114 ON RUBUS IDZXUS. 
But, on the other hand, the North American Rub? themselves appear 
to trace their origin from Japan and the north-east of ‘Asia, whence 
of Asia no Rubi quite identical with A forms, 
though some growing in Japan have so close a relationship to them 
that they may indeed be considered as modified forms. This is 
particularly the case with the simple-leaved Rubi, from which forms 
those with divided leaves have been probably developed (comp. Focke, 
** Die synthetische Methode in der Systematik ” in the ‘‘ Jenaischen 
Zeitschrift fiir Med. und Naturwiss.,” v., p. 107). Thus the Japanese 
species, 2. trifidus, Thunb., R. palmatus, Thunb., and R. Wright, 
A.Gr., very closely resemble R. odoratus, L., and R. Nutkanus, Mog.; 
though I have had no opportunity of learning whether their bark 
scales off from the stem and their fruit. separates from the receptacle. 
On the other hand, there are in Japan, besides R. Jdzus, L., many other 
forms—e.g., R. Coreanus, Miq., R. Thunbergii, Sieb. et Zuce., R. 
Oldhamii, Miq., and particularly R. phenicolasius, Maxwez., and R. 
strigosus, Michx., which much resemble our speci 
pe ch &. Ideus, L., is descended, or 
perhaps to be the primitive form itself. At present it is impossible to 
Rd 
in many other respects so nearly approach R. glandulosus, Bell., that it 
does not seem improbable to me that this species derived its origi 
from some one among them. Perhaps such a return has become con- 
such forms are common both in apan andin North America. 
circumstances now related seem to me to render it very 
All the ci 
probable that the species in question, together with its relations, has 
PUT 0 4 Bm PUNE BE EERE Ge G24 Pk )re TAK omy rm ene ee eae Cae ae 
* For some later remarks of Focke upon R. Leesii, see Oesterr, Zeitsch. 
1870, trans. in Journ, Bot. x, (1872.), p.26-—[Ed, Jews Bett” 
