u 
preceding  and  following  this  period,  which  are  given  in  brackets, 
shows,  as  is  also  seen  in  Tables  13  and  14,  the  extent  to  which  at  the 
beginning  haemolysis  proceeds  slowly  and  towards  the  end  progresses 
with  increased  rapidity.  The  cause  of  this  initial  delay  and  final 
acceleration  of  the  process  of  haemolysis  is  not  (dear ;  it  may 
perhaps  be  dependent  upon  the  jihysical  conditions  under  which  the 
interaction  of  cell  and  haemolytic  agent  occurs.  It  is,  however, 
established  by  the  above  experiments  that  the  reaction  in  question 
during  the  greater  part  of  its  course  resembles  a  chemical  reaction 
monomolecular  in  character,  presenting,  however,  an  initial  retarda¬ 
tion  and  final  acceleration. 
In  Table  16  the  series  of  experiments  given  in  Table  13  is 
repeated,  and  seven  other  series  added.  In  these  the  reaction  rate 
constant  is  calculated  for  the  period  t,  to  t^,  the  starting  point  being 
t^  instead  of  t„  as  in  Table  13.  This  method  of  calculation  has  been 
adopted  because  of  the  difficulty  of  ascertaining  the  precise  moment  at 
which  the  temperature  of  the  emulsion  becomes  identical  with  that  of 
the  water  in  the  bath.  In  these  different  series  of  experiments 
haemolysis  is  seen  to  follow  pretty  closely  a  monomolecular  rate 
except  at  the  beginning  and  end  of  the  reaction,  when  the  respective 
slowing  and  acceleration  already  referred  to  occur,  while  on  the  other 
hand  a  bimolecular  or  multimolecular  rate  is  excluded  by  the  much 
wider  range  of  variation  in  the  value  of  the  constants  calculated  for 
such  reaction  rates.  The  actual  reaction  rate  as  shown  by  the  values 
of  K  varies  from  day  to  day  even  in  the  same  healthy  individual,  the 
variations  being  greater  in  the  series  Y,  to  than  in  the  series 
to  The  cause  of  these  variations,  which  appear  to  lie  outside 
the  range  of  experimental  error,  is  obscure.  It  is  true  that  variations 
occur  in  the  red  blood  cells  employed  in  different  experiments.  In 
particular,  it  was  constantly  found  that  although  the  percentage 
weight  of  wet  red  cells  in  the  emulsions  was  always  the  same,  namely, 
_  0'444  per  cent.,  being  determined  by  means  of  the  haemocrit,  yet 
when  the  red  blood  cell  content  of  the  emulsions  was  determined 
(after  laking  with  distilled  water)  by  means  of  a  haemoglobinometer 
reading  the  result  was  different  in  each  case  Thus  in  the  experi¬ 
ments  recorded  in  Table  16  the  haemoglobinometer  determinations  of 
the  emulsions  were  found  to  vary  from  0’476  per  cent,  to  0'532  per 
cent.  These  variations  are  too  large  to  be  accounted  for  as  due  to 
