46 
Organic  Mattel'  in  Potable  Waters. 
Am.  Jour.  Pharm. 
Jan.,  1883. 
departure  from  the  mean  of  an  individual  determination  for  organic 
carbon  2*89  per  cent.,  for  nitrogen  7'09  per  cent.,  but  the  departure  in 
some  cases  far  exceeds  these  figures ;  however,  we  find  that  the  results 
of  the  combustion  process  are  less  trustworthy  for  nitrogen  than  for 
carbon.  The  average  departure  from  the  mean  in  the  albuminoid 
ammonia  process  was  :  free  ammonia,  2*23  per  cent. ;  albuminoid  am- 
monia, 3'62  per  cent.  In  Tidy's  process  the  average  divergence  was : 
for  oxygen  consumed  in  one  hour,  1*09  per  cent.;  for  oxygen  con- 
sumed in  three  hours,  0*56  per  cent.,  showing  a  greater  irregularity 
during  the  early  stage  than  later.  The  permanganate  yields,  there- 
fore, the  most  closely  concordant  results,  whilst  the  combustion 
process  yields  the  least. 
Extent  of  Agreement  of  the  Results  Obtained  by  the  Different  Pro- 
cesses with  the  Quantities  of  Organic  Constituents  known  to  be  Actually 
Present. — The  loss  of  carbon  by  the  combustion  process  is  considerable, 
and  there  is  a  strong  tendency  to  excess  of  nitrogen.  The  loss  of 
carbon  is  supposed  to  be  due  to  the  evaporation  of  volatile  substances 
such  as  butyric  and  valeric  acids  which  have  been  shown  to  be  present 
in  some  contaminated  waters ;  the  excess  of  nitrogen  is  due  to  the 
presence  of  ammonia  compounds  given  off  by  the  gas  flame  in  the 
neighborhood  of  the  evaporating  liquid ;  for  although  carefully  cov- 
ered, yet  there  is  still  a  slight  communication  with  the  outside  air  by 
means  of  the  notch  in  the  rim  of  the  water-bath  for  the  passage  of  the 
feed-flask  neck.  Suggestions  are  made  so  as  to  show  how  this  access 
of  ammonia  may  be  prevented  by  evaporating  the  water  by  steam,  by 
evaporating  the  water  in  a  vacuum,  etc.  The  loss  of  nitrogen  in 
Wanklyn's  process  is  due  to  volatilization  of  the  amines  during  the 
first  distillation,  and  as  these  compounds  are  not  indicated  by  Nessler, 
they  escape  detection  as  free  "  or  as  "  albuminoid  "  ammonia.  In 
order,  therefore,  to  diminish  this  loss,  a  separate  distillation  should  be 
made  with  alkaline  permanganate  added  at  once ;  also,  to  avoid  the 
uncertain  ending  of  the  collection  of  ammonia,  the  distillation  should 
not  be  stopped  until  the  last  measure  of  distillate  contains  less  than  1 
per  cent,  of  the  whole  .  ammonia  already  collected.  Several  other  • 
alterations  in  the  details  of  the  process  are  also  recommended.  The 
analytical  figures  obtained  by  Tidy's  process  show  that  the  putrescent 
substances  are  accurately  determined,  but  not  so  the  non-putrescent,  as 
their  oxidation  by  cold  permanganate  is  so  slow.  On  the  other  hand, 
by  Kubel's  method  (oxidation  at  100°),  a  loss  of  volatile  matter 
