Amka?ch;  im.m' }    Standardisation  of  Digitalis  Bodies.  105 
cat  method  will  be  seen  by  consulting  the  figures  for  the  different 
standardizations  of  ouabain  made  at  different  times  and  often  with 
"unknown  solutions."  For  example,  the  two<  series  reported  in  the 
table  show  the  following:  r.  If  the  uncorrected  results  are  used 
the  average  of  the  first  series  falls  only  2  per  cent,  below  the 
established  standard,  while  that  of  the  second  series  lies  9  per  cent, 
above  this  standard.  2.  When  the  averages  taken  are  based  upon 
the  corrected  series  (after  the  elimination  of  the  very  abnormal 
animals)  A  shows  an  absolutely  correct  result  and  B  gives  a  figure 
5  per  cent,  above  the  standard.  Ouabain  is  selected  to  show  the 
extent  of  the  variations  obtained  by  standardizations  made  at  dif- 
ferent times  because  it  is  the  purest  of  the  isolated  glucosides  and 
is  the  least  liable  to  variation  in  its  strength.  Similar  close  estima- 
tions have  been  found  in  repeated  tests  of  a  great  many  other 
digitalis  bodies  made  in  this  laboratory,  the  results  often  being 
obtained  by  several  different  observers  and  with  "  unknowns."  An 
instance  of  this  is  found  in  the  case  of  the  digitalis  leaf  which 
supplied  the  tincture  marked  Ger.  A.  This  leaf  yielded  an  assay 
of  approximately  120  mg.  per  kg.  of  cat  as  its  unit  when  other 
preparations  were  used — both  tinctures  and  infusions. 
The  guineapig  method  is  stated  8  to  provide  a  means  of  making 
preparations,  "  which  are  always  within  ten  per  cent,  of  the  same 
strength."  But  Haskell 9  expresses  his  belief  in  the  variability  of 
guineapigs  and  says,  "  I  have  been  unable  to  find  the  report  of  a 
single  series  of  experiments  performed  with  the  object  of  showing 
that  guineapigs  are  not  fully  as  much  influenced  by  adventitious 
circumstances  as  are  frogs."  Vanderkleed  7  replies  to  this  assertion 
citing  the  results  of  a  series  of  43  tests  in  which  only  one  animal 
showed  a  variation  as  great  as  10.4  per  cent.  Houghton  10  found 
these  animals  unsatisfactory,  and  Hale  5  says,  "  certain  animals  sur- 
vived doses  20  per  cent,  larger  than  had  killed  others.  Thus,  4 
were  killed  with  0.5  mg.  per  gram  of  body  weight,  3  lived  with 
0.6  mg.  doses  of  the  same  drug."  Lastly,  Haskell 11  reports  several 
series  of  standardizations  of  ouabain  on  guineapigs  and  shows  the 
occurrence  of  variations,  under  similar  conditions,  ranging  from 
11  to  30  per  cent,  of  the  average  found  in  the  series  under  con- 
8  a.  j.  pharm.  ;  1910;  82;  p.  453. 
9 a.  j.  pharm.;  191 1 ;  83;  p.  201. 
10J.A.M.A.,  1898,  xxxi;  959- 
11  A.  J.  PHARM.;   1912;  84;  p.  24I. 
