no  Standardization  of  Digitalis  Bodies.     \km^£]  S™' 
tance,  especially  in  regard  to  the  relation  of  the  time  of  feeding  or 
drinking  to  that  of  the  testing.  This  can  be  made  uniform  and 
negligible  by  feeding  at  the  same  hour  each  day,  the  best  time  being 
late  in  the  afternoon,  thus  leaving  time  for  the  complete  digestion 
of  food  to  have  taken  place  before  the  animals  are  to  be  used  for 
the  tests  on  the  following  day.  Food  has  also  been  shown  to 
influence  the  reactions  of  frogs  and  guineapigs.  Certainly  the  time 
of  feeding  will  have  as  much  effect  upon  the  pigs  as  on  the  cats, 
for  the  influence  is  merely  one  of  unknown  alteration  in  the  weight 
of  the  animals. 
Sex  influences  the  reactions  of  frogs,  as  has  been  shown  (page 
107),  but  per  se  it  seems  to  have  little  or  no  influence  on  mammals. 
The  occurrence  of  pregnancy  or  lactation  on  the  part  of  cats  does 
have  a  profound,  and  quite  uncertain,  effect  upon  the  susceptibility 
of  the  animals,  and  we  have  seen  variations  from  this  cause  of 
rather  more  than  50  per  cent,  in  rare  instances.  Eckler  reports  the 
finding  of  one  animal  which  took  five  times  as  much  as  the  average, 
but  in  many  hundreds  of  tests  made  in  this  laboratory  we  have 
never  seen  a  tolerance  even  one-fourth  so  great  as  this.  In  view  of 
the  variability  in  pregnant  and  lactating  cats  they  must,  of  course, 
not  be  used  in  biological  assays. 
We  found  that  in  one  or  two  instances  we  could  not  secure  the 
usual  approximate  uniformity  in  reaction  between  the  individual 
animals  of  a  series,  even  when  using  ouabain.  All  other  possible 
factors  being  eliminated,  it  was  suggested  by  Dr.  Hatcher  that  the 
irregularity  in  response  might  be  due  to  a  seasonal  variation  in  the 
susceptibility  of  cats.  The  animals  seemed  to  be  far  less  uniform 
in  reaction  in  the  summer  months  than  at  other  times  of  year. 
Analysis  of  the  records  showed  this  to  be  the  correct  explanation, 
as  is  well  illustrated  in  the  table.  Ouabain  series  A  was  made  in 
the  months  .of  July  and  August  and  shows  considerable  lack  of 
uniformity  between  the  reactions  of  the  several  animals.  Series  B, 
on  the  other  hand,  was  made  in  March  and,  with  one  exception, 
the  variation  between  the  individual  animals  is  very  slight. 
This  seasonal  variation  in  cats  is  slight  as  compared  with  that 
shown  to  exist  in  guineapigs  by  Haskell  (page  105).  Even 
with  the  tendency  to  greater  variation  in  the  susceptibility  of  dif- 
ferent cats  during  the  summer  months,  it  is  possible  to  obtain 
results  which  very  closely  approximate  the  standard  for  a  uniform 
preparation  such  as  ouabain  by  the  running  of  a  small  series  of 
tests.    This  is  shown  by  the  series  Ouabain  A. 
