NOTES ON CARRUTHERSIA AND VOACANGA. 201 
known, with the exception of the fruit, and must now figure as 
Carruthersia pilosa. wers are smaller, much more numerous, 
and arranged in ge inal pani and the leaves very 
e bundant material of Carruthersia scandens 
lected by Mr. Horne, now been received at Kew, with flowers 
) 
and fruit im siti, which enables me to give the following description 
of the fruit :—Follicles elongate, slender, terete, eae striate, 
and minutely puberulous, 4-6 in. long, 2-24 lin. radually 
tapering from middle to an sash point; seed pe we vell developed) 
linear, subcompressed, 5 lin. long, coma of copious hairs, 1-1} in. 
g M 
forests, Viti Levu. Common in many other parts of Fiji.” It has 
also been collected in ios Island of se ip Greeffe (no. 1587), 
There is a specimen sent by Mr. Horne, of w the genus is un- 
certain, as it only ses of a hei; a portion of stem, and a fruit. 
The fo llicles are very divaricate and much —— than mo the 
: es the 
mountain forests near Koro Suli, Witt L vu. Tai uice white, yields 
caoutchouc.” It may therefore prove diferent when more material 
comes to hand. 
The fruit i hae by Seemann has the structure of a 
Melodinus, and most likely belongs to a specimen gathered by him 
in the Island of Orelaa (n. 811), which he refers to M. sca 
Forst. It is . ever different from that Species, and may be 
described as follo 
Melodinus vide ensis, sp. nov. (M. scandens Seem. FI. Vit., 
p. 155, non Forst.). Climbing; stem elabrous, striate. Leaves ovate- 
lanceolate or oblong, obtuse; base attenuate to short siti 
glabrous. Panicle lax; sepals ovate, penis? corolla not w 
developed, but covered outside, as well as the pedicels, &c., as a 
close cinereous scurfy tomentum. Island of Ovalau, Seemaien n. $11, 
As this is the only spstias of Melodinus known from Fiji, and 
gathered Hy Seethitis it is probable the fruit figured as Carruthersa 
belongs to it, and has by some accident been transposed. M. scan- 
dens Forst. differs in the subsessile leaves with subcordate biapea. 
shorter pedicels, cit the broader, more obtuse se 
h 
ned 
present time. In the ‘ Gctirs Piaitarast™ ne is doubtfully ng tare 
abernemontana,—‘* Hue etiam verisimiliter referenda Voacan 
Thou. Gen. Nov. Madagasc. 10, arbor Madagascariensis.” e 
understood the oe rightly, =e described an additional species, 
Com 
Voacanga Dreget ey. (Comm. p. 189), but as no less than three 
ener e since been soit upon this plant, no wonder the 
genus should still pone ee tae The valuable collections 
lately received fro r have done much to clear up the 
om 
mystery. The inaabaaigabir Mr. Baron has collected specimens, 
both in flower and fruit, of what is unmistakably the _“ 
of Thouars, showing the structure to be identical Y ith that of 
