88 
Kynsna, Uitenhage, Albany, Stockenstrom, East London, Kingwilliamstown, Komgha, 
and Kentani Divisions, and then appears in Natal and EE to Lourenco Marques 
The inflorescence in this species is also racemose {a very. abbreviated raceme), 
but specimens have passed through my hands in which the flowers are either solitary 
or 2-4-nate at the apex of short shoots ; this is, however, very rare. The third species, O. 
atyopurpurea, as 1 have limited it, is composed of plants of a shrubby habit, 2-3 ft. high, 
and can usually be recognised at a glance by the branches having a distinct lepidote appear- 
ance due to the presence of numerous prominent lenticels ; the leaves are usually elliptic, 
sharply serrated, and the flowers solitary. It has a similar distribution to OQ. arborea, but 
has also been recorded from Zululand and Barberton. The variety natalitia has been kept 
as a distinct species; it differs from O. afropurpurea in habit, and the branches have not 
the lepidote appearance ; the flowers are rarely solitary, and the leaves longer and lanceolate. 
Though Harvey recognized this variety, he has nevertheless confused it with the species, 
as in “the Cape Government Herbarium there is a sheet (E. and Z. 925) labelled on the 
ticket O. atropurpurea and written up as such in Harvey's handwriting, which is undoubtedly 
O. natalitia. This species is not so common in the south-eastern districts as O. atro purpurea, 
but is recorded from several localities in Natal, and then passes through Swaziland to 
Barberton. Ecklon and Zeyher record it from the Uitenhage Division. The correct limita- 
tion of this species has proved the most troublesome, and at first I was inclined to recognize 
three species. A further examination proved that the first grouping of the specimens was 
very artificial and that this species should be regarded as a variable one with a wide range 
of distribution. O. Chilversii appears to be confined to the eastern forests, 1t resembles 
QO. natalitia in general habit, but is readily distinguished from this species by the larger | 
number of carpels in the ovary, and the flowers are usually much larger. The two species, | 
O). O'Connorii and O. Holstii, which occur in the Woodbush Forest are very distinct from 
all the other South African species. Both are deciduous trees of timber size, and may 
perhaps be confused in herbaria if leaves are only present, but they differ considerably. 
O. Holstii has a rough bark and ord fruits, while O. O’Connorii has a smooth bark 
with kidney-shaped fruits. The remaining species, O. leptoclada, is found in the Barberton 
und Lyde nberg Districts and also in Rhodesia. Specimens in mature foliage are easily 
recognized by the oblanceolate leaves, which are long attenuated at the base. | 
The African species of Ochna were monographed by Diels in 1904 (Enel., *“ Bot. Jahrb., 
Vol. XXXIII, p. 232), to which account the reader is referred for a key to all the known 
African RA 
= — aa AA AA 
The writer was granted facilities by Mr. Kotze for extracting information about the 
species from notes sent to the Forest Department by various forest officers. . 
O. Chilversii.. This species is known in the Cwebe Forest by the natives as “ sibomvu.” | 
The Conservator of Forests, Transkeian Conservancy, remarks: “1 was struck with the 
nice appearance of the wood. It is a soft (hardwood), easy to work. . . . It is plentiful 
in the Cwebe and other coast forests and approximates arborea in size, and could easily be 
exploited. Possibly it might make an excellent hammer handle.” Mr C. C. Robertson 
remarks on the difference between this species and O. arborea. He reported that the `` wood, 
which seems to me far from soft, appears to have the same general character as boxwood, 
and would probably be suitable for similar purposes,” and he suggested that it might be 
tried for whatever purpose O. arborea is used for. The forester at “Cwebe, who has handled 
the wood, describes it as `` easy to work and soft.” 
In the Notinsela Forest the natives know the tree as ““umbomvane umncince,” and 
according to Mr. F. M. Fegen the tree occurs both in the Notinsela and Mpimbo Reserves, 
and appears to be plentiful in the latter locality, and no doubt also in the surrounding 
forests. me 
Forester Chilvers mentions that he has only seen one specimen in the Ingeli Forest 
Reserve. He describes it as a “ small tree, stem 5 inches in diameter; bark not smooth as 
in O. arborea ; leaves not diciduous.” 
