rANHisE.]  DISCUSSIONS    OF   PRINCIPLES,  517 
)f  the  great  folds  and  those  of  the  minute  plications  that  a  very  small  specimen 
jroperly  oriented  gives,  in  many  cases,  the  key  to  the  structure  over  a  large  portion 
)f  the  side  of  a  fold. 
It  is  to  be  noted  that  the  statement  of  these  principles  has  been  in- 
nspired  by  a  study  of  formations  which  have  proved  to  be  wholly  Cam- 
brian or  post-Cambrian. 
The  principles  of  lithological  correlation  enunciated  by  Irving  are  as 
follows,  except  that  series  is  here  substituted  for  group  and  Algonkian 
for  Huronian  so  as  to  make  the  terminology  correspond  with  that  here 
ised: 
Lithological  characters  are  properly  used  in  classification : 
(1)  To  place  adjacent  formations  in  different  series,  on  account  of  their  lithologi- 
3al  dissimilarities  when  such  dissimilarities  are  plainly  the  result  of  great  alteration 
n  the  lower  one  of  the  two  formations,  and  are  not  contradicted  by  structural  evi- 
lence,  or,  if  used  as  confirmatory  evidence  only,  when  such  dissimilarities  are  the 
result  of  original  depositional  conditions. 
(2)  To  collect  together  in  a  single  series  adjacent  formations  because  of  lithologi- 
jal  similarities  when  such  similarities  are  used  as  confirmatory  evidence  only. 
(3)  To  correlate  series  and  formations  of  different  parts  of  a  single  geological  basin 
svhen  such  correlations  are  checked  by  stratigraphy,  and  particularly  by  observa- 
tions made  at  numerous  points  between  the  successions  correlated. 
They  are  improperly  used : 
(1)  To  place  adjacent  formations  in  different  series  on  account  of  lithological  dis- 
similarities when  such  dissimilarities  are  merely  the  result  of  differences  in  original 
lepositional  conditions,  and  when  such  evidence  of  distinction  is  not  confirmed  by 
>r  is  contradicted  by  structural  and  paleontological  evidence. 
(2)  To  collect  in  a  single  series  adjacent  formations  because  of  lithological  sinii- 
arities  when  such  similarities  are  not  confirmed  by  or  are  contradicted  by  other  evi- 
lence. 
(3)  To  establish  general  correlations  between  the  clastic  series  of  different  geologi- 
al  basins,  except  possibly  when  the  gneissic  and  true  crystalline-schist  basement 
ormation  of  one  region  is  compared  with  the  similar  basement  formation  of  another. 
(4)  To  establish  and  determine  any  world-wide  subdivisions  of  the  noneruptive 
asement  crystallines,  i.  e.,  those  which  underlie  the  clastic  series  here  called  Algon- 
dan,  at  least  until  very  much  more  definite  evidence  of  the  existence  of  such  sub- 
[ivisions  be  gathered  than  has  hitherto  been  done. 
In  applying  these  principles  it  must  not  be  forgotten  that  a  bed  of 
me  character  may  thin  out  and  disappear ;  may  gradually  change  from 
b  limestone  to  a  shale,  from  a  shale  to  a  sandstone  or  conglomerate; 
bnd  that  sometimes  the  change  may  be  abrupt,  as  perhaps  upon  the 
opposite  sides  of  an  axial  ridge,  one  side  of  which  faces  toward  the 
cean  and  the  other  toward  an  interior  sea.  All  formations,  however 
widespread,  terminate  somewhere.  A  single  formation  of  a  certain 
ithological  character  can  only  be  assumed  to  be  the  same  bed  in  a  dis- 
trict when  it  has  been  demonstrated  to  be  persistent  over  a  wide  area. 
When  several  characteristic  formations  occur  in  a  definite  order  in  dif- 
ferent parts  of  the  same  district  the  probability  that  they  are  of  iden- 
lical  age  is  greater  than  with  single  beds  found  to  be  lithologically  alike 
|t  separate  points. 
