704 
Theories  of  Blood-Coagulation  \Am'cy?™\m*™' 
serted  into  the  vessel  has  not  been  allowed  to  touch  the  wound,  and 
consequently  no  trace  of  tissue  extract  has  become  mixed  with  the 
blood.  As  a  matter  of  fact,  this  precaution  ought  to  be  taken  regu- 
larly, whatever  may  be  the  species  of  animal  under  experiment,  as  it 
is  quite  a  general  rule  that  tissue  extract  accelerates  coagulation, 
this  auxiliary  influence  being  particularly  noted  for  birds,  because 
this  blood  is  not  so  capable  of  spontaneous  coagulation  as  is  that  of 
mammals. 
Thanks  to  such  methods,  the  separation  of  the  two  constitu- 
ents, cells  and  plasma,  can  be  attained  before  any  beginning  of 
coagulation  and — let  us  emphatically  insist  upon  this  essential  fact — 
before  any  appearance  of  the  coagulating  principle,  thrombin.  There 
is  no  need  to  remind  you  that  serum  yielded  by  coagulation  con- 
tains thrombin. 
We  must  now  try  to  go  further  and  subject  plasma  and  cells 
to  a  closer  analysis.  Let  us  consider  first  the  plasma.  Soluble  cal- 
cium salts  are  necessary  to  coagulation.  How  do  they  act?  Pekel- 
haring  and  Hammarsten  have  shown  the  essential  fact  that 
these  salts  are  not  necessary  to  the  transformation  of  fibrinogen 
into  fibrin  under  the  influence  of  thrombin,  but  are  indis- 
pensable to  the  formation  of  the  latter,  that  is,  to  the  production  of 
thrombin  from  the  mother-substances  already  present  in  the  circulat- 
ing blood.  The  production  of  thrombin  is  prevented  by  oxalate, 
but  on  the  other  hand  decalcification  does  not  prevent  the  coagula- 
tion of  the  fibrinogen  by  completed  thrombin.  Indeed,  it  has  been 
proved  that  blood,  oxalated  immediately  after  withdrawal  from  the 
artery,  remains  permanently  fluid,  no  thrombin  being  ever  detected 
in  it,  whereas  if  serum  yielded  by  normally  clotted  blood  be  oxalated, 
this  oxalated  serum,  added  to  oxalated  plasma,  causes  the  coagula- 
tion of  the  latter.  From  these  facts  it  follows  that  oxalated  plasma 
is  a  most  suitable  reagent  for  the  detection  of  thrombin  in  a  given 
liquid ;  indeed,  an  estimation  of  the  coagulating  power  of  such  a 
liquid  may  be  made  by  taking  into  consideration  the  quantity  of 
oxalated  plasma  coagulated  by  a  certain  amount  of  this  liquid,  or 
the  rapidity  of  the  occurring  coagulation.  But  it  must  be  borne 
in  mind  that,  at  least  when  present  in  serum,  the  activity  of  a  given 
thrombin  depends  not  only  upon  its  quantity  but  also  upon  its  age. 
The  capacity  of  fresh  serum  to  coagulate  oxalated  plasma  decreases 
very  quickly,  by  a  spontaneous  attenuation  of  the  thrombin,  and  this 
