ALBERT

Description: In April 1969 I paid a visit to Ceylon for a week, allowing me to study for the first time the collections of the Department of Agriculture, Peradeniya (PDA), including Thwaites type specimens. My stay was made possible through the Smithsonian Flora of Ceylon Project. The study of Thwaites’ type material revealed some new facts affecting the synonymy of Ochna jabotapita L. and O. obtusata DC. It had previously come to my attention that materials distributed as O. moonii Thw. under number C.P. 1224 belonged to either O. obtusata (BM, BO) or O. lanceolata Spreng. (K, P) (see also the note on page 26 of my revision). I subsequently found that all three species of Ochna in Ceylon were represented on the sheet in PDA, obviously bearing Thwaites’ holotype. From this and accompanying sheets it is clear that the material belonging to O. jabotapita should in fact be designated as the holotype of Thwaites’ species. Consequently, the whole paragraph under O. moonii on page 30 of my revision should be transferred from the synonymy of O. obtusata to that of O. jabotapita. The phrase ‘excl. syn. O. quarrosa L. sensu Moon = O. jabotapita L. ’should be deleted. The type should be referred to as C.P. 1224 p.p. (PDA p.p. holo).

Description: Woody plants, very small undershrubs to tall trees. Leaves distichous or spirally arranged, stipulate, simple, glabrous; midrib prominent on either side. Inflorescences 1- to many-flowered, cymose, racemose, or thyrsoid, bracteate; pedicels articulate. Flowers actinomorphic, bisexual (rarely functionally polygamous). Sepals 5, free or a little connate at base, quincuncial, persistent. Petals 5-10, free, contort, caducous. Staminodes 0-∞. Stamens 5-10-∞; anthers basifix, ± latrorse and dehiscing lengthwise, or with 1-2 apical pores. Carpels 2—5—10(—15), superior, free with 1 ovule, or fused with 2-∞ ovules per carpel; styles fused, basigynous or epigynous; stigmas free or ± fused. Fruit(s) a drupe(s), berry, or capsule. Seeds 1-∞, small or large, sometimes winged, with or without albumen. Taxonomy. There is little doubt that the family of the Ochnaceae represents a natural one among the more primitive in the Guttiferales (= Clusiales or Theales s. l.). Nonetheless, there are striking differences between the genera, even at first sight. It is not difficult to arrange them in a few distinct, supra-generic taxa. A supposed natural system, as far as relevant to the Malesian genera, is as follows: Subfamily Ochnoideae Tribe OCHNEAE ....................... Subtribe Ochninae 1. Ochna 2. Brackenridgea Subtribe Ouratinae 3. Gomphia Subfamily Sauvagesioideae Tribe EUTHEMIDEAE . ..................................4.Euthemis Tribe SAUVAGESIEAE ........................Subtribe Sauvagesiinae 5. Neckia 6. Indovethia 7. Schuurmansiella 8. Schuurmansia

Description: This taxonomic revision includes all Ochnaceae from South and Southeast Asia, Malesia, Australia, and the Pacific Islands. A much wider species concept is applied than in most previous works: 20 species in 10 genera are accepted for the area treated, whereas 199 specific synonyms are listed. A few taxa have been accepted at infra specific level among which Brackenridgea palustris ssp. kjellbergii Kanis is new. It is shown that some old specific names have been overlooked in the past and that the traditional concepts of some species have not been in accordance with the original concepts. All names, currently used for Asiatic species of Ochna L. and Gomphia Schreb. are no longer accepted here, O. jabotapita L. and G. serrata (Gaertn.) Kanis being the correct names for the respective type species. O. fascicularis Blanco is made the type of a distinct section Notochnella (v. Tiegh.) Kanis in the genus Brackenridgea A. Gray. A short history of the taxonomy is given and a partly new suprageneric subdivision of the Ochnaceae is subsequently proposed. Two subfamilies are recognised: the Ochnoideae comprising the tribes Ochneae and Elvasieae Rchb., and the Sauvagesoideae Lindl. including the tribes Sauvagesieae, Euthemideae Planch., and Lophireae Rchb. The Ochneae are newly subdivided in the subtribes Ochninae and Ouratinae (v. Tiegh.) Kanis, and the Sauvagesieae in the subtribes Sauvagesinae and Luxemburginae (Planch.) Kanis.* Some general remarks are made about morphological characters in the family, including some new characters of the pollen. An attempt is made towards a better understanding of the inflorescence types. It is assumed that the genera Ochna and Gomphia migrated from Africa into Asia. Other genera in Southeast Asia, Malesia, Australia, and the Pacific Islands are regarded as long established, because of the more or less relict-like nature of their areas.

Description: This taxonomic revision includes all Ochnaceae from South and Southeast Asia, Malesia, Australia, and the Pacific Islands. A much wider species concept is applied than in most previous works: 20 species in 10 genera are accepted for the area treated, whereas 199 specific synonyms are listed. A few taxa have been accepted at infra specific level among which Brackenridgea palustris ssp. kjellbergii Kanis is new. It is shown that some old specific names have been overlooked in the past and that the traditional concepts of some species have not been in accordance with the original concepts. All names, currently used for Asiatic species of Ochna L. and Gomphia Schreb. are no longer accepted here, O. jabotapita L. and G. serrata (Gaertn.) Kanis being the correct names for the respective type species. O. fascicularis Blanco is made the type of a distinct section Notochnella (v. Tiegh.) Kanis in the genus Brackenridgea A. Gray. A short history of the taxonomy is given and a partly new suprageneric subdivision of the Ochnaceae is subsequently proposed. Two subfamilies are recognised: the Ochnoideae comprising the tribes Ochneae and Elvasieae Rchb., and the Sauvagesoideae Lindl. including the tribes Sauvagesieae, Euthemideae Planch., and Lophireae Rchb. The Ochneae are newly subdivided in the subtribes Ochninae and Ouratinae (v. Tiegh.) Kanis, and the Sauvagesieae in the subtribes Sauvagesinae and Luxemburginae (Planch.) Kanis.* Some general remarks are made about morphological characters in the family, including some new characters of the pollen. An attempt is made towards a better understanding of the inflorescence types. It is assumed that the genera Ochna and Gomphia migrated from Africa into Asia. Other genera in Southeast Asia, Malesia, Australia, and the Pacific Islands are regarded as long established, because of the more or less relict-like nature of their areas.

Description: In table I on page 6 of this volume the names of some taxa have been misspelled. Subfamily Sauvagesoideae Lindl. should read Sauvagesioideae Lindl. The subtribal names Sauvagesinae and Luxemburginae (Planch.) Kanis should be changed to Sauvagesiinae and Luxemburgiinae (Planch.) Kanis respectively. The orthographic errors mentioned have been made consistently throughout the text.\nIn the same table the new subtribe Ouratinae (v. Tiegh.) Kanis was proposed with an incorrect reference to tribus Ourateae (non Engl.) v. Tiegh. (1902). The name Ourateae was introduced by Engler (Nova Acta Leop.-Carol. Akad. 37, 1874, 20) to indicate the tribe that should be called Ochneae, as Ochna L. should be considered the type genus of the family and lower taxa to which it belongs (cf. Art. 19: 3 of the present Code). Engler\xe2\x80\x99s name was invalid, although he gave a Latin description (cf. Art. 32: 2 of the present Code). Van Tieghem correctly used the name Ourateae for a smaller tribe, segregated from his Ochneae. This should be considered as a new name, as he made a new description without reference to Engler\xe2\x80\x99s. The new subtribe is based on the tribe Ourateae v. Tiegh. in Morot, J. Bot. 16 (1902) 33.

Description: Woody plants, very small undershrubs to tall trees. Leaves distichous or spirally arranged, stipulate, simple, glabrous; midrib prominent on either side. Inflorescences 1- to many-flowered, cymose, racemose, or thyrsoid, bracteate; pedicels articulate. Flowers actinomorphic, bisexual (rarely functionally polygamous). Sepals 5, free or a little connate at base, quincuncial, persistent. Petals 5-10, free, contort, caducous. Staminodes 0-\xe2\x88\x9e. Stamens 5-10-\xe2\x88\x9e; anthers basifix, \xc2\xb1 latrorse and dehiscing lengthwise, or with 1-2 apical pores. Carpels 2\xe2\x80\x945\xe2\x80\x9410(\xe2\x80\x9415), superior, free with 1 ovule, or fused with 2-\xe2\x88\x9e ovules per carpel; styles fused, basigynous or epigynous; stigmas free or \xc2\xb1 fused. Fruit(s) a drupe(s), berry, or capsule. Seeds 1-\xe2\x88\x9e, small or large, sometimes winged, with or without albumen.\nTaxonomy. There is little doubt that the family of the Ochnaceae represents a natural one among the more primitive in the Guttiferales (= Clusiales or Theales s. l.). Nonetheless, there are striking differences between the genera, even at first sight. It is not difficult to arrange them in a few distinct, supra-generic taxa. A supposed natural system, as far as relevant to the Malesian genera, is as follows: Subfamily Ochnoideae Tribe OCHNEAE ....................... Subtribe Ochninae 1. Ochna 2. Brackenridgea Subtribe Ouratinae 3. Gomphia Subfamily Sauvagesioideae Tribe EUTHEMIDEAE . ..................................4.Euthemis Tribe SAUVAGESIEAE ........................Subtribe Sauvagesiinae 5. Neckia 6. Indovethia 7. Schuurmansiella 8. Schuurmansia

Description: In April 1969 I paid a visit to Ceylon for a week, allowing me to study for the first time the collections of the Department of Agriculture, Peradeniya (PDA), including Thwaites type specimens. My stay was made possible through the Smithsonian Flora of Ceylon Project.\nThe study of Thwaites\xe2\x80\x99 type material revealed some new facts affecting the synonymy of Ochna jabotapita L. and O. obtusata DC. It had previously come to my attention that materials distributed as O. moonii Thw. under number C.P. 1224 belonged to either O. obtusata (BM, BO) or O. lanceolata Spreng. (K, P) (see also the note on page 26 of my revision). I subsequently found that all three species of Ochna in Ceylon were represented on the sheet in PDA, obviously bearing Thwaites\xe2\x80\x99 holotype. From this and accompanying sheets it is clear that the material belonging to O. jabotapita should in fact be designated as the holotype of Thwaites\xe2\x80\x99 species. Consequently, the whole paragraph under O. moonii on page 30 of my revision should be transferred from the synonymy of O. obtusata to that of O. jabotapita. The phrase \xe2\x80\x98excl. syn. O. quarrosa L. sensu Moon = O. jabotapita L. \xe2\x80\x99should be deleted. The type should be referred to as C.P. 1224 p.p. (PDA p.p. holo).