AmMa"'Srm'}  Law  Applicable  to  Gases  and  Liquids.  237' 
extreme  case  of  a  mixture  of  gaseous  ammonia  and  gaseous 
water. 
This  last  case  I  consider  to  be  most  important  and  most  decisive. 
The  details  are,  "  a  quantity  of  dry  ammonia  was  measured  over 
mercury,  then  a  small  portion  of  water,  which  had  been  accurately 
weighed  in  a  thin  glass  bulb,  was  introduced  into  the  ammonia. 
The  whole  was  then  heated  up  to  ioo°  C,  and  the  volume  of  mixed 
gas  and  aqueous  vapor  noted. 
Corrected  vol. 
at  o'6  and 
Observed  cc.         T.°  P.  mm.  760  mm. 
NH3   39-19  12-5  558-o8  24-008 
NH3  -{-  H20  107-32  102  709-66  72-973 
The  weight  of  water  employed  was  0-0402  grm.  On  calculation 
it  will  be  found  that  0-0402  grm.  H20  yields  49  95  cc.  at  0-6  and 
760  mm.,  whilst  the  experiment  gave  48-965  cc,  which,  bearing  in 
mind  the  experimental  difficulties  of  such  an  experiment,  is  a  suffi- 
ciently close  agreement.  In  short,  there  is  the  most  varied  evidence 
that,  so  long  as  there  is  no  actual  chemical  action  between  the 
molecules  by  a  gaseous  mixture,  the  fact  of  the  molecules  being 
dissimilar  has  no  influence  upon  the  volume." 
Liquids. — The  law  holds  in  the  instance  of  liquids  just  as  in  the 
instance  of  gases.  Such,  I  believe,  is  the  fair  interpretation  of  the 
results  which  have  been  recently  published  in  the  Chemical  News 
and  in  the  Philosophical  Magazine. 
In  July  last  Cooper  and  myself  published  the  preliminary  result 
that  a  strong  solution  of  cane  sugar  occupies  the  same  volume  as 
the  solid  sugar  and  water  of  which  it  is  composed,  and  set  down 
0-  371  as  the  value  of  i,  the  increment  co-efficient  of  cane-sugar. 
In  the  November  number  of  the  Phil.  Mag.,  an  elaborate  investi- 
gation is  published.  Employing  a  half-litre  specific  gravity  bottle, 
and  taking  great  precautions,  and  covering  a  wide  range,  viz.,  from 
1-  3  grms.  up  to  104-6  grms.  of  sugar  in  a  litre  of  solution,  we  obtain 
a  slightly  higher  value,  viz: 
Strength,  i.e.,  Xo.  of 
Grms.  of  Sugar  in  Litre  of 
Solution.  Sp.  gr.  %. 
I'34I  I0OO-52  0-3878 
9-878  1003-84  0-3887 
104-580  1040-60  0-3882 
Over  this  wide  range  it  would  seem  that  i  is  absolutely  constant. 
When  the  strength  is  greatly  increased,  at  749-5  grms.  per  litre,  for 
