608         Detection  of  Ferments  and  Poisons  in  Blood.{^m-£™*i^&vm- 
nizable  quantities  of  zinc  are  to  be  found  in  the  filtrate  only  if  com- 
pounds of  organic  acids  were  present  in  the  blood.  Otherwise 
mere  traces  of  zinc  exist  as  zinc  albuminate,  in  smaller  proportions 
the  more  the  zinc-powder  is  free  from  zinc  oxide.  As  the  deposit 
— a  solid,  reddish-brown,  bulky  mass — retains  no  organic  poisons 
except  traces  of  hydrocyanic  acid  and  carbon  monoxide,  we  have 
in  the  filtrate  all  the  glycosides,  alkaloids,  ptomaines,  toxalbumins, 
enzymes,  amides,  etc. ;  in  short,  the  examination  of  the  filtrate  is 
exclusively  important  for  forensic  chemistry.  Concerning  the  solid 
residue  we  need  merely  remark  that  the  hydrocyanic  acid  may  be 
removed  from  it  by  extraction  with  alcohol  and  carbon  monoxide 
by  exhaustion  in  the  air-pump. 
The  filtrate  is  incomparably  more  suitable  than  the  original  blood 
as  well  for  the  chemical  as  for  the  physiological  detection  of  poisons. 
If  its  action  upon  animals  is  to  be  examined  it  is  only  requisite  to 
shake  up  a  portion  with  a  drop  of  solution  of  sodium  sulphide,  and 
to  filter  off  the  precipitate  of  zinc  sulphide.  If  the  proportions  are 
duly  adjusted  the  liquid,  free  from  zinc  and  sodium  sulphide,  can 
be  injected  subcutaneously  into  a  mouse  or  into  the  circulatory  sys- 
tem of  a  kitten.  However,  the  filtrate  can  often  be  applied  without 
the  elimination  of  the  zinc.  The  bacteria  which  swarm  in  putrid 
blood  are  almost  entirely  left  behind  in  the  residue.  Where  an 
absolute  determination  of  bacteria  is  required  the  filtrate  may  be 
again  passed  through  a  Chamberland  filter,  and  thus  the  last  resi- 
dues of  the  germs  may  be  removed.  If  the  animal  experimented 
upon  exhibits  grave  phenomena  of  poisoning,  a  second  portion  of 
the  original  filtrate  is  mixed  with  a  drop  of  solution  of  potassium 
ferrocyanide  and  a  little  acetic  acid,  the  precipitate,  which  includes 
all  the  zinc  and  the  albuminoids,  is  filtered,  the  filtrate  is  neutral- 
ized and  injected  into  a  second  animal.  If  this  animal,  in  contra- 
distinction to  the  former,  exhibits  no  phenomena  of  poisoning,  the 
symptoms  observed  in  the  former  subject  must  have  been  occasioned 
by  an  albuminoid  poison,  as  in  such  dilute  liquids  other  poisons  are 
not  precipitated  by  potassium  ferrocyanide,  or  at  most  a  portion  of 
strychnine,  which  can  be  easily  detected  in  the  precipitate. 
The  isolation  of  the  poisonous  albuminoid  substance  can  be 
undertaken  only  in  the  main  portion  of  the  original  filtrate,  either 
by  treatment  with  alcohol  in  the  ordinary  manner  or  by  "  salting 
out"  with  ammonium  sulphate.     In  those  toxalbumins,  however, 
