rochleder's  proximate  analysis. 
467 
formed  with  especial  reference  to  its  organic  constituents,  to  form  a  con- 
clusion on  the  normal  composition  of  a  plant  or  one  of  its  parts.  In  such 
analyses,  whether  we  find  ten  or  thirty  per  cent,  of  sugar,  is  of  no  import- 
ance. We  only  learn  by  this  estimation  that  ten  or  thirty  per  cent,  of 
sugar  are  contained  in  a  certain  part  of  a  certain  plant  under  certain  cir- 
cumstances, but  we  do  not  learn  whether  this  quantity  of  sugar  remains 
constant  under  all  conditions,  or  whether  the  quantity  found  is  the  maximum 
or  minimum  quantity  of  sugar,  or  whether  it  lies  between  both  extremes. 
That  which  has  been  stated  with  regard  to  sugar  is  applicable  to  every 
other  constituent.  Whether  with  the  increase  of  one  constituent  the  quan- 
tities of  a  second  and  a  third  increase  or  diminish,  this  and  the  like  are 
questions  which  remain  unanswered.  Quantitative  analyses  will  only  then 
afford  us  information — that  is  to  say,  will  be  able  to  teach  us  the  normal 
constitution  of  a  vegetable  substance — when  they  have  been  performed,  in 
greater  number,  with  materials  which  have  been  collected  from  plants  at 
certain  epochs  of  their  development,  and  from  plants  which,  under  various 
climatic  relations,  have  been  developed  in  various  positions,  and  from  plants 
whose  age  we  know  correctly,  &c. 
These  investigations,  which  are  necessary  to  afford  information  on  the 
normal  constitution  of  a  single  vegetable  structure,  would  claim  the  in- 
dustry of  a  chemist  for  years.  Consequently,  there  is  little  prospect  existing 
that  we  shall  be  in  possession  of  the  results  of  such  investigations  in  our 
time. 
On  the  grounds  just  explained,  we  have  also  no  rule  forjudging  of  the 
correctness  or  incorrectness  of  the  results  of  quantitative  analyses,  when 
several  of  them  have  been  performed  on  one  and  the  same  material  by 
several  chemists,  and  when  the  quantities  of  the  different  constituents  in 
the  various  analyses  differ  considerably.  Individual  exact  quantitative 
analyses,  consequently,  in  proportion  to  the  time  and  labor  bestowed  upon 
them,  are  of  such  little  service,  that  it  is  advisable  to  neglect  them.  Under 
some  circumstances  the  quantitative  estimation  of  a  constituent  in  a  ma- 
terial may  appear  constantly  necessary,  to  determine  accordingly  its  value 
as  a  medicine  or  an  article  of  commerce.  Estimations  of  that  kind  of  con- 
stituent, with  the  neglect  of  all  others,  possess  proportionately  little  diffi- 
culty, and  the  corresponding  method  may  be  easily  devised  when  a  good  qual- 
itative analysis  lies  before  us. 
With  regard  to  these  quantitative  estimations  of  individual  constituents, 
attention  may  here  be  directed  to  circumstances  of  particular  importance. 
The  knowledge  of  the  constituents  of  a  material  to  be  examined  teaches  us 
frequently  that  it  is  quite  impossible  to  separate  two  constituents  from 
one  another  in  such  a  manner  that  one  of  them  can  be  obtained  pure  with- 
out any  loss.  In  such  cases,  or  also  when  the  separation  can  only  be  ac- 
complished by  tedious  operations,  the  possibility  of  the  indirect  estimation 
of  a  constituent  deserves  our  attention.    This  method  for  the  estimation 
