82 
SANDWICH ISLANDS. 
{Myrtacem. 
infimum foliolorum par, stipulis subulatis, racemis axillaribus erectis folio brevioribus, legu- 
mine pendulo lineari compresso membranaceo sub-8-spermo. 
This species seems to have been observed by Gandichaud (Freyc. Voy. p. 94.) to whom we have dedicated 
it ; but he says that the pods are narrow and short, which does not well apply to ours. The whole plant is 
perfectly smooth ; the leaves about six inches long', and the leaflets an inch and a half ; these last are 
on short petioles, about a line in length : the lower pair only are provided with a slender filiform gland, 
rather more than half a line long, which is incrassated and blunt, not acute at the apex as in C. oxyadena, to 
which this species seems allied. The pod is about three lines broad, and 3-4 inches long, very compressed 
and membranaceous, suddenly and shortly attenuated at both extremities. There are complete septa between 
the seeds, which are flat and almost black. 
Ord. XIIL rosacea. Juss. 
1. Osteomeles anthyllidifalia. Lindl. in Linn. Soc. Tr. v. 13. p. 98. t. 8. De Cand. 
Prodr, V. 2. p. 633. — Pyrus anthyllidifolia. Sm. in Rees’ Cycl. 
Ord. XIV. ONAGRARI^. Jnss. 
1. Jussisea angustifolia ; herbacea erecta glabra, foliis subsessilibus lineari-Ianceolatis 
iitrinque acuminatis, floribus brevissime peclicellatis, calycis lobis 4 acutissimis, tubo cylin- 
draceo elongate. DC. — Lam. Diet. 3. p. 331. III. t. 2%Q.f. 3. De Cand. Prodr. v. 3. p. 55. 
We agi'ee with Gaudichaud in referring the species found in Oahu to J. angustifolia. The leaves, how- 
ever, are slightly petiolate, and the lobes of the calyx, although acute, are not remarkably so. Mr. Collie, 
in his notes, remarks that the petals are obovate, rotundate, and emarginate ; their claws, internally, and the 
bases of the stamens, are somewdiat villous. 
Ord. XV. LYTHRARIiE. Juss. 
1. Lyihvxxm. maritimum ; caule fruticoso, foliis oppositis et sparsis lanceolatis acutis basi 
obtusis subtus glaiicescentibus, floribus breviter pedicellatis erectis bibracteolatis 6-petalis 
6-andris. DC. — Humb. et Kuntk, Nov. Gen. v. 6. p. 193. De Cand. Prodr. v. 3. p. 82. 
This we possess, from the same country, from Mr. Menzies, Mr. Macrae, and Chamisso. 
Ord. XVI. MYRTACEM. Juss. 
1. Metrosideros poly7norpha ; foliis oppositis breve petiolatis coriaceis utrinque glabris 
vel subtus sericeo-tomentosis, pedunculis tri-multifloris terminalibus et axillaribus corym- 
bosis, floribus pedicellatis, calycibus ramulisque glabris vel sericeo-tomentosis. — Gaud, in 
Freyc. Voy. p. 482. t. 108 et 109. De Cand. Pi'odr. v. 3. p. 225. 
a. Foliis subrotundo-ellipticis cordatis. — /3. Foliis ovatis vel ovato-ellipticis basi rotun- 
datis. — y. Foliis oblongis basi acutis. — <5. Foliis lanceolatis. 
These varieties, again, but particularly the fii'st, occur either perfectly smooth or more or less tomentose. 
Specimens of all of them are in the Collection, and we possess, in addition, several distinct forms of the 
second variety, collected by Mr. Macrae. Var. ji, was found long ago by Mr. Menzies, but was probably 
confounded by Smith with his ilf. villosa, to which it is indeed too closely allied, and perhaps only distinguish- 
able by the pedicellate flowers. This appears, according to Gaudichaud’s valuable observations, to be a most 
variable plant. Nothing, says he, is more remarkable than the M. polymorpha, which is found with linear 
leaves towards the summit of the mountain, but successively linear-lanceolate, lanceolate, ovate, obovate. 
