508  Review  of  Chemistry  of  the  U.S.P.  {^oVimber.S' 
table  of  the  U.S.P.,  VIII.  In  other  words,  we  have  here  the  rich 
idea  of  some  one  to  apply  a  correction  to  a  method  of  analysis  and 
still  remain  incorrect  to  the  extent  of  from  1-5  to  3  per  cent,  and 
even  more  in  summer.  If  it  take  five  years  to  make  a  fake  correc- 
tion, how  long  will  it  take  to  complete  it  ?  Have  your  chemical 
subdivision,  and  do  not  let  non-workers  pass  on  technical  questions 
of  which  they  know  little  and  understand  less. 
Let  us  consider  that  250-metre  absurdity.  You  will  have  to  go 
rather  high  to  touch  this  low  mark  in  the  city  of  New  York. 
Observations  as  recorded  by  Professor  Cole  show  that  the  mean 
barometric  readings  for  a  number  of  years  in  New  York  are  29-854 
inches,  practically  760  mm.,  equal  no  correction  U.S. P.,  VIII.  He 
has  found  that  the  hourly  variations  in  pressure,  if  plotted  for  the 
year,  follow  the  curve  of  probability,  proving  this  barometric  change 
follows  the  law  of  probability.  In  other  words,  he  is  able  to  calcu- 
late the  number  of  times  a  certain  reading  will  be  made  during  the 
year.  To  avoid  the  negative  sign,  the  deviations  from  28  00  inches 
are  given,  I  take  only  the  i^-inch  differences  to  avoid  too  long  a 
table.  This  is  taken  from  the  Meteorologische  Zeitschrift,  1894,  in 
our  library : 
Deviation  .....  0-95  1*25  1*50  175  2*00  2*25  2*50 
Times  read    .  .  .  .   o  5  31  79  64  18  2 
This  table  then  demands  that  a  pressure  of  2875  occur  79  times 
during  a  year.  Let  us  take  the  readings  of  a  single  year  as  ob- 
served, 1890: 
Reading  095       1*25        1*50        175        2*00        2*25  2*50 
Observed  1  6  25  79  62  14  2 
From  the  above,  you  will  see  that  760  mm.  is  not  the  prevalent 
thing  even  at  sea  level ;  in  the  West  the  changes  are  greater  and 
more  sudden.  This  variation  from  28-95  to  30  50  inches  in  New 
York,  means  on  30  inches  a  variation  of  5-1  per  cent.  Shall  a  small 
5  per  cent,  be  overlooked  in  our  assays  ?  But  the  overlooked  or 
unheard-ol  vapor  tension  makes  another  inch  at  25 °,  and  this  must 
always  be  subtracted  from  the  barometric  reading,  one  on  thirty  is 
another  little  2>%  Per  cent.  Fine  8  per  cent,  was  what  the  U.S.P.  of 
1890  allowed  us  to  run  on,  and  the  grand  U.S.P.,  VIII,  allows  New. 
Yorkers  the  same.  How  considerate  to  our  druggist  in  the  metrop- 
olis.   Did  he  have  a  "  pull "  ? 
