4  Pharmacological  Assay  of  Drugs.     { A "January f  1902 m" 
is  practicable,  I  think  it  will  be  used  in  preference  to  the  pharma- 
cological method,  not  I  believe  because  it  is  more  reliable  or  in  most 
cases  more  accurate,  but  as  on  the  whole  giving  more  assurance  to 
the  operator.  There  is  a  sense  of  finality  in  weighing  an  alkaloid 
obtained  after  long  and  complicated  methods  of  isolation  which 
may  be  absent  from  the  breast  of  the  experimental  pharmacologist 
on  viewing  the  systolic  standstill  of  the  frog's  heart.  The  newer 
method  is  not  a  rival  of  the  older  but  a  substitute  for  it  in  cases 
where  it  cannot  be  applied.  From  a  theoretical  standpoint  of 
course  the  symptoms  elicited  in  animals  is  merely  an  "  indicator 
reaction  "  of  an  exceedingly  complicated  kind  but  based  on  chemi- 
cal processes  just  as  the  indicator  reactions  in  titration.  For 
example,  the  arrest  of  the  frog's  heart  under  digitalis  is  evidence  of 
the  completion  of  a  chemical  reaction  between  the  heart  muscle 
and  the  glucosides  just  as  the  alteration  in  the  color  of  litmus  is  the 
indication  of  a  chemical  change  in  titrating  an  acid.  And  this 
arrest  does  not  indicate  the  presence  of  any  one  glucoside  but  the 
total  strength  of  the  glucosides  with  this  cardiac  action,  just  as  the 
litmus  test  indicates  the  total  acid  in  a  mixture  of  several  acids. 
The  questions  which  arise  at  once  are  how  sensitive  is  the  test 
and  how  constant  does  it  remain  ?  If  the  frog's  heart  reacts  only 
to  large  quantities  of  digitalis  glucosides,  the  test  is  obviously  of 
little  value  ;  and  on  the  other  hand,  if  the  frog  requires  twice  as 
much  digitalis  at  one  time  as  at  another,  the  method  is  equally 
valueless.  These  questions  may  best  be  answered  by  the  narration 
of  a  test  which  I  carried  out  on  a  series  of  frogs  with  a  fluid  extract 
of  digitalis.  I  diluted  it  fifty  times  and  selected  eight  frogs,  four  of 
which  received  I  c.c.  of  the  diluted  extract,  the  other  four  receiving 
0  5  c.c.  The  first  four  all  gave  the  characteristic  reaction,  while 
the  second  set  failed  to  do  so,  although  they  showed  symptoms. 
The  quantity  required  for  the  test  reaction  thus  lay  between  1  c.c. 
and  0-5  c.c.  of  the  diluted  extract,  that  is  between  0-02  and  0  01  of 
the  original.  Four  fresh  frogs  then  received  075  c.c.  of  the  diluted 
fluid  and  three  of  them  reacted,  one  of  them  failing  to  do  so  though 
severely  poisoned.  Another  set  which  received  o-6  c.c.  all  re- 
covered. Finally  two  frogs  poisoned  with  07  c.c.  reacted  typically. 
The  limits  were  thus  between  07  and  0-6  c.c.  of  the  diluted  extract, 
i.  e.y  0-014  and  0-012  c.c.  of  the  original,  the  test  indicating  a  differ 
ence  of  0-002  c.c.    The  dose  of"  this  preparation  ordinarily  employed 
