preceding  investigation,  in  particular,  is  their  action  a  chemical  action, 
or  does  a  purely  physical  process  take  place  ? 
With  a  view  to  finding  an  answer  to  this  enquiry,  it  was  decided 
to  determine  the  reaction  rate  at  which  haemolysis  proceeds  m  the 
heterogeneous  system  formed  by  red  blood  cells  suspended  in  liquid 
containing  a  haemolytic  agent.  This  was  found  to  be  possible  only 
in  the  case  of  quinine  in  the  alkaloidal  state,  the  remaining  three 
haemolytic  agents  causing,  in  addition  to  haemolysis,  decomposition 
of  the  liberated  haemoglobin,  and  thus  preventing  the  estimation  of 
the  amount  liberated  from  being  carried  out. 
The  mode  of  procedure  finally  adopted  for  the  determination  of 
the  reaction  rate  of  quinine  in  the  alkaloidal  state  was  as  follows.-  — 
Method.  About  loo  c.cm.  of  an  emulsion  of  washed  red  cells  in 
0'9  per  cent,  sodium  chloride  solution  containing  0’0358  per  cent.  = 
O'OOioS  M.  of  quinine  was  prepared.  This  was  placed  in  a  water  bath 
kept  within +  0'5°  C.  of  the  temperature  selected  for  experiment.  The 
emulsion  was  shaken  up  every  fifteen  minutes  and  at  the  end  of  an 
hour,  and  subsequently  at  half  hourly  intervals,  samples  of  the 
emulsion  measuring  about  lo  c.cm.  were  pipetted  off  and  the  per¬ 
centage  of  haemoglobin  (in  terms  of  the  equivalent  amount  of  health)’ 
red  cells  in  the  moist  condition)  present  in  the  supernatant  liquid  after 
centrifugalisation,  diluted  if  necessary  with  one  to  three  parts  ot 
distilled  water,  determined  by  means  of  a  haemoglobmometer,  the 
absolute  value  of  the  readings  of  which  had  been  previously  ascer¬ 
tained  in  terms  of  the  equivalent  weight  of  red  cells  of  one  of  us  taken 
as  a  standard.  Inasmuch  as  the  laked  haemoglobin  undergoes  a  slight 
deterioration  during  the  course  of  the  experiment,  a  small  correction, 
representing  the  mean  of  several  determinations,  and  amounting  to 
0'5  division  of  the  haemoglobinometer  scale  per  hour,  was  applied  to 
the  leadings.  This  correction  is  slight  for  the  earlier  haemoglobin 
estimations,  and  negligible  for  the  later  ones.  The  haemoglobino- 
metei  leadings  were  readily  made,  the  variations  between  successive 
readings  of  the  same  .solution  lying  within  +  2  per  cent,  of  the  whole, 
and  for  readings  lying  between  25  and  G5  on  the  scale  being  usually 
identical. 
A  series  of  estimations  of  the  actual  and  percentage  amounts  of 
red  blood  cells  haemolysed  in  successive  periods  of  time  is  given  in 
Table  13.  The  initial  concentra.tion  of  red  cells  in  the  emulsion  was 
