^""sept^'im""""}    Relative  Value  of  Different  Pepsin  Tests.  471 
6.  The  accuracy  of  the  multipliers  7  and  8  is  not  infallible  in  every 
test. 
As  Manwaring  lays  particular  stress  upon  the  question  of  dilu- 
tion, I  think  his  test  is  a  decided  improvement  over  the  U.  S.  P. 
test. 
The  next  good  point  in  his  test  lies  in  the  fact  that  he  does  not  at- 
tempt to  weigh  the  undigested  albumen,  as  is  done  in  the  U.  S.  P.  test, 
and  thereby  does  away  with  a  great  source  of  error ;  but  instead  of 
this  he  figures  the  amount  of  albumen  (?)  digested  upon  a  dry  basis, 
and  then  tries  to  convert  this  dry  basis  by  calculation  into  albumen  on 
the  wet  basis.  In  doing  this  errors  are  apt  to  occur  as  I  have  pointed 
out,  but  I  do  not  think  that  they  are  errors  of  such  magnitude  as  are 
apt  to  be  obtained  with  the  U.  S.  P.  test. 
Finally,  I  wish  to  say  a  few  words  about  a  test  which  I  consider  to 
be  the  only  approach  to  an  accurate  method  of  testing  pepsin  that  I 
know  of.  I  do  not  claim  that  this  test  is  absolutely  accurate  either, 
as  slight  errors  are  apt  to  occur,  which,  however,  do  not  materially  in- 
jure the  final  result.  I  refer  to  the  Kremel  test,  which  was  published 
some  time  since  in  the  Druggists'  Circular. 
In  devising  this  test  Kremel  has  made  a  radical  departure  from  the 
usual  methods,  and  bases  his  test  upon  the  fact  that  under  the  condi- 
tions in  which  artificial  peptic  digestions  take  place,  pepsin  alone  has 
the  property  of  converting  albuminoid  matter  into  peptone,  and  that, 
therefore,  from  an  analytical  as  well  as  from  a  physiological  stand- 
point, the  only  correct  method  is  to  take  the  quantity  of  peptone  pro- 
duced as  a  gauge  of  the  action  of  the  pepsin ;  or  in  other  words,  the 
test  is  made  to  resemble  as  nearly  as  possible  the  conditions  existing 
in  the  natural  process. 
Without  going  into  any  further  detail,  the  test  is  made  as  fol- 
lows : 
One  gm.  of  egg  albumen  (soluble)  dried  at  40°C.  and  pulverized, 
and  0*1  gm.  of  the  pepsin  to  be  tested,  are  placed  into  a  100  cc.  flask, 
and  dissolved  in  50  cc.  of  0.2  per  cent,  hydrochloric  acid.  The  solu- 
tion is  heated  to  38-40°C.  for  three  hours,  and  then  exactly  neutra- 
lized with  sodium  carbonate  ;  it  is  then  heated  on  a  water  bath  to 
90^C.,  and  cooled  after  coagulation  has  taken  place.  The  flask  is  then 
filled  to  the  mark  with  distilled  water,  and  50  cc.  are  filtered  off*  and 
evaporated  to  dryness  in  a  platinum  dish  on  a  water  bath. 
The  residue  is  dissolved  in  hot  distilled  water,  filtered  through  a 
