220    ice;  its  collection,  storage  and  distribution. 
the  lower  tiers  so  as  finally  to  form  a  solid  mass,  if  left  undis- 
turbed for  a  sufficient  length  of  time.  An  ice  house  in  this  city 
was  filled,  and  left  undisturbed  for  four  years.  During  the  fifth 
year  the  proprietor,  finding  his  stock  of  ice  running  low,  ex- 
amined this  house,  and  found  that,  by  melting,  the  ice  had 
lowered  from  twenty  feet  to  four  feet  in  height,  and  was  one 
solid  cake.  He  was  compelled  to  employ  his  plows,  and  get  the 
ice  out  of  it  in  such  pieces  as  he  could  ;  but  it  carried  him  through 
the  season. 
4.  Exportation.- — In  the  first  part  of  the  present  paper  allu- 
sion has  been  made  to  the  difficulties  encountered  in  the  earlier 
efforts  to  transport  ice  to  southern  and  eastern  markets.  The 
first  cargo  was  despatched  by  Mr.  Tudor  in  February,  1806,  to 
St.  Pi  erre,  Martinique.  He  shipped  about  130  tons,  and  of  this 
only  five  tons  arrived  at  its  destination  ;  and  this  trip  was  at- 
tended with  a  loss  of  about  $4,500.  Details  of  the  different  ex- 
pedients resorted  to  to  avoid  this  loss  would  be  interesting,  but 
the  mention  of  one  must  suffice.  On  one  occasion  he  purchased 
several  large  cases  of  flannels,  and  endeavored  by  winding  the 
pieces  in  and  out  and  around  the  ice,  to  protect  it  from  its  natu- 
ral enemy,  a  high  temperature  ;  but  this  expedient  proved  un- 
successful. At  length,  as  ice  houses  were  erected,  and  the  cor- 
rect principles  for  their  construction  were  gradually  developed, 
it  became  apparent  that  the  same  principles  must  obtain  in  pre- 
paring a  ship  to  carry  a  cargo  of  ice, — converting  it,  in  fact, 
into  a  floating  ice  house  ;  and  this  is  the  way  it  is  now  done. 
The  first  thing  is  to  make  an  even  floor  in  the  hold  of  the  ship, 
by  filling  up  the  furrows,  so  to  speak,  each  side  of  the  keel,  with 
what  sailors  term  "dunnage"  consisting  of  fragments  of  lumber 
or  ballast  of  some  kind.  This  gives  a  tolerably  wide  floor  for 
the  lower  tier  of  ice,  which  (the  floor,  not  the  ice)  is  covered  with 
a  layer  of  straw  or  hay,  and  this  again  with  a  thin  layer  of  coarse 
sawdust  or  wood  turnings.  This  allows  the  water  from  the  melt- 
ing ice  to  trickle  down,  and  to  be  removed  by  the  pumps.  But 
as  the  space  at  the  bow  and  stern  of  the  ship  is  necessarily  nar- 
row, and  would  admit  of  the  packing  of  but  one  cake  of  ice  in  the 
extreme  parts  of  it,  which  would  be  attended  with  great  loss  and 
waste  without  any  compensating  advantages,  it  has  been  found 
