Am\Jdir'i9ihoarm'}  Thc  US.P.  Melting  Points. 
also  various  modifications  "of  methods  involving  the  use  of  capillary 
tubes  (including  Piccard's) — using  both  liquid  and  air  baths;  and 
certain  electrolytic  methods.  These  were  applied  in  determining  the 
melting  points  of  three  substances  (naphthalene,  mannite,  and 
anthracene),  melting  at  about  8o°  C,  165  °  C,  and  2000  C.  respec- 
tively. The  results  obtained  for  eaeh  of  these  compounds  under  the 
same  conditions  by  different  methods  were  variable. 
In  1890  Reissert  followed  with  the  publication  of  somewhat 
similar  work.  He  used  only  three  methods,  all  of  which  were 
included  in  Landolt's  investigation,  but  extended  his  experiments 
to  a  much  larger  number  of  compounds  (24).  Here  again  we  find 
differences  in  the  values  obtained  for  the  same  compound  by  differ- 
ent methods — the  divergence  ranging  from  a  few  tenths  of  a 
degree  at  low  temperatures  to  several  degrees  at  high  temperatures. 
Perhaps  the  most  comprehensive  comparative  study  of  different 
methods  that  can  be  found  in  the  literature  is  that  of  Tyrer  and 
Levy,  published  in  the  Year  Book  of  Pharmacy,  1899  and  1900.  In 
the  course  of  their  investigation  nine  different  methods  were  used, 
including  that  described  by  the  British  Pharmacopoeia,  Graebe's, 
Landolt's,  Piccard's,  Loewe's,  Mill's,  Kuhara  and  Chikashige's, 
and  Levy's  acoustical  method.  Twelve  compounds  were  treated, 
ranging  in  melting  point  from  about  40 0  C.  to  about  200 0  C. 
Besides  the  divergence  due  to  the  use  of  different  methods  they 
studied  the  effect  upon  the  melting  point  due  to  varying  physical 
conditions  of  the  compound,  to  the  extent  that  they  determined 
the  melting  point  of  the  commercial  product,  the  same  dried,  and 
the  same  purified  until  there  was  no  further  rise  in  melting  point. 
The  amount  of  divergence  resulting  from  the  use  of  different 
methods  varied  not  only  with  rising  temperature  but  also  between 
different  compounds  melting  at  about  the  same  temperature,  and 
ranged  from  about  0.5 0  at  low  temperature  to  about  30  or  40  at 
high.  With  only  three  compounds,  however,  was  the  comparison 
extended  to  all  nine  methods.  These  compounds  were  spermaceti, 
melting  at  about  43  °,  betanaphthol,  at  about  1220,  and  picrotoxin, 
at  about  2000.  The  range  of  divergence  in  this  case  extended  from 
about  2°  for  the  first  two  compounds  to  over  6°  for  picrotoxin.  The 
increase  in  range  of  divergence  with  increase  in  the  number  of 
methods  tested  upon  the  same  compound  under  the  same  conditions 
is  very  striking  and  convincing. 
Lndoubtedlv  then  the  use  of  different  methods  is  a  real  and 
