﻿492 
  

  

  Smith: 
  Studies 
  in 
  the 
  gexus 
  Lufixus 
  

  

  coming 
  suberect 
  as 
  the 
  petals 
  wither 
  and 
  the 
  ovaries 
  enlarge; 
  

  

  petals 
  bright 
  

  

  pale 
  

  

  defined 
  

  

  banner 
  suborbicular 
  with 
  

  

  ^ptly 
  

  

  naked, 
  at 
  the 
  free 
  edges' 
  below, 
  naked 
  or 
  sparsely 
  ciliate 
  above; 
  

   pods 
  ascending, 
  about 
  six-ovuled; 
  seeds 
  some 
  2-3 
  mm. 
  long, 
  

   angled, 
  dotted 
  or 
  marbled 
  on 
  a 
  pale 
  ground. 
  

  

  Fig. 
  54. 
  LupiNus 
  

  

  / 
  

  

  I. 
  C. 
  F, 
  Baker 
  5357 
  (CPS) 
  ; 
  2, 
  J. 
  Grinnell 
  

  

  (US 
  614572); 
  3. 
  F. 
  S. 
  Daggert 
  (G. 
  L. 
  Moxley); 
  4. 
  H. 
  M. 
  Hall 
  5759 
  (UC) 
  ; 
  

  

  5- 
  /. 
  Tideslrom 
  867 
  (US). 
  

  

  After 
  much 
  deliberation, 
  I 
  am 
  still 
  unable 
  to 
  accept 
  Dr. 
  

   Davidson's 
  L. 
  siihhirsutus 
  as 
  of 
  even 
  varietal 
  rank. 
  That 
  the 
  keel, 
  

   as 
  stated 
  by 
  him, 
  is 
  really 
  non-ciliate, 
  has 
  been 
  verified 
  by 
  my 
  

   dissection 
  of 
  flowers 
  from 
  a 
  type-duplicate 
  specimen 
  in 
  the 
  col- 
  

   lection 
  of 
  Mr. 
  Geo. 
  L. 
  Moxley, 
  of 
  Los 
  Angeles. 
  However, 
  the 
  

   keel 
  is 
  normally 
  ciliate 
  in 
  Dr. 
  Hall's 
  specimens 
  (UC) 
  from 
  Palm 
  

  

  Springs, 
  the 
  type 
  locality 
  of 
  L. 
  snbhirsutiis. 
  

  

  M 
  

  

  is 
  quite 
  variable 
  at 
  Palm 
  Springs 
  is 
  shown 
  by 
  the 
  fact 
  that, 
  before 
  

   Dr. 
  Davidson's 
  description 
  was 
  published, 
  I 
  had 
  listed 
  certain 
  

   specimens 
  from 
  there 
  as 
  var. 
  arizoniciis, 
  and 
  others 
  as 
  var. 
  har- 
  

   hatidiis. 
  The 
  distinction 
  as 
  to 
  nodose 
  and 
  non 
  -nodose 
  hairs 
  I 
  

  

  4. 
  

  

  I 
  have 
  not 
  yet 
  been 
  able 
  to 
  appreciate. 
  

  

  As 
  indicated 
  further 
  by 
  key, 
  the 
  varieties 
  here 
  recognized 
  are 
  

   distinguished 
  by 
  stem 
  and 
  leaflet 
  characters, 
  the 
  often 
  evident 
  

   differences 
  in 
  amount 
  and 
  length 
  of 
  the 
  spreading 
  hairs 
  being 
  

   rejected 
  as 
  quite 
  unsatisfactory 
  for 
  diagnostic 
  use 
  within 
  the 
  limits 
  

   of 
  this 
  species. 
  Each 
  of 
  these 
  varieties 
  could 
  in 
  turn 
  be 
  subdivided 
  

  

  