Panax. ] 
XXXIV. AllALIACEiE. 
103 
petiole 2-8 in. long, robust, with a short, broad, coriaceous, 2-lobed sheath. 
Umbels unisexual, terminal, very numerous, large ; peduncles 2-3 in. long, 
spreading ; rays very numerous, 1 in. long ; pedicels \ in. Flowers large. 
Calyx-margin sinuate. Fruit nearly orbicular, flattened, grooved on the faces, 
with 2 short recurved styles, connate to the middle. — Hook. Lond. Journ. 
Bot. ii. 421. t. 11. 
Common in forests throughout the islands, Banks and Solander, etc. Kermadec 
Islands, M‘ Gillivray. 
10. P. Sinclairii, Hook, f, n. sp. A small tree. Leaves on rather 
short petioles, 3-5-foliolate ; leaflets sessile, 1-2 in. long, very coriaceous, 
not glossy, obovate- or oblong-lanceolate, acute, sharply serrate, veins obscure ; 
petiole 1-H in. long, neither stipulate nor sheathing. Umbels small, uni- 
sexual, on short, terminal peduncles ^-1 in. long or less ; rays few ; pedicels 
very short. Fruit nearly orbicular, compressed, 2-celled, with 2 short recurved 
styles. 
Northern Island: Ruahitie Mountains, etc., Colenso. Auckland?, Herb. Sinclair. 
Sinclair sends both old specimens with normal 5-7-folio)ate serrate leaflets, and youug ones 
with piunatifid leaflets ; Colenso sends the pinnatifid-leaved state as the young of P. Colensoi. 
There are two species of Panax mentioned as natives of New Zealand, in Planchon’s 
‘ Hortus Douatensis ’ (p. 10), and which, being founded on garden specimens that have 
never flowered, are very doubtful: these are P. pentadactylum, Dene, and PI., from New 
Zealand ? (A. penlaphylla, Hort., A. quinquevulnera, Makoy), which Planchon suspects is a 
variety of P. crassifolium, with green not brown leaves, 5- rarely 3-foliolate, and the leaflets 
elegantly curved instead of being stiff and straight as in P. crassifolium and tridactyl um . 
The other is P. tridactylum, Dene, and PI. {A. triplujlla and A. trfoliata, Hort.), of which 
he says that this also is perhaps a variety of P. crassifolium, sometimes bearing only one leaflet, 
whilst the leaflets of P. crassifolium become ternate, especially towards the upper part of 
large trees. Professor Planchon is, however, in some misconception here, for the true P. 
crassifolium is green and not brown, both in a native state and cultivated, and has 3-foliolate 
leaves, except only at the tops of the older branches, where they become 1-foliolate and 
(losing their joint) simple. 
3. SCHEFFLERA, Forst. 
Trees. Leaves digitately 7-foliolate. Umbels racemed. — Flowers poly- 
gamous, not jointed on the top of the pedicel. Calyx-tube minutely 5- 
toothed. Petals 5, valvate. Stamens 5. Ovary 10-celled, with 10 short 
styles united into a cone to above the middle, their tips free. Fruit rather 
fleshy, 10-celled, 10-ribbed. 
I do not find that this old genus of Forster’s is retained in the rearrangement of the Order 
by Decaisne aud Planchon, who, however, I cannot suppose would leave it in Panax, from 
which it differs in habit, inflorescence, the 10-celled ovary, and the absence of any evident 
joint at the summit of the pedicel with the flower, whilst it differs still further from 
Aralia in the valvate petals, digitate leaves with regular serratures, aud 10 styles. 
1. S. digitata, Forst. — Aralia Schefflera, Spr. ; — FI. N. Z. i. 95. A 
small umbrageous tree or large shrub, with stout branches. Leaves on long 
I petioles, digitately 7-1 1-foliolate ; leaflets petiolate, 4-7 in. long, oblong- 
lanceolate, acuminate, membranous, sharply finely serrate; veins diverging, 
delicate ; petiole 3-7 in. long, terete, with a short 2-lobed sheath at the base. 
Umbels racemose on the branches of a very large, unisexual, compound, axil- 
lary panicle, small, in. diam., many-flowered ; rhachis of panicle 1 ft. and 
