﻿54 
  BOTANICAL 
  GAZETTE. 
  [ 
  March, 
  

  

  2. 
  S. 
  MYRTiLLiFOLiA 
  Anders. 
  Sal. 
  Bor.-Amer. 
  28. 
  S. 
  

  

  JVoV(s 
  

  

  160, 
  and 
  DC. 
  Prod 
  

  

  253, 
  mainly. 
  S. 
  myrsinites 
  Hook. 
  FL 
  Bor.-Amer. 
  2. 
  151 
  

  

  mai 
  

  

  'V 
  ^ 
  preuy 
  juue 
  wniow 
  lound 
  througnout 
  the 
  region 
  

   ■of 
  the 
  Canadian 
  lakes 
  from 
  north 
  of 
  Lake 
  of 
  the 
  Woods 
  

  

  (D 
  

  

  **Common 
  on 
  

  

  the 
  portage 
  of 
  the 
  Grand 
  Rapids 
  of 
  the 
  Saskatchewan, 
  near 
  

   Lake 
  Winnipeg'^ 
  (Douglas), 
  Athabasca 
  Lake 
  and 
  river 
  

   (Macoun), 
  extending 
  westward 
  along 
  streams 
  and 
  in 
  low 
  

   situations 
  to 
  the 
  eastern 
  foot-hills 
  of 
  the 
  Rocky 
  Mountains. 
  

   After 
  a 
  careful 
  examination 
  of 
  the 
  collections 
  of 
  Richardson, 
  

   Douglas 
  and 
  others, 
  upon 
  which 
  Andersson 
  founded 
  his 
  S. 
  

   •Novae-Anglian, 
  I 
  am 
  satisfied 
  that 
  (with 
  the 
  exception 
  of 
  

   Drummond's 
  Rocky 
  Mt. 
  specimens) 
  all 
  belong 
  to 
  one 
  

   species, 
  for 
  which 
  S 
  myrtillifolia 
  is 
  the 
  oldest 
  (and 
  best) 
  name. 
  

   It 
  is 
  useless 
  and 
  confusing 
  to 
  attempt 
  to 
  distinguish 
  as 
  varie- 
  

   ties 
  the 
  different 
  forms 
  due 
  to 
  varying 
  conditions 
  of 
  locality 
  

   and 
  exposure. 
  Andersson 
  himself, 
  in 
  the 
  case 
  of 
  Richard- 
  

   son 
  s 
  collections, 
  where 
  aments 
  and 
  leaves 
  of 
  the 
  same 
  plant 
  

   are 
  under 
  separate 
  tickets, 
  has 
  referred 
  the 
  aments 
  to 
  one 
  of 
  

   his 
  varieties 
  and 
  the 
  leaves 
  to 
  another. 
  The 
  species 
  is 
  one 
  

   ol 
  the 
  plain 
  from 
  Lake 
  Winnipeg 
  north 
  to 
  the 
  Arctic 
  circle. 
  

   Urummond 
  s 
  specimens, 
  mentioned 
  above, 
  are 
  more 
  robust 
  

   every 
  way 
  with 
  thick, 
  white-woollv 
  staminate 
  aments, 
  fertile 
  

   immature, 
  leaves 
  wanting, 
  and 
  can 
  not 
  be 
  safely 
  determined, 
  

   though 
  most 
  likely 
  belonging 
  to 
  S. 
  monticola 
  Bebb. 
  

   , 
  3. 
  .^. 
  ARBUscuLoiDEs 
  Anders. 
  Monog. 
  147, 
  fig. 
  8r. 
  S. 
  

   W/.,;^a 
  Anders. 
  DC. 
  Prod. 
  16. 
  2. 
  248. 
  S. 
  aVui/folia 
  Hook. 
  

   Fl 
  Bor.-Amer. 
  2. 
  150. 
  Arctic 
  America 
  (Mieschring 
  and 
  

  

  Rae). 
  Mackenzies 
  R 
  

  

  Rocky 
  Mount 
  

  

  M 
  

  

  The 
  sessile 
  cap- 
  

  

  '^J^r/^1^ 
  ^'^ 
  ^'^^ 
  ^^^^^' 
  ^^^g^ 
  emarginate 
  gland 
  and 
  small 
  

   serrulate 
  leaves 
  constitute 
  a 
  combination 
  of 
  characters 
  so 
  ab- 
  

   solutely 
  unique 
  among 
  American 
  willows 
  as 
  to 
  leave 
  no 
  

   doubt 
  concerning 
  the 
  plant 
  Andersson 
  had 
  in 
  view, 
  nor 
  any 
  

  

  conn 
  tir'^ 
  '-^'^ 
  f'f'^ 
  ^^^^ 
  S. 
  arbuscula. 
  Prof. 
  Ma- 
  

   W^ 
  ' 
  r 
  ° 
  examined 
  Andersson's 
  tvpes 
  for 
  me 
  in 
  the 
  KeW 
  

  

  Rae 
  

  

  Miesch 
  

  

  rpqnprt5.r«ixr 
  +1, 
  1 
  Vr, 
  r 
  "" 
  "leans 
  oi 
  teJJing 
  to 
  wnica, 
  

   [v 
  two 
  or 
  ,^' 
  ' 
  ^^^-^ 
  ^'^^"^ 
  ' 
  ^^^' 
  ^hey 
  compris'e 
  apparent- 
  

   no- 
  Terv 
  Z 
  'P^'f' 
  '^^.^^^ 
  ^^^y 
  imperfect, 
  a 
  few 
  bear- 
  

   thfs 
  tha7tllT? 
  ''f'f'-^ 
  }' 
  ^^"^^ 
  appear 
  p;obable 
  from 
  

  

  belon 
  Is 
  o 
  ,n-l''^i' 
  T^ 
  ''^P^'""' 
  ^^^''"^ 
  ^y 
  Andersson, 
  

   belongs 
  to 
  some 
  other 
  plant 
  than 
  that 
  from 
  which 
  was 
  de- 
  

  

  