Intelligence  and  Miscellaneous  Articles.  157 
not  mentioned  in  the  above  list.  In  these  experiments  a  spectro- 
scope with  a  large  dispersive  and  magnifying  power  was  found  to  be 
required  in  order  to  identify  the  lines  in  the  solar  spectrum,  so  that 
the  hydrogen  spectrum  became  so  reduced  in  brightness  that,  in  order 
to  see  the  fainter  lines,  the  eye  required  to  be  kept  for  some  minutes 
in  the  dark  room,  although  with  a  spectroscope  of  low  power  the 
spectrum  appeared  very  bright  and  full  of  lines. — Monthly  Notices 
of  the  Royal  Astronomical  Society,  December  8,  1871. 
OX  THE  DISENGAGEMENT  OF  HEAT  WHEN  CAOUTCHOUC  IS 
STRETCHED.         BY  PROF.  E.  VILLARI. 
In  a  communication  made  to  the  Royal  Society  in  1857,  Joule* 
describes  some  investigations  on  the  thermal  effects  which  occur 
when  threads  of  various  substances  are  stretched  or  contract.  It 
results  from  an  accurate  investigation  of  this  subject  that  metal 
wires  cool  on  being  stretched,  and  become  heated  on  contrac- 
tion— but  that  threads  of  vulcanized  caoutchouc  have  quite  the 
opposite  comportment,  becoming  heated  when  they  are  stretched,  and 
cooled  when  they  contract.  This  remarkable  fact  has  since  been 
investigated  by  other  physicists,  for  instance  by  Govif,  and  recently 
by  Pierre^,  and  always  with  the  same  result. 
In  my  recent  investigations  on  the  elasticity  of  caoutchouc,  I  had 
occasion  to  repeat  Joule's  experiment  on  threads,  cords,  and  strips 
of  caoutchouc  of  various  dimensions,  and  by  means  of  very  simple 
apparatus.  It  consisted  of  a  thermopile  connected  with  a  galvano- 
meter, and  so  suspended  to  silk  threads  that  it  could  be  easily  lowered 
upon  the  caoutchouc  band  to  be  examined,  or  raised  away  from  it. 
The  strip  of  caoutchouc  was  thus  horizontal,  fastened  at  one  end  to 
a  strong  table,  and  at  the  other  to  a  lever,  so  that  it  could  be  readily 
stretched  and  afterwards  loosened.  Each  time  the  caoutchouc  was 
stretched  or  loosened,  the  thermopile  was  raised,  and  then  placed 
with  its  whole  weight  upon  the  caoutchouc,  so  as  to  determine  the 
change  in  temperature  produced. 
From  the  experiments  made  with  this  apparatus,  it  followed,  in 
accordance  with  the  labours  of  the  above-mentioned  physicists,  that 
caoutchouc  becomes  heated  on  being  stretched,  while  it  cools  on 
contraction.  I  may  mention,  as  a  new  fact  (for  I  am  not  aware 
that  it  has  been  already  observed  by  others),  that  the  absolute  value 
of  the  increase  of  temperature  on  stretching  is  greater  than  the  de- 
crease of  temperature  on  contraction.  This  difference  is  often  ob- 
served when  the  increase  of  temperature  on  stretching,  and  the  decrease 
*  Phil.  Mag.  vol.  xiv.  p.  226.  [This  remarkable  fact  was  first  noticed 
in  unvulcanized  caoutchouc  by  John  Gough,  by  simply  holding  the 
caoutchouc  against  the  lips  and  stretching  it.  See  Nicholson's  Journal, 
vol.  xiii.  (1806)  p.  305. — Ed.  Pogg.  Annalen.'] 
t  Les  Mondes,  April  22,  1809. 
X  Ibid.  April  8  and  May  6,  1869. 
