532 
Methods  for  Digitalis  Assay. 
Am.  Jour.  Pharm. 
August,  1920. 
dog.  By  use  of  a  number  of  dogs  an  average  could  be  selected,  but 
time  and  the  expense  of  labor  and  animals  will  not  permit  this 
procedure.  Again,  to  compare  sample  and  standard  on  the  same 
animal  by  means  of  consecutive  injections  is  not  practicable  be- 
cause the  elimination  of  digitalis  is  so  slow  that  the  animal  never 
fully  returns  to  its  original  condition  to  permit  comparing  two  in- 
jections even  of  the  same  sample. 
For  these  reasons,  attempts  to  use  this  as  the  end-point  have  been 
abandoned  and  the  death  of  the  animal  or  in  the  case  of  frogs,  stop- 
page of  the  heart  in  systole  selected. 
Not  only,  therefore,  can  the  action  of  drugs  on  the  human  be 
deduced  from  that  on  the  lower  animals,  but  also  the  degree  of  the 
action  can  be  measured  and  the  exact  manner  of  action  explained. 
Lloyd's  statement  in  reference  to  killing  a  turkey  and  curing  a 
man  may  have  been  true  in  those  days  when  the  science  was  un- 
developed. At  present,  however,  the  fact  of  killing  is  of  little 
importance;  the  vital  points  in  the  experiment  are  the  amount  that 
killed,  the  action  of  smaller  doses  and  the  possibility  of  applying  the 
substance  clinically  to  relieve  a  pathological  condition. 
There  is  no  known  medicinal  substance  which,  taken  in  excessive 
quantities,  does  not  induce  toxic  symptoms  and  in  most  cases,  death. 
It  is  not  valuable  because  of  this  but  in  spite  of  it. 
The  question  raised  by  Rusby  is  entirely  logical  and  apropos. 
If  a  substance  is  standardized  solely  on  the  basis  of  its  M.  L.  D. 
with  no  regard  to  the  characteristic  effect  which  that  substance 
may  be  expected  to  produce,  common  sense  suggests  that  such  a 
test  is  inadequate  and  should  be  used  only  if  no  verifying  effect  is 
available.  We  will  attempt  to  apply  this  test  of  adaptability  to 
the  various  methods  for  the  assay  of  the  digitalis  series,  disregarding 
the  question  of  accuracy  and  keep  in  mind  only  whether  the  test 
shows  digitalis  glucosides  and  no  other  poison  to  be  present. 
The  Gold  Fish  Method,  while  little  known  or  practiced,  is  a 
proposed  means  of  assaying  digitalis  by  its  toxic  action  on  gold 
fish,  the  drug  being  mixed  in  various  dilutions  with  the  water  in 
which  they  swim.  The  end-point  is  the  dilution  which  kills  after 
a  certain  period  of  contact  and  with  due  regard  to  certain  factors, 
such  as  the  temperature  of  the  medium.  It  has  the  advantage  of 
being  one  of  the  simplest  and  probably  cheapest  of  all.  But  for 
the  fact  that  there  is  nothing  typical  of  digitalis  in  the  action  of  the 
drug  on  the  test  animal  there  is  no  reason  why  it  should  not  have  a 
