490  Cat  Method  for  Testing  Digitalis^  {%ciT'i,mh'' 
My  results  with  cr\  stalline  ouabain  would  indicate  that  the  lethal 
dose  of  this  substance  varies  considerably  with  different  animals. 
It  seems,  then,  irrational  to  estimate  the  value  of  a  preparation  of 
digitalis,  from  its  supposed  equivalent  of  a  body  which  is  in  itself, 
for  any  given  animals,,  an  unknown  quantity.  The  authors  of  this 
method  claim  that  crystalline  ouabain  will  exactly  replace  digitalis 
in  regard  tO'  its  toxicity  on  the  cat.  It  seems  to  me,  however,  that 
there  might  be  some  variance  in  its  power  to  exactly  replace  differ- 
ent samples  of  digitalis  depending  on  the  proportion  of  active  prin- 
ciples present  and  the  conditions  of  these  principles,  whether  or  not 
decomposed.  Since  the  amount  of  digitalis  to  be  injected  which  will 
represent  50-75  per  cent,  of  the  required  amount  is  an  unknown 
quantity,  it  necessarily  follows  that  the  amount  of  ouabain  required 
to  complete  the  experiment,  even  if  its  toxicity  could  be  exactly 
known,  is  an  unknown  quantity.  Therefore,  not  knowing  the  amount 
of  ouabain  required,  the  rate  of  injection,  which  probably  plays  an 
important  part,  cannot  be  known.  Lastly,  the  time  required  to  kill, 
being  dependent  on  the  rate  of  injection,  constitutes  another  unknown 
factor.  So,  when  testing  a  sample  of  digitalis,  one  has  to  deal  with 
six  or  more  unknown  factors.  This  requires  an  operator  of  con- 
siderable experience  and  skill. 
Summary. 
Considering  the  results  of  this  work,  together  with  my  experi- 
ence with  the  other  methods,  I  am  led  to  make  the  following  state- 
ments in  conclusion  : 
The  cat  method  of  Hatcher  and  Brody  is  unquestionably  the 
most  complicated  and  difficult  of  all  the  American  methods,  requir- 
ing an  operator  of  considerable  experience  in  animal  experimentation. 
It  is  not  a  method  that  will  be  found  convenient  and  generally 
serviceable  by  the  retail  pharmacist. 
It  is  more  time-consuming  than  the  other  methods,  rec[uiring  con- 
stant attention  when  started. 
The  item  of  expense,  like  that  of  the  guinea-pig  method,  is 
decidedly  in  its  disfavor. 
The  procuring  of  a  sufficient  number  of  suitable  animals  is  a 
practical  impossibility  for  the  manufacturing  pharmacist  having  a 
large  number  of  preparations  to  test.  This  may  also  l)e  the  source 
of  much  unpleasantness  and  trouble. 
