Am.  jour.  Pharm.  ) 
Febv  1878.  j 
The  Saponin  of '  Sar sap ar ilia. 
83 
of  water,  or  evaporation  of  the  alcohol  and  chloroform,  white  flocks 
of  parigenin  are  formed  whilst  sugar  remains  in  the  solution.  As  in 
similar  cases  the  fluorescence  of  parillin  is  very  persistent.  An 
unweighable  quantity  heated  with  a  few  drops  of  strong  sulphuric  acid 
in  a  water-bath  gives  a  liquid  that  can  be  diluted  with  100  cc.  of  acid 
without  losing  its  fluorescence,  but  dilution  with  water  causes  its  imme- 
diate disappearance.  After  saturation  with  ammonia  the  liquid  does 
not  again  show  the  green  shade.  In  this  behavior  and  the  color  parillin 
gives  with  a  little  sulphuric  acid  in  the  cold  lie  the  best  means  at 
present  known  for  its  detection.  It  is  noticeable  that  the  "  saponin  " 
of  digitalis,  to  be  presently  mentioned,  Schmiedeberg's  digitonin,  also 
gives  this  fluorescence,  but  not  cyclamin.  Three  analyses  of  parillin 
(smjlacin)  given  in  Gmelin  on  the  authority  of  Henry,  Peterson  and 
Poggiale,  agree  fairly  well  with  the  figures  obtained  by  Klunge  in  two 
analyses.  But  Professor  Fllickiger  believes  that  these  specimens  were 
contaminated  with  parigenin.  Parillin  dissolved  in  warm  water,  which 
does  not  take  up  parigenin,  filtered,  and  reprecipitated  by  alcohol,  gave 
between  2  and  3  per  cent,  less  corbon,  or  as  a  mean  of  three  analyses, 
Q— 60*4  ;  H=9-  Three  other  analyses  of  another  sample  gave  the 
following  figures,  showing  still  less  carbon  : 
C  ...    57*66       56-80  56*4 
H  .  .    8-27  8-3 
These  figures  appear  to  show  a  remarkable  relation  between  parillin 
and  the  saponin  prepared  by  Rochleder,  Schwarz  and  von  Payr  from 
the  u  soap-root "  erroneously  attributed  to  Gypsophila  Struthium,  which 
had  the  formula  C64H106O36.  If  this  formula  be  written  C32H53018,  the 
next  lower  in  a  homologous  series  of  "  saponins"  would  have  the  for- 
mula C31H51018.  Possibly  this  is  the  place  of  the  "  saponin  "  found 
by  Schmiedeberg  in  commercial  digitalin,  and  named  "  digitonin  "  ;  he, 
however,  attributed  to  it  the  formula  C31H52Oir  Should  there  really 
be  a  homologous  series  of  "  saponins,"  the  eighth  step  upwards  from 
Rochleder's  saponin  would  be  the  compound  C32H53018-|-8CH2— 
C40H69O18.  This  would  require  57*3  per  cent,  of  carbon,  and  8*2  per 
cent,  of  hydrogen,  figures  not  irreconcilable  with  those  obtained  in  the 
last  three  analyses  of  parillin,  whilst  the  mean  of  the  previous  three 
analyses  would  agree  with  the  formula  of  a  saponin  C32H53018-f-i6CH2 
=C48H85018,  which  would  require  60*7  per  cent,  of  carbon  and  9  per 
cent,  of  hydrogen.  Subsequently,  however,  Rochleder  has  published 
the  formula  C32H54018  for  his  saponin,  which  agrees  better  with  the 
results  of  its  decomposition,  and  also  pretty  closely  with  the  saponin 
prepared  by  Christophsohn  from  Levant  soap  root,  quillaia  bark,  Sapo- 
naria  officinalis  and  Agrostemma  seeds,  which  he  believes  to  be  identical 
as  obtained  from  all  four  sources.  As,  however,  this  would  only 
slightly  alter  the  hydrogen,  it  is  not  inconsistent  with  the  homologous 
nature  of  the  "saponins."  From  these  and  other  considerations  it 
appears  probable  that  there  exists  a  series  of  saponins  with  the  general 
formula  Cn  H2n  -10O18. 
