164 CRAIB—HARDY SPECIES OF ENKIANTHUS. 
from the branchlet. Another feature is found in the nerves of 
the lower surface where only the lower part of the midrib is 
elevated. In flower the shrub may be easily recognised by the 
terminal umbels of few, white, urn-shaped flowers each with 
five very prominent nectaries. 
E. subsessilis (syn. E. nikoensis), which is not very common 
in cultivation, agrees with E. perulatus in having white urn- 
shaped flowers and in the pedicels in fruit being straight. The 
close, fine reticulation of the leaf affords a useful diagnostic 
mark, as does also the row of short white erect hairs along the 
midrib on the upper surface. The flowers again have not got 
the prominent nectaries, and they are arranged in a raceme 
and not in an umbel. 
As already indicated, the species of the campanulatus type 
may be grouped in two series: (1) the western series which 
may be referred to as the deflexus series after the name of the 
‘first described species, and (2) the eastern or true campanulatus 
series. The most reliable character for separating these two 
groups is found in the pistil, which is hairy in the former and 
glabrous in the latter. Besides this character we find that 
plants of the deflexus series are not as a rule so bushy as those 
of the true campanulatus series, that the former have larger, 
coarser leaves and larger flowers, and that the raceme is much 
more abbreviated. These characters, however, while holding 
for the majority are by no means absolute, e.g. as will be noted 
in due course, one of the species of the true campanulatus series 
is in habit very similar to the plants of the other series. 
The deflexus series is represented here by the well-known 
E. Himalayan plant and, of the more recent introductions, by 
two of Wilson’s Chinese plants, both of which have been referred 
by their collector to the Himalayan species, as also by some 
younger plants of still more recent introduction which have not 
yet flowered. In the absence of flowering material these latter 
must meanwhile be left out of consideration. 
Although I have referred to this series as the deflexus series 
I cannot at present follow Schneider and Wilson in adopting 
Griffith’s name deflexus for the Sikkim plant, which is the plant 
in cultivation. Griffith gave the name deflexus to his Bhutan 
plant, and of his original collection I have seen but poor dried 
material. Recent collections made in Bhutan by Mr. R. E. 
Cooper contain specimens of an Enkianthus which are exactly 
similar to Griffith’s plant, and since Griffith’s and Cooper’s 
material agrees exactly with nothing I have seen from Sikkim’ 
1 am forced to keep the two plants distinct and to refer to the 
Sikkim plant as E. himalaicus. : 
Again I cannot agree with Wilson iti referting his plants to 
