382  Prof.  Morley  and  Mr.  Tomlinson  on  Tensile 
continuously  for  about  three  weeks.  Our  overstraining  loads 
are  beyond  this  amount,  being  but  little  below  the  values 
necessary  to  cause  fracture  in  a  very  short  time. 
Materials  and  Apparatus. 
The  aluminium  on  which  the  tests  were  carried  out  was 
all  cut  from  a  1^  inch  round  rolled  bar  supplied  by  the 
British  Aluminium  Company,  the  composition  of  which  is 
given  as 
Al  99*55  per  cent. 
Fe  0-31 
Si    0-14 
The  bar,  the  section  of  which  was  found  by  a  micrometer 
gauge  to  be  1*251  inch  diam.  and  very  uniform,  was  tested 
as  received,  without  machining  of  any  kind.  The  tensile 
strength  of  turned  specimens  was  found  to  be  from  9  to  10 
tons  per  square  inch. 
The  tests  were  carried  out  on  the  50-ton  Wicksteed  single- 
lever  testing-machine,  fitted  with  hydraulic  traverse,  in  the 
Engineering  Laboratory  of  University  College,  Nottingham. 
Any  given  load  can  on  this  machine  be  applied  and  read 
accurately  to  g-J-Q  ton.  The  extensions  were  measured  over 
a  length  of  8  inches  by  means  of  the  latest  type  of  Ewing 
extensometer"*,  and  with  this  instrument  it  was  possible  to 
measure  correctly  to  ^-^0  inch. 
The  application  and  adjustment  of  the  load  to  a  fixed  value 
usually  occupied  from  15  to  20  seconds,  but  on  account  of 
the  creeping  which  occurs,  even  in  the  new  metal,  and  more 
particularly  in  the  overstrained  specimens,  it  was  found 
desirable  to  take  extensometer  readings  at  regular  intervals 
of  one  minute. 
For  aluminium  an  overstraining  load  of  11  tons  was 
chosen,  this  being  well  above  the  yield-point  and  causing  a 
permanent  extension  of  about  rl,j  inch  on  8  inches  and  a 
contraction  of  diameter  from  1*251  to  1*245  inches.  The 
heating  of  the  specimens  was  carried  out  in  a  cylindrical 
bath  of  water  20  inches  high  and  9  inches  diameter,  the 
temperature  of  which  was  kept  at  99°  C.  by  passing  in 
steam  from  a  ^-inch  steam-pipe  under  the  control  of  a  valve. 
The  cooling  to  the  temperature  of  the  atmosphere  took  place 
in  a  large  rectangular  tank  under  a  water-tap.  In  order  to 
bring  out  clearly  the  difference  between  two  load-extension 
*  Proc.  Rov.  Soc.  vol.  Iviii.  1895. 
