AmbJc°turi92iiarm:'j       Theories  of  Blood-Coagulation  709 
added  to  blood  flowing  from  the  artery,  prevents  its  coagulation.  By 
centrifugalization  and  pipetting  off,  a  clear  plasma  is  obtained,  which 
remains  fluid,  even  when  platelet  emulsion,  or  tissue  extract,  or 
lipoidic  cytozyme,  is  added.  This  is  easily  understood ;  both  mother- 
substances,  serozyme  and  cytozyme,  are  equally  necessary  to  the 
production  of  thrombin;  it  is  of  no  use  to  add  one  of  them  if  the 
other  is  absent.  But  such  a  plasma,  which  we  may  for  the  sake  of 
brevity  call  "phosphate  plasma,"  clots  under  the  influence  of  throm- 
bin or,  which  naturally  is  the  same,  when  both  mother-substances 
are  added.  It  behaves  as  an  excellent  reagent  for  the  detection  of 
thrombin.  Since  its  composition  closely  resembles  that  of  the  orig- 
inal plasma;  it  may  be  considered  as  being  fibrinogen  dissolved  in  a 
normal  medium. 
But  tricalcic  phosphate  is  endowed  with  a  property  which  is 
remarkably  useful  for  technical  purposes.  As  is  well  known,  it  is 
capable  of  being  dissolved  in  physiological  solution  under  the  in- 
fluence of  a  current  of  carbonic  oxide  gas.  Consequently,  phosphate 
which  has  been  added  to  plasma  and  has  absorbed  the  serozyme, 
after  having  been  thoroughly  washed^  can  liberate,  thanks  to  its 
own  dissolution,  the  active  principle  it  had  withdrawn.  In  this  way, 
Delange  and  myself  succeeded  in  bringing  about  the  isolation  of 
serozyme,  which,  on  the  addition  of  cytozyme  extracted  from  plate- 
lets, gave  abundant  thrombin.  Thus  in  the  course  of  the  whole 
experiment  the  factors  which  determine  coagulation  are  in  reality 
subjected  to  an  analysis  followed  by  a  synthesis. 
As  mentioned  above,  our  assumption  that  serozyme  and  cyto- 
zyme are  the  generators  of  thrombin  involves  the  idea  that  those 
mother-substances  really  unite  to  form  a  compound,  which  is  throm- 
bin. This  ought  to  be  demonstrated  also  with  regard  to  pure  cyto- 
zyme, I  mean  a  cytozyme  in  the  condition  of  a  lipoidic  extracted 
matter.  If  the  union  really  occurs,  we  may  anticipate  that,  if  a  given 
quantity  of  serozyme  be  mixed  with  a  sufficient  amount  of  cyto- 
zyme, thrombin  will  be  engendered  and  the  serozyme  will  be  ex- 
hausted; in  other  words,  it  will  be  no  longer  capable  of  yielding 
fresh  thrombin  when  a  new  amount  of  cytozyme  is  added.  Such 
is  indeed  the  case.  Serum  yielded  by  coagulation  of  recalcified 
platelet-free  oxalated  plasma  is  divided  into  two  parts,  lipoidic 
cytozyme  being  added  to  one  of  them,  the  other  portion  being  kept 
as  it  is.    In  the  tube  containing  both  serum  and  cytozyme,  thrombin 
