
 
		only  on  its  entering  the  black  fea,  but  never  in  going  out j 
 “   becaufe,  fays  he,  it  does  not  ordinarily go  out.  The  ièafon 
 “   of this  is,  that  o f  all  the  other  fpecies  o f fiilies, the  Sardine 
 *j  alone  fwims  up  the Danube ;  and when it  comes  to  thè diramation  
 of  the  river,  is  carried  into  the  Adriatick.’ - *   We  
 ihall  a  little  further  on  take notice  again  o f  this  voyage  o f  the  
 >  let  it  fuffice  juft  now  to  have  hinted,  that  fome,  not  
 without  reafon,  derive  the  origin  o f  the  refluent  lakes  o f  thefe  
 parts,  and  the  large  fifties  that  are  found  in  them,  from  a  dira-  
 mation  o f  the  Danube,  or  of  fome  large  river  communicating  
 with  it,  fuch  as  the  Sava,  perhaps  open  in  former  times,  and  
 now  covered  by  the  ruins  o f  mountains,  and  deviated.  This  
 is  a  point  o f natural  hiftory  nearly  connected  with  other  points  
 o f  ancient  hiftory,  and  which  deferves  to  be  accurately  dif-  
 eufled.  -{* 
 O f the  Caverns  and  Gulphs. 
 There  are  feveral  gulfs  in  this  iiland  of CherJb,  though  not  fo  
 many  as  in  the  neighbouring  continent,  and  particularly  o f  
 Iflria.  We  vifited  two  o f them,  but  could  not  go  down  into  
 the  firft  for want  o f time,  and the neceffary  helps,  fuch  as  ropes,  
 ladders,  lights,  &c.  It  is  fituated a  ihort mile  from  the  habitation; 
 *  Arift.  Hift.  Anim.  Lib.  VIIT.  Cap:  ig : 
 + Under the  title of the jmall ijlands adj-acent to Cherfo tmd'Ofero I  (hall treat more  
 at  large  on  the  changes  wrought  in  litrra  by  the  fiMdendes  o f  the  marble  
 ftrata.  The  lakes,  the  dreadful  caves,  the  gulfe,  and  the  vait  depofites of land  
 that are  to  oe  feen  on  the  hills  of that  province,  and  the  adjacent  little  fandy  
 iflands,  have  all,  in  my  opinion,  one  common origin,  from which,  however,  I  
 would  except  fome  particular  effeifs,  as proceeding  from  caufes  purely  local.  
 T h is  fubject would require  to be treated exprefsly by iti elf. 
 tation  o f the  courteous  Signor  Jacopo  Colombls,  at  3 .  Pietro di  
 Mezzavia. 
 The  ground  all  around  is  covered  with  very  fliarp  and  cutting  
 pieces  o f rocks,  among which  fome  few  trees  grow,  and  bruih-  
 wood  exceedingly  cloie  together.  This  incommodious  road  
 goes  bending a  little  in  declivity,  as  you  come  nearer  the  mouth  
 o f  the  gulf,  and  there  the  trees  are  pot  fo  rare.  The  maffes  
 that  furround  the  cavern  are  very  large,  and  parts  o f a  ftratum  
 two  feet  and  a half thick,  disjoined  by  the water,  and  fubverted  
 for want o f   an  equal  bafe,  yet  fupported,  a.s  it were  hanging  in  
 the  air  by  the  large  roots  o f  the  trees,  The  aperture  is  about  
 fifteen  feet wide;  and  is  eroded  by  the  toot of a large  ilex  The  
 branches o f  that  old  tree  that  incumber  the  entrance,  render  it  
 ftill  more  dark  and  frightful.  The  depth  of  this  cavern  cannot  
 be  afcertained.  A   paufe  is  feen  from  the  brink,  which  may  be  
 about  twenty  feet  below j  but  the  gulf  do.es  not  end  there,  
 another  lateral  hole  opening  near  it,  through which  the  ftones  
 thrown  in  by  the  curious  roll,  and  rebound a  long  time  before  
 they  ftop.  O f  this  abyfs,  whofe  fubterraneous  obfeurity  muft  
 be  vaft  and  intcrpfting,  I  hope  one  day  to  be  able  to  give  a  
 more  particular  account,  i f   circumftances  permit me  to  return  
 to  that  iiland;  as  I  look  upon  exait  defcriptions  o f  the  interior  
 parts  of our  globe,  to  be  equally  ufeful  to  mineralogy,  as  anatomical  
 diffeftions  o f  the  human  body  are  to  phyfick.  In  the  
 northern  part  of Cherfo,  which  we  had  not  time  to  vifit,  there  
 are many  precipices,  and magnificent  grottos,  caverns  and  dens  
 well worthy  o f  obfervation,  at  leaft,  i f   they  correfpond  to  the  
 defcriptions  given  h s   of them.  The  rent  of  the mountain  near  
 Smergo,  where  they  ufually  pafs  from  Cherfo to  the  neighbouring  
 iiland  of Vaglia,  muft  be  a  magnificent  fpeftacle  to  a  geologift, 
 K   k   k   2  i f