52 BALFOUR—NEW SPECIES OF RHODODENDRON. 
1.8 cm. broad. Stamens 10 unequal shorter than corolla, 
longest about 3.2 cm. long with anther 3 mm. long, shortest 
about 2.2 cm. long with anther 2 mm. long; filaments yellow 
only slightly widened downwards, puberulous at the base. Disk 
glabrous. Gynaeceum about 3.4 cm. long shorter than corolla 
a little longer than stamens; ovary about 6 mm. long very 
narrow cylindric eglandular grooved truncate densely clad with 
ascending fasciate few-branched hairs with sharp points form- 
ing an adpressed tomentum and a tuft round base of style ; 
style floccose to above middle, eglandular. 
id W.-Yunnan. Western flank of the Tali Range. Lat. 
25° 40’ N. Alt. 10,000 ft. In open thickets. Shrub of 4-6 
ft. Flowers deep rose with crimson markings. G. Forrest. 
No. 13,736. May 1917. 
This plant has special botanical interest. Rh. neriiflorum, 
Franch. is its nearest ally, and Rh, dimitrum differs in the 
leaf-form—leaves tapered never rounded at the apex and the 
shape tending to oval; in the green not glaucous under surface 
of the leaf—Rh. dimitrum wanting the long epidermal wax- 
secreting papillae of Rh. neriiflorum; in the glabrous under- 
leaf midrib—not coated profusely with floccose hairs as in Rh. 
nertiflorum; in the paler coloured calyx-lobes abundantly 
spotted with crimson; in the puberulous corolla-tube and 
puberulous staminal filaments—these being glabrous in Rh. 
nertiflorum. Rh. dimitrum is a plant of the western flank of 
the Tali Range, a region from which apparently only a small 
number of plants have come to us. Delavay, to whom our first 
knowledge of Yunnan Rhododendrons is due, collected mainly 
on the eastern flank and in the area immediately to the north 
and north-east of Tali itself—about Langkiung. On the 
eastern flank of the Tali Range Rh. neriiflorum must be not 
uncommon. We have it from Delavay under Nos. 294, 2061, 
and from Forrest under Nos. 4140, 4144, 4147, 4164, 6766, 6780, 
11,617. The flora of the western flank shows differences from 
that of the eastern flank, as Forrest has acutely observed, but 
the differences are often not easily focussed. Here in Rh. 
dimitrum we have a species which illustrates modification of a 
phylum on opposite sides of the range, and the differential 
characters are easily recognised, although the determining 
factors are not yet apparent to us. On the Shweli-Salween 
divide occur plants which in the dried specimens collected by 
Forrest under Nos. 8939, 11,911, 11,921 do not offer marks 
sufficient to separate them from Rh. neriiflorum of the eastern 
flank of the Tali Range, yet they suggest differences which, if 
we knew more of the plants, we might appraise as of some dis- 
tinctive value. For the present we have to reckon them as 
