413 
A revision of Davidsonia (Cunoniaceae) 
Gwen J. Harden and John B. Williams 
Abstract 
Harden, G.J. 1 and Williams, J.B. 2 0National Herbarium of New South Wales, Royal Botanic Gardens, 
Sydney, NSW 2000, Australia; 2 Botany Division, University of New England, Armidale, NSW 2351, 
Australia), A revision of Davidsonia (Cunoniaceae). Telopea 8(4): 413-428. The genus Davidsonia 
(Cunoniaceae) is revised on the basis of available herbarium collections, field studies over its 
geographic range, fresh and preserved materials and previous accounts. Formerly considered a 
monotypic genus, Davidsonia is found to comprise three species; D. pruriens F. Muell., the type 
species, from north-eastern Queensland; D. jerseyana (F. Muell. ex F.M. Bailey) G. Harden & J.B. 
Williams, a known variety from north-eastern New South Wales here elevated to specific rank; and 
D. johnsonii J.B. Williams & G. Harden, a well-known but hitherto undescribed species from north¬ 
eastern New South Wales and south-east Queensland. Each species is fully described and 
illustrated, and the lectotypification of D. pruriens is clarified. 
Introduction 
When describing tire genus Davidsonia in 1867, Ferdinand von Mueller referred it to 
the family Saxifragaceae. Later, when the Cunoniaceae were separated from that 
family, Davidsonia was traditionally placed in the Cunoniaceae, though Engler (1930) 
urged its exclusion because of the stiff hairs, alternate leaves, short stamens and 
exalbuminous seeds. Bange (1952) segregated Davidsonia from the Cunoniaceae into its 
own family, the Davidsoniaceae, and this has generally been regarded as a unigeneric 
family endemic to eastern Australia. The characters emphasised by Bange (op. cit.) 
when he described the family Davidsoniaceae were its spirally arranged leaves, non- 
endospermic seeds and epitropous ovules. Cunoniaceae sens, strict, usually has 
apotropous ovules but epitropous ovules have also been recorded in Spiraeanthemum 
in that family (Dickison 1980). Airy Shaw (1973), Dahlgren (1980), Takhtajan (1980), 
Cronquist (1981), and Mabberley (1997) have maintained Davidsoniaceae as a separate 
family, closely related to the Cunoniaceae. However, Hutchinson (1973) retained the 
genus in the Cunoniaceae without comment, while Ingle and Dadswell (1956) and 
Dickinson (1980) argued that the genus should remain in the Cunoniaceae on the 
evidence of its wood anatomy. Carpological evidence (Doweld 1998) supports the 
segregation of the Davidsoniaceae from the Cunoniaceae; however, the evidence also 
emphasises the families' close affinities. Recent research using DNA sequencing 
supports the inclusion of Davidsonia in the Cunoniaceae (Bradford pers. comm. 1997, 
1999; Angiosperm Phylogeny Group 1998). 
The genus Davidsonia was described in 1867 by Mueller with one species, D. pruriens. 
The southern populations of D. pruriens were described as a distinct variety, var. 
jerseyana, by Bailey in 1900. The brief description of var. jerseyana was that it was 
'smaller in all its parts'. The type variety is restricted to north-eastern Queensland, and 
the variety jerseyana confined to north-eastern New South Wales, though erroneously 
listed by Bange (1952) as also occurring in southern Queensland. These two taxa are 
markedly allopatric and differ consistently in inflorescence size and position, anther 
shape and size, pyrene morphology and size, arid leaflet venation. 
