233. 
ditter from those of ?. amurense in being dull on the upper surface 
and having glabrous margins. Nor does there appear to be any 
marked difference between sachalinense and amurense in the 
colour of the branchlets, It seems desirable, therefore, to treat 
the cultivated sachalinense as a variety of amurense. 
_ (2) Sargent’s identification of the cultivated sachalinense with 
P. amurense var. sachalinense, Fr. Schmidt, may, on the other 
hand, be accepted. Both trees differ from typical amurense in 
having a non-corky bark, and relatively broader leaflets. 
(3) The identification with var. sachalinense of the Japanese 
wee, hitherto known as P. amurense, is open to question. The 
leaflets of wild specimens from Japan certainly resemble those of 
var. sachalinense in shape, but the labels -aftord no information as 
to the nature of the bark, which furnishes the most important 
character of the variety. Furthermore, the figures of the bark of 
Japanese amurense given by Shirasawa, Ic. Ess. For. Jap. ii. t. 33, 
ff. 27, 28 agree, in the writer’s opinion, with typical amurense 
rather than with var. sachalinense. The question is recommended 
to the attention of Japanese botanists, as it can be elucidated most 
satisfactorily by a study of the living trees. 
It is, moreover, complicated by the existence in Japan of 
another variety or form of P. amurense, which Dode has described 
as an independent species, P. Lavallet, from a tree of Japanese 
origin cultivated in the Arboretum at Segrez.* This differs from 
P. amurense, according to Dode, in the leaflets, which are lanceo- 
late-ovate instead of lanceolate, and-have the nerves of the lower 
surface clothed with stiff white hairs. It has been in cultivation 
in England for about 40 years, and has hitherto passed under the 
names P. japonicum and P. sachalinense. Apart from the in- 
dumentum, it seems to differ in no respect from the usual 
Japanese form of P. amurense, and it may therefore be named 
P. amurense var. Lavallei. 
A specimen in the Kew Herbarium, gathered by Tschonoski in 
1865 in subalpine woods in the province of Nambu, Nippon, agrees 
with var. Lavallei. Sargent referred the corresponding specimen 
in the Gray Herbarium to P. japonicum, from which the Kew 
specimen differs in the cuneate base of the leaflets, and the nature 
of the indumentum. 
Phellodendron japonicum, Maxim., appears to be confined to 
base and softly pubescent on the lower surface with rather curly 
hairs, and the larger fruits. 
A third species which is in cultivation at Kew is P. chinense, 
Schneider, of which P. sinense, Dode, is a synonym. This is a 
native of Hupeh, and may easily be recognised by its very com- 
pact inflorescence. rE 
Phellodendron macrophyllum, Dode, and P. Fargesit, Dode, are 
known to the writer from description only, and do not appear to 
be in cultivation. The former has very large leaflets, up to 20 cm. 
# Bull. Soc. Bot. France, lv. p. 648 (1909). 
