Two  New  Tests  for  Sugar.  443 
kited  with  water,  differs  totally  in  character  from  that  formed  in  sac- 
charine liquids.  The  limit  of  sensibility  of  Fehling's  test  is  0*0008 
per  cent.,  and  that  of  Trommer's  test  is  0*0025  per  cent. 
2.  If,  instead  of  the  alpha-naphthol  in  the  preceding  test,  an  alco- 
holic solution  of  thymol  of  similar  concentration  be  employed,  the 
remaining  manipulations  being  the  same  as  before,  a  deep  red  varying 
from  cinnabar  to  carmine  is  produced ;  dilution  with  water  brings  the 
color  to  carmine,  and  after  a  time  there  separates  a  flocculent  precipi- 
tate, which  dissolves  with  a  pale-yellow  color  in  alcohol,  ether,  and 
caustic  potash,  but  with  a  bright  yellow  in  ammonia. 
The  delicacy  of  this  reaction  is  about  the  same  as  of  that  with  alpha- 
naphthol. 
After  many  experiments  had  shown  the  trustworthiness  of  the  re- 
sults given  by  these  tests,  it  wTas  interesting,  on  account  of  their  ex- 
ceeding delicacy,  to  apply  them  to  the  solution  of  the  disputed  ques- 
tion whether  normal  human  urine  does  or  does  not  contain  sugar. 
The  results  of  the  first  attempts  were  so  decided  that  the  urine  ex- 
amined appeared  to  be  diabetic.  The  urine  of  a  number  of  perfectly 
healthy  individuals  was  therefore  examined,  but  with  precisely  the 
same  results.  The  tests  were  made  with  alcoholic  solutions  of  alpha- 
naphthol  and  thymol,  exactly  as  has  been  described,  and  the  extraor- 
dinary delicacy  of  the  reactions  can  be  better  understood  by  the 
statement  that  normal  urine  diluted  to  from  one  hundred  to  three 
hundred  times  its  volume  with  water  still  gives  a  recognizable  reac- 
tion. When  the  urine  is  diluted  to  four  hundred  times  its  volume, 
the  test  shows  no  result. 
In  order  that  there  might  be  no  question  as  to  sugar  being  the 
actual  cause  of  the  reaction,  the  following  substances  were  examined 
and  gave  negative  results  with  both  alpha-naphthol  and  thymol : 
urea,  creatine,  xanthine,  uric  acid,  allantoine,  hippuric  acid,  succinic 
acid,  phenol,  pyrocatechin  and  indican. 
These  results  fully  confirm  the  opinion  advanced  by  Briicke,  and 
supported  by  many  other  observers,  that  normal  urine  constantly  con- 
tains sugar. 
These  tests  may  be  applied  in  two  different  manners  in  order  to  dis- 
tinguish a  normal  from  a  diabetic  urine. 
1.  A  specimen  of  normal  urine  and  of  that  to  be  tested  are  equally 
diluted  with  water  to  about  one  hundred  times  their  volume ;  the 
same  quantity  of  each  is  then  tested  just  as  has  been  described;  if 
Am.  Jour.  Pharm. 
Sept.,lSS6. 
