230  Gelsemic  Acid.  {^^mF* 
of  analyses  being  made  with  no  better  success  than  before.  A 
description  of  this  latter  method  with  apparatus  is  given  here,  since 
it  has  answered  admirably  in  the  analyses  of  various  derivatives  of 
gelsemic  acid. 
In  the  combination  flask  (Fig.  i)  from  IO  to  20  grammes  of  chro- 
mic anhydride  are  introduced,  followed  by  the  gelsemic  acid  which 
has  been  accurately  v/eighed  off  in  a  small  thin  glass  tube,  this  is 
placed  in  a  nearly  upright  position  in  the  flask,  in  order  to  avoid 
contact  with  the  Cr03  before  the  proper  time.  After  securing  all 
the  joints  of  the  apparatus,  a  slow  current  of  pure  oxygen  gas  is 
passed  through  the  entire  apparatus  until  practically  all  of  the  air 
has  been  removed,  after  which  the  current  is  regulated  to  about  20 
Fig.  t. 
bubbles  a  minute,  which  is  continued  throughout  the  entire  combus- 
tion, unless  the  reaction  becomes  violent,  when  the  current  should 
be  temporarily  closed.  By  slightly  shaking  the  flask  the  gelsemic 
acid  is  caused  to  spill  out,  and  is  distributed  through  the  chromic 
oxide,  then  the  concentrated  sulphuric  acid  which  is  contained  in 
the  separation  funnel  is  allowed  to  trickle  over  the  mixture  very 
slowly,  regulating  the  flow  according  to  the  energy  of  the  reaction. 
Finally,  when  the  reaction  is  over,  sufficient  acid  is  added  to  make 
a  bulk  of  about  20  to  30  c.c.  The  flask  and  contents  are  then  cau- 
tiously heated,  increasing  gradually  till  just  short  of  boiling  tem- 
perature, which  is  kept  up  for  a  period  of  fifteen  to  twenty  minutes. 
The  gases  given  off  pass  up  through  a  well-cooled  condenser  into  a 
tube  which  is  filled  with  spun  glass,  well  covered  with  lead  peroxide, 
