only surpassed in beauty by the species of Stenogyne, which flower in the late 
winter months. Their large curved corollas, which are borne in large whorls, 
vary in shades from deep magenta to crimson, pink, yellow and pure white, in- 
terlacing trees or gracefully festooning branches, or, as is often the case, 
forming dense carpets covering the ground in small spots to the exclusion of 
other plants. The handsomest species are St. Kamehamehae and St. longiflora. 
Another very important feature of the vegetation is the tribe Lobelioideae, 
of which most of the species found here are new to science; they belong nearly 
all to the Sect. III., Palmaeformes, and are more or less closely related. The 
most interesting is Cyanea aculeatiflora, which, as the name implies, is covered 
with spines even to the lobes of the corolla; another peculiar new species is 
Cyanea hamatiflora with broad sessile leaves, dark-red flowers, and large purple 
fruits; the latter plant is most common on Puukakai, where it reaches a height of 
15 to 20 feet, similar to C. aculeatiflora. C. macrostegia with lobed leaves is not 
uncommon, and so is C. atra, but differing from the specimens found on West 
Maui. Cyanea ferox is here a shrub 15 feet in height with straight ascending 
branches, which together with the stem are covered with thorns; the leaves of 
this latter species are sinuate and remind one somewhat of Cyanea Grimesiana. 
Besides these tall species, two subherbaceous ones are found in the dense 
shady moss forest, the taller one of the two, Cyanea Bishopii (a new species, but 
first collected by the late E. Bishop, and referred by Hillebrand doubtfully to 
Cyanea Kunthiana?), with purple flowers, is the most common, but flowers, un- 
like the other species, in the winter months. As we cross Waikamoi, where we 
meet again with Gunnera petaloidea and Hillebrandia sandwicensis, the Ha- 
Wwailian begonia, the forest becomes more uniform. At the edge of Waikamoi 
proper, we find Lobelia macrostachys and a species of Wikstroemia, probably a 
new species. The writer crossed this forest belt from Olinda to Honomanu and 
followed along the ditch trail to Kailua. A forest as described in the above pages 
covers this stretch of land, and it may be remarked that at about 3000 feet ele- 
vation, above Honomanu, there are two clumps of Palms, Pritchardia arecina 
Bece. This palm, discovered by the writer, is new to science, and is described by 
Beccari in Webbia Vol. IV. Lower down along the ditch trail proper the forest 
has died for miles, the cause being still wnascertained. All the Ohia trees are 
dead, and only here and there a species of Tetraplasandra is struggling for exist- 
ence. Since the death of the tree growth the lobeliaceous Clermontia macro- 
carpa, so common on Oahu, has become almost the sole underbrush, with here 
and there a species of Cl. arborescens. 
What has been said of this forest belt up to Honomanu holds good for 
Keanae and Nahiku, the only exception being the presence of Siderorylon rhyn- 
chospermum at Nahiku, besides several species of Cyanea. 
The forests spoken of by Hillebrand at Ulupalakua have entirely disappeared 
and only remnants of them can be found. Cheirodendron Gaudichaudii is still 
common, besides Suttonia, and Ohia lehua; numerous still is the araliaceous 
70 
