ROSA INVOLUTA VAR. NICHOLSONII 49 
Ase the same station near emia in Central France. Two o 
some are hooked and stout, but less so and only curved or declining 
in others. The minor arma ture is very abundant, and in those 
and 
all its armature weak and slender. rer soon 7, small a ed ral 
he 
densely glandular. Petioles glabrous or pore. sO, inal ‘glan- 
dular, with many very small prickles or acicles. Stipules usually 
broad, pee glandular on back, auricles very variable in form, 
- Somewhat connivent, a good deal Ses — en ular on back, 
though “vite ciliate, persisting till at least September. Styles 
canno: 
There is no doubt about this being a pimpinellifolia x rubign- 
nosa hybrid, the typical form of which has not been found in 
Britain, but — — near it have been gat iad) at Boxley 
Warren, E. Kent, and by Mr. Barley. “ Caputh, E. Perth. These 
differ primarily in 1 thei glandular ——— — fruit, and 
Sepals, which lat so are more entir e armature is more 
completely lak, like that of R. pimpineliolia The flowers of 
the Boxley plant are rose, not white as in the type. I do not 
know the een of the a specimens. 
Though these may be sufficient pion unds for giving a varietal 
name to our British form, I hesitate to do re Great Sana 
must be expected in hybrids, and the naming of several differe 
forms has already given rise to great confusion in the Seapine 
folia x tomentosa group. 
A INVOLUTA var. NicHOLsoNII 
Crépin in rg de la Soc. Roy. de Bot. de Belg. xxi. p. 119 (1882). 
‘* Leaflets of medium or rather large size, broadly oval ae pee 
compound. Outer sepals with 1 . “4 appendages wers two. 
Branches and branchlets densely setige 
I have seen no specimen of thia, 80 can only supplement 
JOURNAL OF Botany, May, 1910. [SupPLEMENT. ] é 
