18 
True  Zero  of  Thermometers. 
f  Am.  Jotre.  Pharm. 
\     Jan.  1, 1873. 
the  indication  of  the  0°  when  placed  in  melting  ice  or  snow.  Accord- 
ing to  the  observations  on  the  supersaturation  of  water  with  cold  (see 
Chemical  News,  vol.  xxvi,  p.  107)  thermometers  are  much  less  vari- 
able than  is  generally  supposed,  and  the  cause  of  the  differences  which 
are  observed  is  probably  due  to  an  error  made  in  the  determination  of 
the  0°.  It  may  be  readily  conceived  that  unless  special  precautions 
are  taken  at  the  time  of  the  gradation  of  the  thermometer,  the  water 
in  which  it  is  plunged,  and  which  is  supposed  to  be  precisely  at  the 
temperature  of  melting  ice,  may  in  reality  be  slightly  above  that  tem- 
perature ;  this  is  the  case  if  the  walls  of  the  vessel  containing  the  wa- 
ter and  ice  admit  more  heat  to  the  water  than  the  melting  ice  can 
overpower:  this  is  natural;  ice  does  not  melt  instantaneously,  but 
only  in  the  ratio  of  its  surface,  and  in  proportion  to  the  difference  of 
the  temperature  of  the  water  in  which  it  floats,  and  its  own  tempera- 
ture ;  and  it  is  consequently  quite  possible  that  the  water  which  con- 
tains the  ice  is  not  at  a  temperature  of  0°.  The  colder  the  water  the 
more  slowly  will  equilibrium  of  temperature  be  established  between 
the  two  bodies,  and  in  the  same  ratio  will  the  chances  of  error  be 
greater.  The  error  of  indications  of  the  thermometers  brought  on  by 
time  either  depends  upon  a  modification  in  the  glass,  as  usually  admit- 
ted, or  it  is  due  to  the  result  of  an  erroneous  estimation  of  the  0°.  In 
the  first  case  the  alteration  would  rather  tend  to  plus  in  one  case  and 
minus  in  the  other,  and  there  is  no  plausible  reason  why  it  should  be 
otherwise;  in  the  second  case  (erroneous  estimation  of  the  0°)  the 
error  should  be  always  plus,  because  the  water  must  be  above  0U.  My 
experiments  have  confirmed  these  views.  I  have  taken  seven  ther- 
mometers with  the  gradations  engraved  on  the  stem  and  made  by  one 
of  the  best  makers;  only  one  of  these  instruments  has  been  found  to 
indicate  0°  correctly,  all  the  others  indicated  a  difference — 
2  indicated  +  0-1 
1  "       +  0-2 
2  "  +  0-3 
1       "       +  04 
Not  one  of  these  thermometers  indicated  below  0°. 
The  determination  of  the  0°  by  placing  the  thermometer  in  melting 
ice  is  therefore  not  an  absolutely  certain  method  of  operating.  In 
order  to  find  the  true  0°  another  plan  must  be  followed,  which  is  that 
found  and  described  by  me,  and  called  terminus  of  congelation.  The 
operation  is  carried  on  as  follows  : — A  glass  vessel  is  placed  in  a  re- 
