AKNOTT ON SOUTH AFIUCAN PLANTS. 233 



petals; Phylica with cucullate petals, and the ovary con- 

 tained in the bottom of the calyx-tube, and Sonlangia which 

 differs from Phylica by the ovary filling up the whole calyx- 

 tube. These at least are the more prominent differential 

 characters. Reissek has further subdivided Phylica into 

 Tylanthus which has the calyx-segments ovate and acute, and 

 a short indistinct conical style, and Phylica proper with 

 subulate calyx-segments, and a clavate or filiform style; and 

 has added a new genus Petalopogon, having subulate calyx- 

 segments, a short style, and cucullate fringed petals. Keep- 

 ing these in view, I shall indicate how far Drege's specimens 

 agree with such characters. Ph. fortuosa is a Tylanthns. — 

 Ph. sqiiarrosa agrees with all the characters of Tylanthus, 

 except the calyx, the segments of which are subulate as in 

 Phylica. — Ph. bicolor accords with Phylica^ except that the 

 calyx-segments are ovate and acute, as in Tylanthus. — Ph. 

 imberbis is a Tylanthus, as are also Ph. ericuides, Ph. parvi- 

 flora, a, and No. 6775. — No, 6777, a, is a true Phylica. — 

 No. 1917, a, and alsoP^. Thnnhergiana are species of Tylan- 

 thus, but the sepals have a subulate point. — Of Ph. cylin- 

 drica I have no flowers — No. 6*779, which is the same as 

 Sieber's flora mixta, No. 90, and apparently Ph. capitata, L, 

 belongs to Trichocephalus, where also must be brought Ph. 

 spicafa. No. 6787, 6788 a, 6790, 6752 b, Ph. callosa, and Ph. 

 stipularis. — Ph. retrorsa agrees with Trichocephahis in the 

 petals, but the calyx-segments are ovate. — Ph. plunwsa, No. 

 6770, 6772, Ph. pedicellata, Ph. rosmarimfolia. Ph. parvi- 

 fiora, c, e, and Ph. oleoides, all exhibit the characters of 

 Sonlanffia. 13esides these I may mention a species which was 

 in Mr Harvey's first distribution, No. 202, which I i)ad called 

 Trichocephalus Harveyi, fforibus capitatis, petalis apice dilata- 

 tis cucullatis margine membranaceo fimbriato, ovario glabro, 

 ramis junioribus villosis, foliis exstipulatisangustis basi cordatis 

 margine revoluto subtus incanis. But this must, I presume, 

 be referred to Petalopogon, (and perhaps to the species 

 already described, but the leaves are not cuspidate) akhough 

 I cannot see any material character to separate it from Tri- 



