90 SANDWICH ISLANDS. [Convolvulacece. 



(Gaud, in Freyc. Voy. p. 98,) is the same, since no description has been given. Mr. Brown takes notice of 

 two other species found in the Sandwich Islands, both with the segments of the corolla quite naked; these, 

 however, we have not seen. The habit of our plant is so similar to Leucopogon obovatus, Labill., that 

 Sprengel has united them without attending to the position or number of the bractese at the base of the calyx. 



Ord. XXVII. APOCYNEiE. Juss. 

 1. Alyxia sulcata; foliis oppositis ovatis obtusis submembranaceis utrinque nitidis 

 obsolete parallelim venosis, pedunculis axillaribus solitariis 3-floris folio dimidio brevioribus, 

 fructibus olivaeformibus longitudinaliter multi-sulcatis. 



This is a very handsome species, and perhaps, as we at first thought, the A. olivoeformis of Gaudichaud 

 (Freyc. Voy. p. 451,) found in the same islands; but there are too many points of discrepancy to allow 

 of our joining them. The leaves are never, that we see, in threes; nor are they acute at both extremities, 

 as in Gaudichaud's plant ; nor do we think he would have neglected to notice the numerous longitudinal 

 furrows on the fruit. From A. scandens, Forst., this differs in many respects, particularly in the furrowed 

 fruit, which has a perfectly even surface in that species ; a character we omitted to remark at page G6, from 

 not being at that time acquainted with the present plant. 



1. Cerbera parviflora ; foliis quaternis oblongis obtusiusculis parallelim venosis planis, 

 cymis pedunculatis axillaribus ramosis divaricatis. — " Forst. Prodr. n. 121." Willd. Sp. PI. 

 v. 1. p. 1222. Roem. et Schult. Syst. Veget. v. 4. p. 438. 



This species is omitted by Sprengel, and placed by Willdenow and subsequent authors among the doubtful 

 species. It is very closely allied to C. maculata, Willd., nor can we easily point out any character to 

 separate the two, except the shape of the leaves, which in our plant are very much broader, and not at all 

 spotted, as is well represented in the other species by Jacquin, (Ic. Rar.ii. t. 331.) There are constantly four 

 leaves in each whorl. The bracteas are small, and, towards the base, furnished with several spinous processes, 

 or teeth. 



Ord. XXVIII. CONVOLVULACE^. Juss. 



1. Convolvulus tuberculatus ? Desv. in Lam. Encycl. v. 3. p. 545. — Ipomaea tubercu- 

 lata ? Roem. et Schult. Syst. Veget. v. 4. p. 208. 



The specimen is not in flower, and is otherwise imperfect. 



2. Convolvulus Cairicus. Vahl. — Bot. Mag. t. 699. Spreng. Syst. Veget. v. 1. p. 592.— 

 Ipomsea palmata. Forsk. JEgypt. p. 43. Roem. et Schult. Syst. Veget. v. 4. p. 207. 



3. Convolvulus purpureus. Linn. — Bot. Mag. t. 113 et 1005. Spreng. Syst. Veget. v. 1. 

 p. 597. — Ipomaea purpurea. Lam. — Roem. et Schult. Syst. Veget. v. 4. p. 232. 



4. Convolvulus Pes Caprm. Linn. — Spreng. Syst. Veget. v. 1. p. 609. — Ipomaea maritima. 

 Br.— Roem. et Schult. Syst. Veget. v. 4. p. 249. Bot. Reg. t. 319. 



5. Convolvulus Batatas. Linn. — Spreng. Syst. Veget. v. 1. p. 607. — Ipomsea Batatas. 

 Lam. — Roem. et Schult. Syst. Veget. v. 4. p. 218. 



6. Convolvulus ovalifolius. Vahl. — Roem. et Schult. Syst. Veget. v. 4. p. 288. Spreng. 

 Syst. Veget. v. 1. p. 613. 



In the specimens before us, the branches are long and slender, from which we might almost conclude 

 the plant to have been prostrate and not erect. We possess, in our herbarium, a plant from Mr. Menzies, 

 also from the Sandwich Islands, which is a remarkable variety of the present species, being densely pubescent, 



