ADecimSeri896m-}      Alcohol  in  Volumetric  Analysis.  667 
attention  to  this  apparatus,  having  used  such  a  one  for  many  years, 
with  most  satisfactory  results.  It  consists  of  a  box  made  of  good, 
sound  wood,  free  from  loose  knots  or  cracks,  the  joints  being  well 
closed.  It  should  be  made  of  planed  boards,  so  that  paper  pasted 
on  the  inside  will  adhere  closely ;  the  lid  should  fit  tightly  and  the 
edges  of  the  box  should  be  lipped  with  soft  leather  or  sheet  rubber. 
A  tray  must  be  provided  in  which  freshly  burnt  lime  is  to  be  placed, 
and  upon  the  sides  of  box  cleats  are  fastened  at  convenient  distances 
to  support  wire  frames,  upon  which  the  substances  to  be  dried  are 
placed.  The  case  should  then  be  closed  tightly  by  means  of  hooks 
of  such  a  character  that  the  closer  they  are  pressed  to  their  support- 
ing pins  the  tighter  they  draw  the  lid  down.  Lozenges  of  a  hygro- 
scopic character  can  be  dried  in  two  days'  time,  so  that  they  become 
brittle.  Tough  roots,  like  gentian  and  althaea,  after  having  been  kept 
for  a  time  in  the  case,  can  be  readily  ground.  Tartaric  acid,  bicar- 
bonate of  potassium  and  other  moist  chemicals,  if  kept  for  a  time 
in  the  drying  box,  will  be  found  in  excellent  condition  for  use. 
Leaves  containing  volatile  ingredients,  which  are  liable  to  be  dissi- 
pated by  heat,  will  soon  be  fit  for  powdering  and  found  to  retain  their 
natural  components  unaltered. 
IS  ALCOHOL  A  SOURCE  OF  ERROR  IN  VOLUMETRIC 
ANALYSIS? 
BY  IyYMAN  F.  Kkbi^er. 
About  a  year  and  a  half  ago,  while  making  a  careful  study  of  the 
relative  sensitiveness  of  the  various  indicators,  the  writer  and  Mr. 
C.  H.  La  Wall  observed  what  appeared  to  be  an  affirmative  answer  to 
the  above  question.  A  few  experiments  were  made  by  the  writer, 
with  various  samples  of  commercial  alcohol,  when  it  was  found  that 
they  all  exercised  a  disturbing  influence.  It  was  found,  however, 
that  alcoholic  solutions  of  the  same  strengths  yielded  fairly  con- 
cordant results,  i.  e.,  the  disturbing  factor  was  constant.  Moreover, 
a  15  per  cent,  solution  did  not  appear  to  affect  the  results.  Should 
future  investigations  prove  this  observation  to  be  correct,  then  it 
would  be  necessary  to  make  due  allowance  for  all  volumetric  work 
in  which  alcohol  forms  an  essential  solvent.  This  would  not  only 
be  the  case  with  strong  alcohol,  but  for  various  strengths  also.  It 
might  be  necessary  to  establish  a  co-efficient  of  disturbance.  A 
