520  The  Netherlands  Pharmacopoeia.  {A^VeSVPi9™* 
furthermore  directed  that  if  the  pharmacist  has  any  question  as  to 
the  authenticity  of  the  sample  examined,  he  shall  compare  it  with 
the  powder  of  a  drug,  the  identity  of  which  is  known  or  has  been 
established. 
It  should  be  emphasized  that  this  is  a  very  important  rule  to 
observe.  Considerable  stress  is  laid  upon  the  size  of  the  tissues  and 
cell-contents,  and  the  manner  of  stating  the  measurements  is  quite 
a  novel  one. 
As  every  one  knows  there  are  certain  discrepancies  in  the  litera- 
ture with  regard  to  the  length  of  fibres,  size  of  starch  grains,  number 
of  carpels,  etc.  In  order  to  prepare  accurate  statements  based  on 
statistics  which  would  allow  for  uncertain  variations  the  latter  were 
arranged  in  such  a  manner  that  the  curve  of  probability  would  indi- 
cate the  different  possible  errors.  This  is  arrived  at  as  follows :  The 
number  indicating  the  minimum  size  is  subtracted  from  that  of  the 
maximum  size,  and  the  half  of  the  remainder,  known  as  the  quarter 
(quartilis),  is  taken  to  represent  the  probable  mathematical  error. 
The  average  (or  medium)  size  is  obtained  either  by  subtracting  the 
quarter  from  the  maximum  or  adding  it  to  the  minimum.  The 
average  size  or  the  average  number  is  indicated  by  M  and  the 
probable  variation  above  or  below  this  is  indicated  by  Q,  and  thus 
we  find  under  the  description  of  the  fibres  of  powdered  cinnamon 
that  M  is  equivalent  to  489.5  p  and  Q  equals  55.7  /x,  and  thus  one 
would  expect  to  find  a  variation  ranging  from  433.8  to  545.2  fi. 
In  those  cases  where  the  curve  of  probability  of  error  would 
become  asymmetric,  as  in  the  measurement  of  potato  or  maranta 
starch  grains,  the  size  is  indicated  by  an  arithmetical  ratio  repre- 
sented thus  (x).  Thus  under  amylum  solani  we  find  that  the  error 
of  asymmetry  in  length  of  the  grain  is  given  as  23  /*,  and  is  expressed 
as  x  =  23  /x.  This  means  that  after  making  the  measurements  of  a 
large  number  of  grains  in  a  certain  field,  the  sum  of  the  sizes  of  the 
grains  divided  by  the  total  number  examined  gives  x,  or  the  arith- 
metical ratio. 
Again,  in  the  study  of  starch  grains,  the  position  of  the  point  of 
origin  of  growth  or  hilum  is  indicated  by  a  fraction,  the  numerator 
and  denominator  indicating  respectively  the  distance  of  the  nucleus 
from  the  two  ends  or  the  two  extremes  of  the  longer  diameter. 
Thus,  1/1  would  indicate  that  the  nucleus  is  central  as  in  wheat 
starch;  %,  as  in  the  case  of  potato  starch  grains  where  the  nucleus 
is  %  the  diameter  from  the  narrow  end.    In  the  case  of  maranta 
