291 
approach  to  specimens  from  Catsworth,  Hants.,  which 
Dingier  refers  confidently  to  a  form  of  R.  caryophyllacea 
Ghr.  (non  Bess.),  to  which  he  has  given  the  name  of 
R.  tomentella,  var.  anonyma  (ined.  ?).  The  form  is  remark¬ 
able  for  the  glandular  development  on  the  branches,  and 
the  more  or  less  numerous  subfoliar  glands.  Dingier 
thinks  it  may  be  a  hybrid  with  R.  Eglanteria.  I  have 
seen  somewhat  similar  forms  from  near  Huntingdon  and 
from  Surrey.  I  fear  that  the  Wadenhoe,  and  possibly  the 
other  gatherings,  may  contain  a  mixture.  The  late  Mr. 
Ley  told  me  that  he  gathered  from  several  bushes  what  he 
believed  to  be  the  same  form.  I  have  in  consequence 
received  from  him  at  least  three  forms,  one  of  which 
belongs  certainly  to  the  Deseglisei  sub-group,  so  probably 
a  mixture  has  been  distributed  here  also. — A.H.W.-D. 
R.  canina  L.,  var.  insignis  Desegl.  (1)  Near  Waden¬ 
hoe,  Northants,  v.c.  82,  Aug.  8,  1910.  (2)  Near  Ruardean, 
W.  Glos.,  v.c.  84,  Sept.  8,  1910.  This  is  a  composite 
parcel. — A.  Ley.  My  specimen  from  Wadenhoe  (with 
glabrous  leaflets)  is  certainly  R.  insignis  Desegl.,  but  I 
understand  that  some  of  those  distributed  have  the 
leaflets  hairy  on  the  midribs,  so  that  they  obviously  cannot 
belong  here.  A.H.W.-D.  This  cannot  be  R.  insignis 
Desegl.,  whatever  may  be  the  worth  of  that  as  a  species 
or  variety.  Here  we  have  the  midribs  hairy,  though  the 
hairs  seem  to  be  partly,  if  not  wholly,  deciduous.  I 
suppose  it  must  be  put  as  a  var.  of  R.  dumetorum  Thuill., 
though  it  is  one  of  those  forms  which  show  that  the 
distinction  between  that  and  R.  canina  is  sometimes  verv 
slight  indeed. — W.B. 
R.  canina  L .,  var.  aspernata  (Desegl.).  Lyde  Green, 
Pucklechurch,  W.  Glos.,  v.c.  34,  June  27  and  Sept.  21, 
1910.  Ida  M.  Roper.  I  do  not  know  this  interesting 
rose ;  but  the  fruit  of  R.  aspernata  Desegl.  should  be 
ovoid,  whereas  these  are  globose.  Styles  villous ;  leaflets 
five,  small,  varying  from  lanceolate  to  orbicular-ovate, 
their  teeth  simple  or  slightly  compound,  mostly  gland- 
tipped.  The  general  appearance  is  that  of  a  R.  obtusifolia, 
with  glabrous  leaves  and  very  glandular-hispid  fruit, 
rather  than  of  R.  canina. — E.S.M.  This  agrees  best 
with  R.  aspernata  Desegl.,  but  the  armature  of  the 
peduncles  in  typical  specimens  is  stouter. — A.H.W.-D. 
