﻿Vol. 
  53.] 
  SIGILLARIA 
  AND 
  GLOSSOPTEEIS 
  IN 
  SOUTH 
  AFKICA. 
  317 
  

  

  from 
  different 
  horizons 
  in 
  the 
  Gondwana 
  Series 
  of 
  India 
  have 
  in 
  

   several 
  instances 
  formed 
  the 
  types 
  of 
  new 
  species 
  founded 
  on 
  quite 
  

   inadequate 
  grounds. 
  It 
  requires 
  but 
  a 
  brief 
  examination 
  of 
  Feist- 
  

   mantel's 
  figures 
  and 
  some 
  acquaintance 
  with 
  the 
  commoner 
  forms 
  of 
  

   Glossopteris 
  to 
  confirm 
  this 
  statement. 
  Zeiller 
  has 
  recently 
  shown 
  

   good 
  reasons 
  for 
  identifying 
  Feistmantel's 
  G. 
  communis 
  with 
  

   G. 
  indica, 
  Schimp. 
  1 
  It 
  is 
  not 
  proposed 
  at 
  present 
  to 
  attempt 
  a 
  

   general 
  revision 
  of 
  the 
  specific 
  determinations 
  of 
  tbis 
  Southern 
  

   Hemisphere 
  genus 
  of 
  ferns, 
  but 
  to 
  demonstrate 
  the 
  extreme 
  diffi- 
  

   culty 
  of 
  drawing 
  satisfactory 
  distinctions 
  between 
  such 
  forms 
  as 
  are 
  

   included 
  under 
  G. 
  Broivniana, 
  G. 
  indica, 
  and 
  G. 
  angustifolia. 
  In 
  

   1861 
  Bunbury 
  2 
  figured 
  and 
  described 
  several 
  specimens 
  of 
  Glosso- 
  

   pteris 
  from 
  the 
  neigbbourhood 
  of 
  Nagpur, 
  Central 
  India, 
  and 
  insti- 
  

   tuted 
  some 
  new 
  specific 
  names. 
  The 
  type-specimens 
  of 
  Bunbury 
  

   are 
  now 
  in 
  the 
  Museum 
  of 
  the 
  Geological 
  Society, 
  and 
  in 
  the 
  

   Bunbury 
  Collection 
  in 
  the 
  Botanical 
  Museum, 
  Cambridge. 
  It 
  is 
  

   difficult, 
  or 
  indeed 
  impossible, 
  to 
  admit 
  the 
  existence 
  of 
  adequate 
  

   grounds 
  for 
  some 
  of 
  Bunbury's 
  species, 
  and 
  he 
  himself 
  calls 
  

   attention 
  to 
  the 
  variation 
  in 
  the 
  direction 
  and 
  number 
  of 
  the 
  veins 
  ; 
  

   he 
  expresses 
  the 
  opinion 
  that 
  no 
  satisfactory 
  specific 
  distinction 
  can 
  

   be 
  drawn 
  between 
  the 
  Australian 
  and 
  Indian 
  forms 
  of 
  G. 
  Broivn- 
  

   iana. 
  Tenison- 
  Woods, 
  3 
  in 
  speaking 
  of 
  the 
  variation 
  in 
  venation- 
  

   characters, 
  admits 
  that 
  the 
  distinctions 
  drawn 
  between 
  many 
  of 
  

   the 
  fossil 
  leaves 
  are 
  not 
  such 
  as 
  would 
  rank 
  as 
  specific 
  features 
  

   in 
  recent 
  fern-fronds. 
  In 
  Zeiller's 
  interesting 
  paper 
  on 
  African 
  

   plants, 
  already 
  alluded 
  to, 
  attention 
  is 
  called 
  to 
  the 
  striking 
  variation 
  

   in 
  form 
  and 
  venation 
  exhibited 
  by 
  the 
  6rZossop?ms-leaves. 
  The 
  

   drawings 
  which 
  he 
  gives 
  of 
  Brougniart's 
  type-specimens 
  illustrate 
  

   in 
  a 
  marked 
  degree 
  the 
  close 
  agreement 
  as 
  regards 
  the 
  venation 
  in 
  

   G. 
  Browniana, 
  G. 
  indica, 
  and 
  G. 
  angustifolia. 
  In 
  one 
  of 
  Zeiller's 
  

   specimens 
  the 
  veins 
  on 
  the 
  two 
  sides 
  of 
  the 
  midrib 
  are 
  by 
  no 
  means 
  

   identical, 
  but 
  exhibit 
  a 
  striking 
  difference 
  as 
  regards 
  the 
  angle 
  and 
  

   direction. 
  4 
  While 
  endeavouring 
  to 
  avoid 
  the 
  dangerous 
  and 
  un- 
  

   scientific 
  practice 
  of 
  needlessly 
  multiplying 
  specific 
  names, 
  we 
  must 
  

   be 
  careful 
  to 
  bear 
  in 
  mind 
  the 
  possibility 
  of 
  carrying 
  too 
  far 
  the 
  

   system 
  of 
  linking 
  together 
  distinct 
  types 
  by 
  a 
  long 
  series 
  of 
  inter- 
  

   mediate 
  forms. 
  It 
  is 
  no 
  easy 
  matter 
  to 
  decide 
  as 
  to 
  the 
  best 
  course 
  to 
  

   adopt 
  with 
  regard 
  to 
  G. 
  Browniana, 
  G. 
  indica, 
  and 
  G. 
  angustifolia-, 
  

   but 
  my 
  impression 
  is 
  that 
  these 
  three 
  names 
  cannot 
  be 
  maintained 
  as 
  

   standing 
  for 
  three 
  well-marked 
  species 
  of 
  Glossopteris. 
  Attention 
  has 
  

   been 
  called 
  by 
  several 
  writers 
  to 
  the 
  variations 
  in 
  both 
  form 
  and 
  

   venation, 
  and 
  an 
  examination 
  of 
  similar 
  single 
  fronds 
  of 
  recent 
  ferns 
  

   seems 
  to 
  demonstrate 
  the 
  want 
  of 
  a 
  definitely 
  fixed 
  form 
  among 
  

   living 
  species. 
  The 
  test 
  of 
  size 
  is 
  always 
  dangerous, 
  and 
  cannot 
  as 
  

   a 
  rule 
  be 
  regarded 
  as 
  a 
  taxouomic 
  character 
  of 
  much 
  value. 
  From 
  

   the 
  widespread 
  occurrence 
  in 
  Australia, 
  Africa, 
  and 
  India 
  of 
  layers 
  

  

  1 
  Bull. 
  Soc. 
  geol. 
  France, 
  ser. 
  3, 
  vol. 
  xxiv. 
  (1896) 
  p. 
  368. 
  

  

  2 
  Quart. 
  Journ. 
  Geol. 
  Soc. 
  vol. 
  xvii. 
  (1861) 
  p. 
  325. 
  

  

  3 
  Proc. 
  Linn. 
  Soc. 
  N. 
  S. 
  Wales, 
  vol. 
  viii. 
  (1883) 
  p. 
  122. 
  

  

  4 
  Op. 
  supra 
  cit. 
  pi. 
  xvii. 
  fig. 
  2. 
  

  

  