100  PROCEEDINGS  OE  THE  GEOLOGICAL  SOCIETY.  [May  1 893, 
rostrcita,  cannot  be  collected  at  all  above  this  zone.  Finally, 
several  small  but  well-known  species,  such  as  Conus  dormitor  and 
Buccinum  desertum,  here  make  their  first  appearance. 
The  Upper  Barton  division  is  represented  as  having  a  development 
of  90  feet,  and  is  far  more  variable  both  in  its  fauna  and  lithology  ; 
in  fact  it  comprises  within  itself  a  triplicate  subdivision,  of  which 
the  upper  portion  contains  the  Hampshire  Upper  Bagshots,  which 
are  also  known  as  the  Headon  Hill  Sands  (the  Long-Mead-End 
Sands  of  the  Authors).  This  Upper  Barton  as  thus  constituted  is,  in 
fact,  a  somewhat  artificial  division,  its  base  being  the  Chama- bed. 
The  fauna  of  the  Chama- bed  numbers  altogether  170  known 
species,  and  differs,  as  a  whole,  more  from  that  of  the  Lower  and  even 
Middle  Barton  than  does  that  of  the  Lower  Barton  from  the  Upper 
Bracklesham.  The  greater  number  of  what  may  be  considered  the 
typical  Barton  species  have  disappeared.  It  is  difficult  to  say 
positively  whether  the  change  was  from  deeper  to  shallower  water, 
or  the  reverse  ;  but  the  shell-bed  preceding  the  Chama-bed  plainly 
indicates  a  long  period  when  no  mud  was  being  deposited,  and  the 
Chama- bed  itself  was  formed  in  clear  water.  It  serves  to  show, 
like  so  many  other  abrupt  transitions  in  the  Eocene,  that  a  relatively 
slight  change  in  physical  conditions  makes  a  far  greater  impression 
on  the  succession  of  life  in  a  formation  than  would  be  occasioned  by 
an  enormous  lapse  of  time  without  such  a  change.  This  is  the  last 
appearance  of  Chama  in  England. 
An  analysis  of  the  entire  Barton  fauna  is  given  towards  the  end  of 
the  paper,  with  a  table  enumerating  also  those  species  which  ascended 
from  below,  and  those  which  continued  into  the  Headon  Series.  A 
somewhat  critical  examination  of  the  Edwards  collection  shows 
about  527  varieties  of  Mollusca  from  the  Barton  Beds  entitled  to 
specific  rank,  and  the  Authors  are  not  of  opinion  that  this  number 
will  ever  be  greatly  exceeded.  Fossils  belonging  to  other  groups 
bring  the  fauna  to  a  possible  total  of  600.  The  list  comprises  23 
Vertebrates,  47  Invertebrates  other  than  Mollusca,  257  Gasteropods, 
and  150  Bivalves,  exclusive  of  over  120  undetermined  species. 
The  Barton  Series  in  Hampshire  is  now  separated  from  its 
representative  in  the  London  Basin  by  an  interval  of  about  60 
miles — a  break  in  all  probability  due  to  post-Eocene  denudation. 
There  can  be  no  doubt  that  exact  correlation  in  this  case  must  be 
attended  with  much  difficulty ;  still  the  Authors  roundly  assert  that 
the  fauna  precludes  the  idea  of  any  correspondence  between  the  Upper 
Bagshot  Sands  of  Surrey  and  the  sand}^  beds  of  the  Upper  Barton, 
