Minimum  Deviation  through  a  Prism.  609 
all  opposite  results  were  fault)".  We  have  proved  experi- 
mentally, that  in  the  cases  where  the  expansion  was  found  too 
great  (Cantone)  there  were  errors  in  the  measurements ;  and 
in  the  cases  where  the  expansion  was  found  too  small  (Quincke 
and  Wullner  &  Wien),  we  proved  also  by  experiments,  that 
the  variation  of  the  dielectric  power  agrees  in  sign  and  value 
with  the  observed  discrepancy.  Accordingly  there  remain 
no  observations  leading  to  the  conclusion  that  there  are  other 
forces  in  electrostriction  than  the  pressure  o£  the  charged 
armatures. 
We  are,  Gentlemen, 
Aachen  and  Danzig,  YoiTrs  faithfully, 
February  1906.  "'  A.  WuLLNER  ;   M.  WiEN. 
LIII.  Note  on  "  Minimum  Deviation  through  a  Prism.'3 
By  R.  Chaktres  *. 
0 
0+$ = minimum.     6'+cf>'=  constant,     Sin  6  =  p  sin  &.     Sin  <£  =  ^  sin  </>' 
Make  F  C  B  =  0,     FB  C  =  0, 
then  F  is  to  be  a  maximum,  while  D  is  constant. 
(1)      .'.    cosF-cosD=^(l--2),  will 
But  since 
be  a  minimum. 
/*sin  (fl'  +  M 
~  sin  (#  +  </>)     ' 
.*.     n,  as  also  (  1 2  ),  will  be  a  maximum  when  0  +  (f> 
is  a  minimum. 
.'.    that  (1)  should  be  a  minimum  xy  must  be  a  maximum  ; 
x  =  ?/,      6'  =  (/)',     and  6  =  (f>. 
*  Communicated  by  tlie  Author,  being  reprinted,  with  additions,  from 
the  Phil.  Mag.  vol.  vi.  p.  529  (1903). 
