FLORA VITIENSIS. 111 
to augment the number still further by two from Viti (S. Katakata and Graffei), and one from the New 
Hebrides (S. Maegillivrayi, sp. nov., Seem. mss. in Herb. Mus. Brit.; arbuscula; ramis pubescentibus ; 
foliis oppositis ovatis v. ovalibus acuminatis basi acutis, calloso-serratis utrinque glabris, insigniter pen- 
ninerviis, nervis subrectis; paniculis folia excedentibus; floribus 4-meris, diplostemonibus ; squamulis 
hypogynis oblongis truncatis basi connatis, intus barbatis; folliculis ignot.—Aneitum, New Hebrides 
(M‘Gillivray! n. 59). In habit resembling S. Samoense and Katakata). 
l. S. Vitiense, A. Gray, Bot. Wilkes, p. 669. t. 83; glabrum ; foliis oppositis et verticillatis 
obovato-ellipticis oblongisve obtusis basi attenuatis integerrimis paucivenosis paniculas excedentibus ; 
froribus 4-5-6-meris diplostemonibus ; squamulis hypogynis cuneatis v. oblongis truncatis v. emargi- 
natis glabris; folliculis 4 monospermis; semine superne alato.—At Bua Bay and Macuata coast of 
Vanua Levu (U. S. Expl. Exped.) ; Viti Levu (Dr. Greffe! n. 16, ex parte). 
2. S. Greeffei, (sp. nov.) Seem. ; glabrum; foliis oppositis ellipticis acuminatis basi acutis inte- 
gerrimis v. versus apicem obscure denticulatis insigniter penninerviis ; paniculis folia multo exce- 
dentibus; floribus ignotis; folliculis 4; czet. ignot.—Buke Levu, Kadavu (Greffe! n. 16, ex parte). 
Discovered by Dr. Greffe, whose specimen was kindly sent to me by Dr. F. Müller, and nearest to 
S. Vitiense, from which it differs by the shape of the leaves, and very large panicles. I have only seen 
one fruiting specimen, all the follicules having shed their seed. Leaves on petioles 3-1 inch long; blade 
23-8 inches long, 1-1} inch broad, on the upper side more prominently feather-veined than S. Vitiense ; 
veins from 7-9 on each side of the midrib. Panicles twice or thrice as long as the leaves. 
3. S. Katakata, (sp. nov.) Seem. in Bonplandia, vol. x. p. 36. (Tab. XVII.) ; arborea; ramis 
petiolis pauiculisque puberulis; foliis oppositis ovatis v. ovato-oblongis acuminatis basi rotundatis v. 
obtusis minute denticwlatis insigniter penninerviis, nervis curvatis, utrinque costis nervisque ex- 
ceptis, glabris; paniculis folia excedentibus; floribus 4-meris diplostemonibus; squamulis hypogynis 
obovato-oblongis emarginatis intus barbatis; folliculis 4 extus puberulis; cet. ignot.—Nomen 
vernac. Vitiense, * Katakata."—Kadavu (Seemann! n. 196). 
A. middle-sized tree; petioles 1-11 inch long; blade of leaf 4-5 inches long, 23-3 inches broad. 
Primary veins more curved than in S. Macgillivrayi. Bracts lanceolate, opposite. Flowers polygamous. 
My specimens have male flowers only, whilst Milne's fruiting specimens, from which Fig. 5 and 6 of our 
Platé are copied, have shed the seeds. ; 
Expnanation OF PLATE XVIL —Fig. 1, a male flower; 2, the same, with the calyx removed ; 3 and 
4, front and back view of one of the eight hypogynous scales, alternating with the stamens; 5, branchlet 
with ripe fruit; 6, the folliculi after shedding their seed :—all, with exception of Fig. 5, magnified. 
Orvo XLVIII. HEDERACEJE. 
It is a matter of importance to ascertain the æstivation of all the plants regarded by us as Umbelli- 
feræ and Araliaceæ, because, for systematic purposes, most useful characters are to be found in it; and as 
there is a vast number of species to be examined, it is desirable there should be a good many observers. 
I am more and more convinced that the only clear line between Umbelliferæ and Araliaceæ, or, as I should 
prefer calling the latter, Hederaceæ, can be drawn by relying upon the characters derivable from the 
æstivation of the corolla. Not one of the distinctive characters usually assigned to the two Orders holds 
good, and in elementary books this state of things is extremely puzzling to the student. _ If, for instance, a 
beginner wished to find Hydrocotyle and Crithmum by means of the analytical key given in one of the 
leading British Floras, he could never hope to succeed. After getting to the neighbourhood of the Orders 
in which they are placed in that work, he would meet with the following enigma :— i 
; * 83. Petals imbricate in bud. 34. 
Petals valvate in bud. 35." 1 ar. : 
Of course, having ascertained the petals to be valvate in bud, he would continue his search by turning to 
“35,” and there find,— : 5 
“Fruit a berry. Styles several. Leaves alternate. Araliaceæ. 2: 
Now as neither Hydrocotyle nor Crithmum have a berried fruit or more than two styles, his search for the 
two genera by means of this analytical key would have come to an end. I have quoted this example, out of 
many, because it came ready to hand, and it illustrates the practical inconvenience arising from our neg- 
lecting the zestivation. ‘The conspectus of the calycifloral Orders given in Hooker and Arnott’s ‘ British 
Flora,’ places the student in the same dilemma. 
