466  Note  on  Assay  of  Formaldehyde.       {^"ctober  ivn^' 
These  results,  therefore,  Hke  the  results  given  in  Table  III, 
show  that  when  only  a  comparatively  small  excess  of  the  KCN 
is  used,  the  effect  of  a  considerable  variation  in  the  temperature 
manifests  itself  quite  markedly  in  the  results;  but  when  the  excess 
of  KCN  is  increased  to  about  one-half  of  an  equivalent,  the  results 
obtained  at  40°  C.  are  only  slightly  lower  than  those  obtained  at 
5°  C. ;  while  the  results  given  in  Table  III  show  that  a  variation 
of  the  temperature  between  5°  C.  and  30°  C.  had  no  appreciable 
effect  on  the  found  value  of  -the  formaldehyde  solution  when  the 
excess  of  KCN  was  about  one-half  of  an  equivalent.  We  may  con- 
clude, therefore,  that  the  ordinary  changes  in  room-temperature 
during  the  year  will  not  appreciably  affect  the  results  when  the 
excess  of  KCN  is  as  much  as  one-half  of  an  equivalent. 
A  closer  examination  of  the  potassium  cyanide  method  seemed  to 
show  also  that  there  is  nothing  inherent  in  the  basic  principle  of  this 
method  v\^hich  would  prevent  its  application  directly  to  the  concen- 
trated formaldehyde  solutions.  That  such  is  really  the  case  may 
be  seen  from  the  following  experiment:  To  0.5505  Gm.  of  a  sample 
of  U.S. P.  formaldehyde,  weighed  out  in  a  closely  fitting  glass-stop- 
pered Erlenmeyer  flask  of  about  150  c.c.  capacity,  there  were  added 
100  c.c.  of  an  approximately  n/io  KCN  solution  and  the  two  solu- 
tions well  mixed.  This  mixture  was  then  added  to  a  mixture  of 
35  c.c.  n/io  AgNOg  and  20  c.c.  10  per  cent,  nitric  acid  in  a  200  c.c. 
measuring  flask,  the  Erlenmeyer  well  washed  and  the  washings 
added  to  the  silver  solution,  the  whole  made  up  to  200  c.c,  thor- 
oughly shaken,  and  filtered  through  a  dry  filter.  100  c.c.  of  this 
filtrate  were  found  to  require  2.00  c.c.  n/io  KCNS.  This  would 
make  the  formaldehyde  in  the  0.5505  Gm.  of  the  sample  equivalent 
to  67.75  c.c.  n/io  AgNOo  which  would  correspond  to  0.20325  Gm. 
HCHO  or  36.92  per  cent.  On  analyzing  a  diluted  solution  of  this 
sample,  the  result  obtained  corresponded  to  36.98  per  cent.  It  is 
thus  seen  that  not  only  is  the  potassium  cyanide  method  applicable 
to  concentrated  formaldehyde  solutions  by  previously  diluting  them 
with  water  but  that  it  may  even  be  applied  to  the  concentrated  solu- 
tion directly,  just  as  in  the  case  of  the  hydrogen  peroxide  method. 
It  has  already  been  pointed  out  above  that  in  the  case  of  the 
concentrated  formaldehyde  solution  the  H2O2  method  seems  to 
48  c.c.  of  this  KCN  solution  was  equivalent  to  47.4  c.c.  n/io  AgNO,. 
