5  36  A  Ikaloidal  Method  for  Fluid  Extracts.    { Am6cJtoberPi8%m" 
Coffee. — William  E.  Weiss,  Ph.G.,  obtained  concordant  results 
when  he  applied  the  method  to  fluid  extract  of  coffee  in  duplicate 
assays.  The  percentages  of  caffeine  shown  by  eleven  samples  of 
this  preparation,  said  to  have  been  made  from  the  green  berries, 
were  as  follows:  074,  0  82,  0.74,  0-88,  046,  0-54,  0-54,  0-14,  0-30, 
0-28  and  014;  two  samples,  which  had  been  made  from  roasted 
berries  and  intended  for  making  syrup,  contained  0-38  and  0  27  per 
cent,  of  the  alkaloid.  These  results  also  indicate  the  poor  quality 
of  the  product  on  the  market. 
Guarana. — Edythe  Weston,  Ph.G.,  found,  upon  assaying  eight 
brands  of  fluid  extract  of  guarana  from  the  market,  that  the  prepa- 
ration usually  yielded  from  2-62  per  cent,  to  3-50  per  cent,  of 
caffeine,  although  an  exceptional  case  showed  5-97  per  cent.  The 
exact  percentages  were  as  follows  for  the  duplicate  assays : 
I. 
2. 
3- 
4- 
5- 
6. 
7- 
8. 
2*60 
2-88 
3-00 
3-20 
3 '34 
3*34 
3  "44 
5*96 
2-64 
2*94 
3*46 
3"40 
3*42 
3-48 
3'56 
5-98 
The  firm  that  supplied  sample  No.  1  claimed  it  to  be  a  standard- 
ized preparation,  containing  4- 50  per  cent,  of  caffeine,  or  that  100 
c.c.  would  yield  4-5  grammes  of  this  alkaloid.  They  also  recom- 
mended that  Professor  Lloyd's  method  of  assay  be  tried  as  a  means 
of  estimating  the  caffeine.  Five  trials  were  made  by  this  method, 
but  the  figures  given  in  the  table  could  not  be  exceeded.  A  sam- 
ple of  the  powder  from  which  this  fluid  extract  was  made  was  sub- 
mitted to  the  same  treatment  as  in  the  assay  of  the  preparation. 
The  residue  from  the  evaporation  of  the  chloroformic  solution  con- 
tained a  fatty  substance;  this  was  separated  by  treating  with  acidu- 
lated water,  and  its  weight  was  then  deducted  from  the  first  con- 
stant weight.  By  this  means  2  72  and  2-98  per  cent,  of  caffeine 
were  found.  From  these  results  it  might  be  concluded  that  this 
method  of  assay  is  also  applicable  to  the  powdered  drug.  The 
fatty  substance  was  not  encountered  in  working  with  the  fluid 
extracts.  In  both  of  the  assays  made  of  Nos.  3  and  8,  the  magmas 
when  treated  with  chloroform  yielded  dark-colored  solutions  through 
small  quantities  of  iron,  which  were  dissolved.    An  ash  determina- 
