148  The  Metameric  Amyl  Nitrites.  {Amkarch\*?™' 
gives  results  varying  considerably.  The  alkaloids  in  the  ethereal  solu- 
tion ought  to  be  submitted  to  the  process  a  second  and  even  a  third 
time,  bat  even  with  this  improvement  the  whole  of  the  cinchonidine  is 
not  obtained,  and  the  results  vary  much,  but  the  composition  of  the 
bye-product  is  more  uniform  than  in  the  other  processes.  The  crystal- 
lization test  has  the  same  advantages  as  the  bisulphate  test  if  the  crys- 
tallization is  repeated  often  enough,  and  is  the  process  which  is  least 
influenced  by  the  presence  of  hydro-bases.  It  is,  however,  tedious. 
The  process  of  the  German  Pharmacopoeia  depends  on  the  fact  that 
the  precipitate  produced  by  ammonia  in  solutions  of  the  alkaloids  is 
soluble  in  excess  of  ammonia,  but  that  much  less  ammonia  is  required 
for  quinine  than  for  the  other  alkaloids.  The  excess  of  ammonia  re- 
quired varies,  however,  very  considerably  with  the  temperature. 
THE  METAMERIC  AMYL  NITRITES. 
By  Peofessoe  Duxstax  and  W.  Lloyd  Williams. 
The  liquid  which  has  hitherto  been  known  to  scientific  chemists  as- 
pure  amyl  nitrite  is  a  mixture  of  the  metameric  nitrites,  corresponding 
to  a-amyl  alcohol  or  iso -butyl  carbinol  (CH(CH3)2CH2CH2X02)  and 
/9-amyl  alcohol  or  secondary  butyl  carbinol  (CH(C2H5)CH3CH2N02). 
The  relative  proportions  of  these  two  nitrites  in  the  mixture  are  de- 
pendent on  the  composition  of  the  "  amyl  alcohol w  from  which  they 
have  been  prepared.  The  respective  quantities  of  these  two  meta- 
meric alcohols  contained  in  different  specimens  of  fusel  oil  are  subject 
to  variation.  The  quantity  of  the  a-amyl  alcohol,  which  is  optically 
inactive,  is  always  much  larger  than  that  of  /9-amyl  alcohol,  which  is 
lsevorotatory.  In  the  portion  of  fusel  oil  which  has  been  repeatedly 
fractionated  between  127°— 132°,  there  is  said  to  be  usually  about  13 
per  cent,  of  the  active  alcohol.  These  alcohols  cannot  be  separated  by 
fractional  distillation,  since  the  a-amyl  alcohol  boils  at  128°  C.  and 
the  /9-amyl  alcohol  at  131°  C.  Their  separation  is  a  matter  of  very 
great  difficulty,  and  it  has  seldom  been  accomplished. 
Our  first  experiments  related  to  the  properties  of  the  mixture  of 
these  two  nitrites.  The  statements  of  different  observers  about  its 
properties  are  inconsistent,  the  boiling  point  being  variously  stated  at 
almost  every  temperature  between  90°  and  99°  C.  These  discrepancies 
are  no  doubt  in  part  due  to  the  employment  for  the  preparation  of  the 
