Native Hawaiian Cotton {Gossypium tomentosum Nutt .) 1 
S. G. Stephens 2 
Although the wild cotton, Gossypium to- 
mentosum Nutt., is one of the more common 
of the few endemic species which still survive 
on the coastal plains of the Hawaiian Islands, it 
remains relatively unknown to the geneticist. 
Elsewhere it has been grown with indifferent 
success in experimental culture. Under such di- 
verse conditions as those found in the West 
Indies, southern Mexico, the U. S. cotton belt, 
and in greenhouse culture, it flowers sparingly 
and even less frequently sets seeds. As a con- 
sequence, experimental studies have been very 
restricted, and cytogenetic analysis has been con- 
fined almost entirely to the few crosses which 
have been made with annual forms of the 
related New World species, G. barbadense L. 
and G. hirsutum L. To the technical difficulties 
may be added the lack of representative collec- 
tions of the species in culture. The few accessions 
studied have usually been obtained from the 
more readily available Oahu populations, and 
less frequently from Molokai. These have been 
supplied to cotton geneticists through the cour- 
tesy of resident Hawaiian botanists, J. F. Rock, 
O. Degener, A. Mangelsdorf, and others, and 
patiently resupplied as fast as the stocks in 
culture expired. 
As seen in culture, the different accessions 
show little morphological variation (finely to- 
mentose vegetative parts, sulphur -yellow and 
spotless flowers, long anther filaments partly 
fused in pairs, nectariless leaves, and undifferen- 
tiated seed fibers ) . More recently, visitors to the 
islands, unacquainted with the taxonomy of the 
species, have sent seed samples to the mainland, 
and the plants grown from some of these appear 
to differ rather widely from earlier "authentic” 
accessions. 
It has long been recognized that an under- 
standing of the degree of relationship existing 
between G. tomentosum and the other 52- 
1 Contribution from Genetics Department, Univer- 
sity of Hawaii. Manuscript received May 23, 1963. 
2 On leave of absence from Genetics Department, 
North Carolina State College, Raleigh, N. C. 
chromosome species of Gossypium is funda- 
mental for a satisfactory interpretation of the 
evolution of the genus as a whole. It is some- 
what disturbing for the cytogeneticists not to 
know if the limited samples of the species 
studied in culture are at all representative of its 
natural range of variation, nor to be confident 
that some of the "off-types” which have been 
collected recently may safely be attributed to the 
effects of hybridization with those 19th-20th 
century introductions of barbadense and hirsu- 
tum which still persist in the islands ( Stephens, 
1963). 
In the spring semester, 1963, a temporary 
appointment to the staff of the Genetics De- 
partment, University of Hawaii, gave me the 
opportunity to make a field study of the native 
species. I should like here to record my thanks 
to the University of Hawaii for inviting me to 
their campus and to my home institution, North 
Carolina State College, for granting me the 
necessary leave of absence. Thanks are also due 
to the National Science Foundation for defray- 
ing a major part of the travel expenses (NSF 
Grant G-14203). 
GEOGRAPHICAL RANGE 
G. tomentosum has never been found beyond 
the limits of the Hawaiian Islands. Statements to 
the contrary have arisen through nomenclatural 
confusion (Watt, 1907:69-71 ) . 3 Hillebrand’s 
statement ( 1888:51) that it occurred on all the 
Hawaiian Islands probably referred only to some 
or all of the eight major islands of the group — 
Niihau, Kauai, Oahu, Molokai, Lanai, Kahoo- 
lawe, Maui, and Hawaii. The smaller leeward 
islands from Nihoa and Necker and beyond to 
Midway and Kure were studied by the Tanager 
Expedition, and the list of vascular plant collec- 
3 The matter has recently been re-investigated by Dr. 
R. Wilbur (Pacif. Sci. 18(1) :101— 103, 1964), who 
concludes that ”G. sandvicense Pari.” is the correct 
name for the Hawaiian native cotton, not "G. tomen- 
tosum Nutt.” as used throughout this article. 
385 
