  fragments  by  the  epithelium  of  the  renal  tubules.  March.ard.i 
Kenutniss  der  feineren  Veranderungen  der  Nieren  bei  Hamoglobinausscheidung,  I 
Arch.,  1883.  B.  91,  S.  267,  who  describes  a  diapedesis  of  red  cells,  also  adopts  tins  view. 
