108 FLORA VITIENSIS. 
Orvo XLVII. SAXIFRAGACEZE. 
In 1836, when Endlicher published his list of South Sea Island plants, he was able to enumerate 
only four species of Savifragacee from tropical Polynesia, viz. Codia montana, Forst. (New Caledonia, 
Forster), Weinmannia parviflora, Forst. (Tahiti), Geissois racemosa, Labill. (New Caledonia), and Brous- 
saisia arguta, Gaud. (Sandwich Islands). Since then the number has so rapidly increased, that I am ac- 
quainted with no fewer than 38 species, distributed over 7 genera, viz. Geissois (5), Cunonia (5), Wein- 
mannia (8), Spireanthemum (6), Pancheria (7), Codia (5), and Broussaisia (2). I hold the two species 
of the latter genus sound ones. Broussaisia arguta, Gaud., A. Gray, Bot. Wilkes, t. 87, was -collected in 
Oahu (Macrae! in Mus. Brit.), B. pellucida, Gaud. Bonit. t. 9, in Hawaii by Deal, and also in Captain 
Cook’s second voyage, with the schedule “ Hydrangeoides. Frutex 9-pedalis; fl. pallide rubicundi; 
Hawaii in sylvis." 
I may here express my conviction that Sazifragacee and Crassulacee should be fused into one great 
Natural Order. At present they are kept apart merely by imaginary differences. The calyx is 5-sepalous 
in several Sawifragacee, as it usually is in Crassulacee. The hypogynous scales, upon which Lindley laid 
undue stress in Crassulacea, are present in Spireanthemum ; there is every possible degree of transition 
from apocarpous to syncarpous fruits; and there are inferior, semi-inferior, and free ovaries. - 
I. Geissois, Labill. Ser. Austr. Cal. p. 50. t. 50; Brongn. et Gris, in Ann. Sc. Nat. Ser. 5. 
t. i. p. 368. Calyx liber, 4-partitus, sepalis acutis margine crassis, intus sepe hirsutis, caducis, 
eestivatione valvatis. Corolla 0. Stamina 12-16, hypogyna. Ovarium oblongo-conicum, basi in 
‘discum annularem crenatum angustum expansum, 2-loculare. Ovula co, biseriata, ascendentia. 
Styli 2, graciles, basi connati; stigmata parva, acuta. Capsula elongata, styli basi simplici cuspidata, 
sed margine tumido cincta. Pili in tota planta pellucidi, rigidiusculi, quasi articulati, subulati, pruritum 
momentaneum excitantes.—Hab. in Tahiti, Huahine, etc."— Sol. Prim. Fl. Ins. Pacif. p. 336 (ined.). 
This plant is allied to C. ovifera, but perfectly distinct, the calyx being very different in the two species. 
There are besides in tropical Polynesia the following Cueurbitacece :— 
Sycios australis, Endl. Fl. Norfolk, p. 67 (1833): S. Fretensis, Hook. fil. in Lond. Journ. Bot. vol. vi. 
p (1847) ; Walp. Ann. vol.i p. 817: 8. angulata, Forst. Prodr. n. 363 (non Linn.); Hook. fil. Fl. 
: Zealand, vol. i. p. 72, ex parte.— Norfolk Island (Bauer), New Zealand (Banks and Solander! in Mus. 
Brit.), New South Wales (fide A. Gray). Forster and Hooker fil. have regarded this species as identical 
with S. angulatus of America, but the two seem to be quite distinct. A. Gray (Bot. Wilkes, p. 648) has 
already pointed out that the flower and fruit of S. australis are not larger than those of S. parviflorus, and 
less than half the size of those of S. angulatus. There are besides other distinctions. S. angulatusg Linn. 
is covered with long, floccose, often glandulose hair on the peduncles and fruit, its tendrils are 4- or nore 
gencrally 5-fid, and its fruit sparingly covered with spines, whilst S. australis is without the long floecose 
hair, has always 3-fid tendrils, and its fruit is densely covered with spines. Besides, the form of the leaf is 
different in the two. 
Sycios pachycarpus, Hook. et Arn. Bot. Beech. p. 83; A. Gray, Bot. Wilkes, p. 650, t. 80.—Oahu 
(Macrae! Lay and Collie, Gaudichaud), Maui, Sandwich Islands (U. S. Expl. Exped.). 
Sycios macrophyllus, A. Gray, Bot. Wilkes, p. 651, t. 81.— Hawaii, Sandwich Islands (U. S. Expl. Exped.). 
x PT DC A. Gray, Bot. Wilkes, p. 652, t. 82.— Hawaii, Sandwich Islands (Macrae! U.S. 
xpi. Exped.). : : 
2 Melothria Samoensis, A. Gray, Bot. Wilkes, p.641.—Samoan Islands (U. S. Expl. Exped.). 
Zehneria Baueriana, Endl. Fl. Norf. n. 126.— Norfolk Island (Bauer, fide Endl.). 
Guillemin (Zeph. Tait.) mentions a Cucurbitacea, of which only a single branch was collected by Ber- 
tero and Moerenhout, and which he refers, with a mark of doubt, to Trichosanthes. 1t is said to be called 
“ Patara” by the natives, and is stated to have palmate leaves, with seven large lanceolate leaflets; the 
flowers are unknown. Ellis (‘Polynesian Researches,’ vol. i. p. 360) says:—‘ Patara is a root growing wild 
in the valleys, in shape and taste resembling a potato, more than any other root found in Tahiti. It is 
highly farinaceous, though less nutritive than the Yam; the stem resembles the Woodbine or Convolvulus. 
The natives say the flower is small and white; I never saw one, for it is not cultivated, and but seldom 
sought, as the tuberous root is small, and more than two are seldom found attached to the same vine or 
stalk."— The Patara will probably be found to be no Cucurbitacea at all, but Dioscorea pentaphylla, Linn., 
whieh I find mentioned in Cuzent's list under the native name of * Paauara ;" but Patara is probably the 
correct name. The * Paauara” proper, of Tahiti, is, from all I ean learn, identical with Dioscorea aculeata. 
