Amjour,  Pharm.  ^       Theories,  of  Blood-Coagulation  707 
lets,  we  observed  that  the  former  serum  contains  a  much  larger  quan- 
tity of  thrombin  than  the  latter.  Consequently,  the  platelets  actively 
participate  in  the  production  of  the  coagulating  principle.  This  fact 
can  be  proved  more  distinctly  by  the  following  experiment :  A  sedi- 
ment constituted  exclusively  of  these  small  cells  is  obtained  by  vig- 
orously centrifugalizing  oxalated  plasma,  previously  freed  of  its 
red  and  white  corpuscles,  but  containing  still  its  platelets.  This 
platelet  deposit,  thoroughly  washed,  is  emulsified  in  physiological 
solution,  and  one  drop  of  the  turbid  emulsion  thus  obtained  is  added 
to  a  certain  quantity  of  a  serum  which,  being  derived  from  the  slow 
coagulation  of  recalcified  oxalated  plasma  that  has  been  freed  of 
its  own  platelets,  is  by  itself  very  poor  in  thrombin.  Within  20 
or  30  seconds,  the  mixture  becomes  capable  of  coagulating  a  suit- 
able amount  of  oxalated  plasma  almost  instantaneously;  in  other 
words  the  reaction  of  serum  with  platelets  generates  an  abundant 
supply  of  thrombin.  It  must  be  pointed  out  that  this  experiment 
closely  resembles  that  of  Morawitz  except  that  platelets  instead  of 
tissue  cells  are  added  to  the  serum.  Tissue  cells  and  platelets  both 
contain  one  of  the  generators  of  thrombin,  which  may  be  called 
cytozyme.  The  second  one,  the  serozyme,  exists  in  the  serum.  It 
can  be  easily  demonstrated  that  the  reaction  between  serum  and 
platelets,  that  is  to  say,  between  serozyme  and  cytozyme,  takes  place 
only  in  the  presence  of  soluble  calcium  salts,  no  thrombin  appearing 
if  the  serum  has  been  decalcified  before  the  introduction  of  the 
platelets.  Moreover,  I  shall  further  insist  on  the  fact  that  the 
two  substances  unite,  that  they  really  consummate  each  other.  In- 
deed, experience  shows  that  when  a  serum  has  been  treated  once 
with  platelets,  and  has  yielded  thrombin,  this  serum  is  subsequently 
incapable  of  reacting  with  a  new  amount  of  platelets,  its  serozyme 
having  been  exhausted  by  the  first  reaction.  It  follows  that  a  serum 
produced  by  the  coagulation  of  a  plasma  rich  in  platelets  and  which 
of  course  contains  much  thrombin,  is  considerably  less  rich  in 
serozyme,  that  is,  considerably  less  capable  of  reacting  with  new 
platelets,  than  is  a  serum  derived  from  a  plasma  deprived  of  most 
of  its  platelets.  This  is  precisely  what  experience  shows.  It  is, 
therefore,  highly  advisable  always  to  employ  for  serozyme  a  serum 
obtained  by  the  coagulation  of  oxalated  plasma  which  has  been  care- 
fully freed  of  its  platelets  before  recalcification. 
Serozyme  is  a  thermolabile  substance,  easily  destroyed  by  heat. 
