THE MONTANE FLORA OF FIJI. 209 
part of a ridge stretching to the sea. This part of the plain is intersected by 
the streams collected on the ridge. Groves of Kavika (Eugenia moluccensis), 
Leba (E. neurocalyx), and Ivi (Inocarpus edulis) occur on the banks, of 
course originally planted and now sheltering the banana plantations between 
the ramifications of the streams. The soil here seems chiefly soapstone, 
which bakes and cracks in the sun. Beyond this cultivation zone Eugenia 
rivularis and Pittosporum Brackenridgei, with Hoya australis twining up their 
stems, were flowering in the woods fringing the streams and characteristic 
of this Leeward country. 
On the lower slopes of the mountain, ÆRynchospora aurea abounded in 
marshy areas, and in some open places Yaka (Pachyrrhizus trilobus) and Lageno- 
spora Pickeringii 5 otherwise the reeds, sometimes over 6 ft. high, luxuriated 
up to the top. The native tracks run like tunnels through this growth, quite 
invisible to the ordinary eye and impassable in European clothes. The 
summit, 3000 feet in height, was the usual knife-edge type, broadening out 
in the highest part, where, the reeds (associated with Balantium stramineum 
and Gleichenia linearis) not being very high, a magnificent view was obtained 
over the open country and the forest-clad ranges behind. Fine wild lemon- 
trees fruiting abundantly were growing towards the edge, which was fringed 
by the trees of the soak areas on the mountain-side. In these tree-clad 
soaks the ground was quite parched underneath, and there was little under- 
growth beyond some ferns. The breadth of the stone-swept tract testified to 
the source of the stream and showed how the water must pour down from the 
summit in the rainy season. These trees could not possibly hold sufficient 
water to feed the innumerable streams which rise on the steep slopes of this 
Koro Levu ridge, and one must suppose that the luxuriant reed-growth, 
covering the ground thickly where trees cannot grow, not only binds the soil 
together, but acts as a reservoir for holding water. 
Near Tavua, the nearest coast-town, there is a great deal of sugar-cane 
cultivation on the alluvial flats of the Tavua river, which are apparently of 
old mangrove formation. About two miles off the road are some hot springs, 
to reach which you have to pass over an undulating ridge and furrow country, 
the furrows generally forming small savannas of the grotesque Pandanus 
odoratissimus. My pony fed greedily on the leaves of this plant, which, cut 
up and mixed with the refuse of molasses, is the staple food of the mules 
belonging to the Sugar Company. The hot springs formed tiny streams, 
covered by a thick yellow scum of algæ, probably bleached by the action of 
the water, and fringed so thickly with overhanging grasses and sedges, that 
I did not believe the guide, when he put his foot in and suddenly withdrew 
it to show the water was hot. The temperature, however, proved to be 59° C. 
The Algæ collected were Navicula viridis, Gomphonema intricatum, Nitzschia 
Palea, Phormidium tenue, P. laminosum, and P. luridum, all species previously 
known in similar situations. 
LINN. JOURN.—BOTANY, VOL. XXXIX. Q 
