i8o 
The  US.P.  Melting  Points. 
f  Am.  Jour.  Pharm. 
t      April,  1910. 
cause  or  causes  of  such  conflicting  values,  or  of  reasonable  protests, 
should  be  determined  and,  if  possible,  eliminated — if  not  completely, 
at  least  to  as  great  a  degree  as  is  practicable. 
The  question  naturally  arises — "  What  are  the  causes  of  this 
divergence  and  what  is  the  remedy?  "  I  would  summarize  the  main 
causes,  though  perhaps  imperfectly  and  incompletely,  as  follows : 
1.  The  great  variety  of  methods  used  in  melting  point  deter- 
minations. 
2.  Varied  individual  manipulation,  including  the  so-called  "  per- 
sonal factor,"  and  especially  the  rate  of  heating. 
3.  Differences  in  the  physical  condition  of  the  compounds. 
4.  The  use  of  thermometers  differing  widely  in  their  construc- 
tion or  range,  or  both. 
5.  The  application  or  omission  of  emergent-stem  correction  and 
the  manner  of  making  it. 
6.  Widely  varying  interpretations  of  just  what  the  melting- 
point  is  (which  might  be  considered  to  include  the  apparent  use  of 
decomposition  point  as  equivalent  to  melting  point). 
The  remedy  may  be  indicated,  to  a  greater  or  less  degree,  by  a 
brief  discussion  under  each  of  these  different  headings. 
A  description  and  detailed  discussion  of  all  the  methods  for 
melting  point  determinations  that  I  have  so  far  found  described  in 
the  literature  would  doubtless  be  interesting  and  instructive  but 
would,  I  fear,  unduly  tax  not  only  your  patience  but  also  your 
endurance,  and  mine.  Some  of  them — designed  to  eliminate  certain 
specific  difficulties  in  obtaining  accurate  results — are  ingenious,  more 
or  less  complicated,  devices  which  impress  me  as  being  rather  fan- 
tastic, and  impossible  of  general  application. 
The  methods  prescribed  by  some  of  the  pharmacopoeias  are,  in 
the  main,  simple  and  practical  but  have  not  been  sufficiently  devel- 
oped, it  seems  to  me,  to  insure  the  degree  of  uniformity  in  results 
of  which  they  may  be  capable. 
That  the  use  of  different  methods  constitutes  a  real  cause  of 
divergence  in  results  is,  I  believe,  pretty  generally  recognized  but 
may  perhaps  be  more  emphatically  indicated  here  by  citing  some 
very  good  work  in  demonstration  of  this  fact. 
In  1889  Landolt  published  the  results  of  a  very  careful  investi- 
gation to  test  the  comparative  accuracy  of  several  methods,  includ- 
ing the  determination  of  the  melting  points  and  of  the  freezing 
points  of  compounds  with  thermometers  dipping  into  the  substance  ; 
