Vegetative Anatomy of Dubautia^ Argyroxiphium^ and 
IF ilkesia ( Compositae ) ^ 
Sherwin Carlquist^ 
Because Dubautia, Ar^roxiphium, and W il- 
kesia are endemic Hawaiian genera of uncer- 
tain position within the Compositae and are 
characterized by species markedly different in 
habit, a more thorough knowledge of ana- 
tomical structure in these genera and in 
putatively related genera is desirable. The pur- 
pose of this study is to explore the variation 
pattern of anatomical characters in vegetative 
organs of Duhautia, Ar^roxiphium, and W il- 
kesia, and to suggest which of these appear to 
be important in indicating relationships 
among the genera and to other genera. The 
data may also be helpful in outlining natural 
groups within the genus Duhautia, 
Information concerning secondary xylem 
of Dubautia is included in a separate study 
(Carlquist, 1958). The peculiar leaves of 
Ar^roxiphium, and comparison of them with 
leaves of Wilkesia, form the subject of an 
earlier paper (Carlquist, 1957^). The present 
paper concerns the vegetative anatomy of the 
three genera in other respects. For the pur- 
poses of this study, the writer interprets Du- 
bautia as including Kailliardia, a treatment 
offered by Keck (1936: 24-28) and accepted 
by St. John (1950). 
The dubious position of Dubautia can be 
appreciated by comparing the works of au- 
thors who have dealt with the genus. Al- 
though recent authors agree that Dubautia 
and Railliardia, if both genera are recognized, 
are closely related, earlier authors recognized 
both genera and relegated them to widely 
separated portions of the family. Bentham 
and Hooker (1873: 393), for example, place 
Dubautia near Ar^roxiphium and W ilkesia, 
whereas they locate Railliardia {loc. cit., p. 
^ Manuscript received November 20, 1957. 
2 Claremont Graduate School, Rancho Santa Ana 
Botanic Garden, Claremont, California. 
442) between Railliardella, a genus tradi- 
tionally placed in Senecioneae, and the Juan 
Fernandez Senecioneae Robinsonia and Rhetim- 
dendfon. Although the systematic positions of 
Ar^roxiphium and Wilkesia have been in 
doubt, they have been interpreted as belong- 
ing to the tarweeds (Heliantheae, subtribe 
Madinae) by such authors as Hoffmann (1890: 
248). Hoffmann, however, places Dubautia 
and Railliardia in the subtribe immediately 
preceding Madinae, Galinsoginae. Skottsberg 
(1931: 56; 1956: 211) finds Dubautia and 
Railliardia possibly related to Robinsonia and 
Rhetinodendron, as well as to a New Guinea 
genus of Senecioneae, Brachionostylum. Keck 
(1936: 8) agrees, although he emphasizes the 
relation of Dubautia to Ar^roxiphium and 
Wilkesia, which he excludes from Madinae 
and places in Galinsoginae; and he suggests 
that Dubautia {sensu lato), Ar^roxiphium, and 
Wilkesia form a related endemic group. The 
fact that he relates Dubautia to elements of 
both Heliantheae and Senecioneae is not in- 
congruent, in his opinion, because Dubautia 
may be regarded as a genus transitional be- 
tween the tribes. Dubautia, therefore, would 
seem to be a critical genus in the delimitation 
of the tribes of Compositae. St. John (1950: 
240) finds no genus with close affinities to 
Dubautia, and Sherff (1935) declines to com- 
ment. Because the Heliantheae mentioned 
above are American, demonstration of rela- 
tionship between them and Dubautia, Ar^ro- 
xiphium, and Wilkesia would be important by 
validating an example of Hawaiian-American 
phytogeographical affinity. 
MATERIALS AND METHODS 
The author’s collections (listed below) 
from which portions were preserved in Car- 
noy’s Fluid (3 parts absolute ethyl alcohol: 
195 
