the  Descent  of  Glaciers  by  their  Weight  only.  43 
a  glacier  bend  itself  down  ?     My  answer  to  this  question  shall 
be  founded  on  Mr.   Mathews's  own  experiment.      He  placed 
a  plank  of  ice  2$  inches  thick  between  two  supports  six  feet 
apart,  and  found  that  it  sank  in  the  middle  until,   in   seven 
hours,  under  the  influence  of  a  thaw   it  had   deflected  seven 
inches.     Of  this  experiment  Mr.  Mathews  says,  "I  regard  it  as 
absolutely   subversive  of  the  Canon's  theory  that   the  descent 
of  glaciers  by  their  weight  alone  is  a  mechanical  impossibility." 
These  are  strong  words.     I  can  only  understand  them  by  sup- 
posing that  Mr.  Mathews  has  other  reasons  than  he  has  alleged 
for  considering  his  experiment  conclusive   on  the  question  at 
issue;  for  the  conditions  under  which  his  ice-plank  bent  and  the 
conditions  under  which  a  glacier  descends  are  entirely  different 
and  have  nothing  in  common.     A  glacier  is  nowhere  an  ice-plank, 
nor  is  it  anywhere  placed  flatwise  between  two  fixed  supports 
and  left  to  bend  vertically  with  nothing  beneath  it  to  rest  upon. 
If  the  Mer  de  Glace  descends  by  bending,  it  must  bend  in  the 
direction  of  its  length,  because  it  is  in  the  direction  of  its  length 
that  it  descends.     To  present  a  parallel  to  the  descent  of  a  gla- 
cier, Mr.  Mathews  should  have  placed  his  ice-plank  between  two 
supports,  not  flatwise,  but  endwise,  and  should  have  observed 
whether  then  it  bent  in  the  direction  of  its  length.     Moreover 
the  Mer  de  Glace  does  not  descend  vertically,  but  on  an  inclined 
plane  of  4°  52'.     The  ice-piank  should  then  have  been  placed  on 
a  plane  inclined  at  that  slope,  and  it  should  have  been  observed 
whether  when  so  placed  it  bent  lengthwise.     And  even  if  it  de- 
scended then  by  bending,  it  would  not  have  descended  under 
similar  conditions  to  those  under  which  the  Mer  de  Glace  de- 
scends— unless,  placing  a  number  of  such  ice-planks  on  one 
another,  those  nearer  to  the  surface  had  descended  faster  than 
those  further  from  it,  notwithstanding  that  they  were  all  frozen 
together.     To  apply  Mr.  Mathews's  experiment  to  the  case  in 
hand,  I  will  suppose  his  ice-plank  to   be  lengthened  from  6 
feet  to  1398  feet.     It  will  then  be  about  as  long,  I  believe,  as  the 
Mer  de  Glace  is  wide  at  Les  Fonts.     I  will  further  suppose  it  to 
become  2.2, GOO  feet  thick,  in  which  case  it  will  be  as  thick  as  the 
Mer  de  Glace  is  long,  measuring  from  the  origin  of  the  Glacier 
de  Lechand  -to  the  Montanvert.     This  colossal  ice-plank,  1398 
feet  long  and  22,600  feet  thick  (or  deep),  I  will  then  suppose  to 
be  laid  down  in  a  uniform  rectangular  channel  at   a  slope   of 
1"  52' ;  and,  the  sides  and  bottom  of  the  channel  being  perfectly 
smooth,  1  will  suppose  the  ice-plank  or  glacier  not  to  be  frozen 
to  them,  but  prevented  from  slipping  by  two  stops  projecting  a 
short  distance  from  the  opposite  sides  of  the  channel  at  the  end 
of  the  glacier.     Now  I  apprehend  that,  according  to  Mr.  .Ma- 
thews's theory,  this  glacier  would,  by  its  weight  only,  bend  in  the 
