466 
kqchleder's  proximate  analysis. 
binations,  and  of  whose  existence  we  know  nothing.  With  vegetable  sub- 
stances also  which  contain  only  known  bodies,  a  method  for  their  quanti- 
tative determination  cannot  be  supplied,  because  the  bodies  most  known 
are  still  too  little  investigated  with  exactness  to  afford  grounds  for  the 
construction  of  a  method  of  separation  from  a  knowledge  of  their  proper- 
ties, and  the  properties  of  their  compounds.  Another  circumstance  ren- 
ders a  quantitative  estimation  almost  impossible,  even  when  we  know  the 
means  of  separating  the  constituents  from  one  another  :  namely,  a  com- 
plete exhaustion  of  a  vegetable  substance  under  examination  is  almost  im- 
practicable, because  it  is  not  possible  to  comminute  a  substance  in  such  a 
way  that  each  cell  is  broken  and  its  contents  exposed  to  the  action  of  a 
fluid.  There  also  always  remains  behind  a  quantity — sometimes  greater, 
sometimes  smaller — of  the  constituent  to  be  estimated  in  the  material 
under  examination.  It  may  be  easily  determined  by  an  experiment  how 
much  quinine  can  be  extracted  from  a  sample  of  cinchona  bark,  while  the 
difficulties  of  the  experiment  would  be  considerably  increased  to  determine 
how  much  quinine  is  contained  in  such  sample  of  bark,  as  a  little  quinine 
always  remains  in  the  bark.  I  believe  that  it  is  quite  useless  to  give  the 
methods  here  which  have  been  devised  to  determine  quantitatively  one  or 
the  other  constituent  in  a  certain  vegetable  substance  ;  for  example,  the 
starch  in  seeds  or  bulbs,  the  sugar  of  beetroot,  the  quinine  in  cinchona 
bark,  &c.  To  conduct  a  quantitative  analysis,  a  complete  qualitative  analy- 
sis must  first  be  undertaken,  and  the  individual  constituents  must  be  ac- 
curately studied,  and  not  only  must  their  relations  to  solubility  and  their 
combinations  be  known,  but  their  decomposition  products,  the  properties 
of  the  same,  and  the  quantity  in  which  they  appear,  &c,  must  have  been 
correctly  studied.  In  a  particular  case,  a  method  for  the  quantitative 
analysis  of  an  individual  part  of  a  certain  plant  may  be  devised  when  the 
troublesome  and  tedious  labor  above  mentioned  has  been  performed  in  that 
particular  case.  But  it  is  not  impossible  to  supply  a  general  method  for 
this  purpose,  because  a  single  body  previously  unknown,  when  present 
with  other  well-known  constituents,  must  always  cause  an  alteration  in 
the  entire  order  of  the  analysis.  The  majority  of  quantitative  analyses, 
of  which  we  have  no  deficiency,  are  regarded  as  useless.  They  can  only 
claim  to  be  considered  as  qualitative  analyses.  Independently  of  pharma. 
ceutical  or  technical  bearings,  which  in  particular  instances  may  render 
desirable  an  approximately  correct  quantitative  determination  of  one  or 
the  other  constituent  of  a  certain  material,  an  exact  quantitative  analysis 
can  only  have  for  its  object  the  construction  of  a  true  representation  of  the 
normal  constitution  of  the  parts  of  a  plant.  But  this  object  has  never 
been  accomplished  by  an  analysis,  however  correct.  We  are  far  from 
being  in  a  position,  from  a  dismembered  analysis  of  an  ash,  to  draw  a 
valid  conclusion  on  the  normal  soil  constituents  of  a  plant,  and  quite  as 
far  from  being  able,  from  the  individual  analyses  which  have  been  per- 
