i86 
The  US. P.  Melting  Points. 
Am.  Jour.  Pharii 
April.  1910. 
effect  that  the  presence  of  moisture — also,  of  course,  impurity — 
may  produce  upon  the  melting  point  of  a  compound.  For  example, 
in  this  connection,  their  results  with  acetanilide,  phenacetin,  and 
antipyrine  show  a  variation  in  melting  point  of  about  i°  between 
the  commercial  and  commercial-dried  forms  and  a  variation  of  from 
0.6°,  in  the  case  of  phenacetin,  to  nearly  30,  in  the  case  of  anti- 
pyrine, between  the  dried  and  the  purified  forms.  The  question  of 
moisture  as  a  cause  of  variation  in  the  melting  points  of  pharma- 
ceutical compounds  would  seem  to  be  easily  disposed  of  by  requir- 
ing a  definite  period  of  adequate  desiccation  for  the  finely  powdered 
substances  before  the  melting  point  determination  is  made. 
The  question  of  impurity,  however,  it  seems  to  me,  is  a  much 
more  difficult  one.  To  attempt  an  exact  standardization  of  melting 
points  for  a  class  of  compounds  in  which  a  certain  percentage  of 
impurity  is  permissible  opens  a  wide  field  of  investigation,  that  so 
far  as  I  can  find,  has  barely  been  touched  upon,  and  constitutes,  at 
least  in  theory,  a  very  complicated  problem,  the  discussion  of  which 
is  far  beyond  the  scope  of  this  paper.  Any  possible  difficulty  in 
dealing  with  this  factor,  however,  does  not,  in  my  opinion,  justify 
neglect  and  tolerance  of  other  causes  of  divergence  which  can  be 
readily  eliminated. 
Passing  rapidly  over  the  topics  remaining  for  discussion : 
The  fact  that  the  use  of  thermometers  of  different  construction 
induces  more  or  less  variation  in  results  suggests  the  desirability  of 
adopting  an  official  thermometer — or  set  of  thermometers — as  a 
part  of  the  official  method,  and  requiring  that  they  be  standardized. 
True,  such  a  step  introduces  the  objectionable  element  of  expense, 
but  surely  not  in  a  prohibitive  degree,  even  for  a  very  modestly 
equipped  laboratory. 
That  the  application  or  omission  of  emergent-stem  correction  is 
a  real  and  serious  cause  of  divergence  is  strikingly  illustrated  by  the 
fact  that  if  a  thermometer  of  average  construction — used  in  connec- 
tion with  the  apparatus  we  have  recommended — registers  200 0  as 
the  melting  point  of  a  compound,  the  correction,  in  most  cases,  will 
amount  to  30  or  40.  Obviously  uniformity  of  practice  in  this  respect 
should  be  required,  and  in  the  cause  of  accuracy  the  correction 
should  be  applied.  The  manner  of  making  the  correction  may  also 
cause  variation  and  should  therefore  be  clearly  defined — or  better 
still,  if  official  thermometers  were  adopted,  official  corrections  could 
be  made,  plotted  in  a  curve  on  co-ordinate  paper  and  published, 
