466 
APPENDIX  I. 
Draba  verna  L.  Key  to  subspecies,  from  Bony 
&  Foucaud’s  “  Flore  de  France,”  Yol.  II.,  p.  220,  (1895). 
Bouy  &  Foucaud’s  Key  to  Erophila  is  subjoined,  not 
as  solving  the  difficulties  of  classification,  but  only  as  a 
help  to  some  better  understanding. 
Mr.  E.  G.  Baker  writes  that  the  definite  separation 
of  I.  Bijidce  from  II.  Simplices  does  not  work  out  very 
well  in  the  case  of  E.  hirtella  Jord.,  of  which  Jordan 
says  “ pilis  patulis  scepius  bifurcatis .”  Also  that  Jordan’s 
E.  majuscula  has  leaves  “  oblong-obovate,”  and  De 
Candolle’s  E.  prcecox  leaves  “lanceolate,”  and  that  there 
are  similar  differences  in  the  number  of  seeds,  E.  steno- 
carpa  Jord.  having  “about  40,”  and  E.  hirtella  Jord. 
“  80—35.” 
1  Hairs  all  or  nearly  all  simple,  very  rarely  with 
bifid  hairs  intermixed  ;  silicules  elliptic  or  oblong  ; 
seeds  14 — 24.* 
Hairs  all  or  nearly  all  bifid,  some  trifid,  rarely  with 
simple  hairs  intermixed. 
2  Leaves  broadish,  ovatef  or  oblong-lanceolate,  spread¬ 
ing  horizontally  on  the  ground  ;  silicules  elliptic 
or  oblong,  little  or  not  at  all  attenuate  at  base. 
D.  glabrescens  Nob. 
Leaves  lanceolate,  erect  or  ascending ;  silicules 
oblong,  much  attenuate  at  base.  D.  hirtella  Nob. 
3  Silicules  ovate-suborbicular,  or  obovate-rotundate, 
very  obtuse  ;  seeds  8— 24. 
Silicules  of  a  different  form. 
*  The  number  of  seeds  refers  to  each  loculus  of  the  silicule. 
t  “  Ovale  ”  has  two  meanings : — 
1.  In  a  general  sense,  its  equivalent  is  “oval.” 
2.  In  a  botanical  sense,  its  equivalent  is  “  ovate.” 
To  translate  “ obovale ”  as  “  oboval  ”  would  be  meaningless;  and  therefore 
it  appears  better  to  adhere  to  “ovate”  and  “  obovate  ”  respectively,  although 
“ovate”  as  applied  to  the  silicule  of  an  Erophila  gives  rise  to  other  difficulties 
of  fact. 
I 
2 
3 
