﻿Rydberg: 
  Notes 
  on 
  Rosaceae 
  51 
  

  

  name 
  for 
  the 
  Swamp 
  Rose, 
  or 
  what 
  has 
  usually 
  been 
  known 
  as 
  

   R. 
  Carolina^ 
  is 
  thus 
  i?. 
  palustris 
  Marsh., 
  a 
  very 
  appropriate 
  name. 
  

   The 
  synonymy 
  of 
  this 
  species 
  is 
  much 
  involved; 
  see 
  North 
  

   American 
  Flora. 
  

  

  West 
  of 
  Lake 
  Michigan, 
  especially 
  in 
  the 
  Green 
  Bay 
  region, 
  

   R. 
  palustris 
  has 
  more 
  pubescent, 
  smaller, 
  and 
  less 
  bluish 
  green 
  

   leaflets. 
  This 
  form 
  may 
  represent 
  a 
  distinct 
  species 
  or 
  geograph- 
  

   ical 
  variety. 
  

  

  II. 
  Rosa 
  dasistema 
  Raf. 
  

  

  f 
  

  

  Some 
  specimens, 
  collected 
  especially 
  in 
  Missouri 
  and 
  named 
  

  

  4 
  

  

  R. 
  Carolina, 
  differ 
  from 
  R. 
  palustris 
  in 
  having 
  a 
  smaller 
  hypanthium 
  

  

  + 
  

  

  deddedly 
  pear-shaped 
  and 
  acute 
  at 
  the 
  base, 
  instead 
  of 
  depressed- 
  

   globose 
  and 
  rounded 
  below, 
  and 
  also 
  in 
  having 
  fewer 
  leaflets, 
  

   usually 
  five, 
  rarely 
  seven, 
  instead 
  of 
  seven 
  or 
  nine. 
  These 
  speci- 
  

   mens 
  agree 
  closely 
  with 
  the 
  description 
  of 
  R. 
  dasistema 
  Raf. 
  

   The 
  type 
  of 
  the 
  latter 
  was 
  from 
  Indiana, 
  from 
  which 
  state 
  I 
  have 
  

   seen 
  no 
  specimens. 
  This 
  fact, 
  however, 
  does 
  not 
  disprove 
  the 
  

  

  identity 
  of 
  the 
  Missouri 
  specimens 
  with 
  Rafinesque*s 
  species^ 
  

   These 
  specimens 
  are: 
  

  

  Missouri: 
  Paw 
  Paw 
  Junction, 
  Bush 
  226; 
  Kennett, 
  Trelease 
  

   ///; 
  Campbell, 
  Bush 
  log, 
  6214; 
  Butte 
  County, 
  Bush 
  3688. 
  

  

  12, 
  Rosa 
  gemella 
  Willd. 
  

  

  Although 
  this 
  has 
  usually 
  been 
  regarded 
  as 
  a 
  synonym 
  of 
  

  

  R. 
  blanda, 
  the 
  description 
  shows 
  that 
  it 
  can 
  not 
  be 
  that 
  species- 
  

   The 
  description 
  of 
  the 
  prickles 
  of 
  R. 
  gemella 
  (slender 
  and 
  curved) 
  

   would 
  exclude 
  it 
  from 
  R. 
  hlanda. 
  There 
  is, 
  however, 
  a 
  plant, 
  

   though 
  rather 
  rare, 
  which 
  answers 
  Willdenow's 
  description. 
  

   It 
  is 
  related 
  to 
  R. 
  Carolina, 
  but 
  difi^ers 
  in 
  the 
  curved 
  prickles, 
  the 
  

  

  corymbose 
  inflorescence, 
  and 
  the 
  entire 
  sepals. 
  It 
  may 
  be 
  

   confused 
  also 
  with 
  R. 
  virginiana 
  and 
  R. 
  palustris. 
  From 
  the 
  former 
  

   it 
  differs 
  in 
  the 
  thin 
  dull 
  leaflets, 
  the 
  slender 
  more 
  curved 
  prickles 
  

  

  more 
  

  

  and 
  the 
  low 
  habit. 
  

  

  palust 
  

  

  but 
  are 
  much 
  weaker, 
  while 
  its 
  leaflets 
  are 
  broad 
  and 
  rather 
  short 
  

  

  m 
  ■ 
  

  

  coarsely 
  serrate. 
  The 
  following 
  specimens 
  belong 
  here: 
  

  

  more 
  

  

  Yarmouth, 
  Howe 
  

  

  Massachusetts: 
  Ipswich, 
  Nash, 
  Morong; 
  Provincetown, 
  HoU 
  

  

  lick 
  . 
  

  

  