Native Hawaiian Cotton — Stephens 
395 
to the key accompanying the figure. Solid circles 
refer to the parental species ( barbadense to the 
left, and tomentosum to the right of the chart ) . 
Open circles represent the plants under study, 
the "Texaco” plants being distinguished from 
the "Piiiokahe” plants by dots placed within the 
circles. 
It is evident that although most of the plants 
resemble barbadense with respect to the charac- 
ters measured, they all deviate in one or more 
characters in the direction of tomentosum. This 
would be expected if they were hybrids segre- 
gating from backcrosses of tomentosum to bar- 
badense. A few of the plants had open bolls, 
and the seeds they contained bore fibers of two 
distinct kinds- — medium to long lint fibers, more 
or less adherent to the seed-coat; and coarse, 
strongly adherent fuzz fibers. Both lint and fuzz 
fibers were light brown in color. These fiber 
characteristics are typical of barbadense X to- 
mentosum hybrids studied previously in culture. 
Several visits were made to the hybrid popu- 
lations from the beginning of March through 
May, but at no time was insect activity among 
them very great. The pollen-feeding beetle 
( Conotelus ) was quite common inside the open 
flowers, but it is not likely that it would be 
effective as a cross-pollinating agent. A few 
small ants were also found in the flowers from 
time to time, but the only "prospects” likely to 
be concerned in natural crossing were the rare 
visits by honeybees, and one solitary visit by a 
syrphid fly, which were observed. Similarly, it 
was noted that Conotelus was common inside 
the flowers of dooryard barbadense in the same 
area, and rather infrequently honeybees were 
seen to enter the flowers also. A few carpenter 
bees and hornets were observed around the 
plants, but their attention appeared to be con- 
fined to the extrafloral nectaries, and they were 
not seen to enter the flowers. During the same 
period honeybees in particular, syrphids, and 
occasionally carpenter bees and small moths 
were seen to be actively visiting the flowers of 
Sida and a variety of garden ornamentals in the 
neighborhood. 
Thus it appears that rather infrequent visits 
by honeybees were most likely responsible for 
the origin of the hybrid populations, and for 
the subsequent backcrossing and intercrossing 
which have probably taken place. However, the 
Br - D <6 6 O 
cfcove 75 65-75 below 65 
Bra o' o o 
below 40 40-60 cfeove 60 
u. 0-0 0 
above 45 35-45 below 35 
FIG. 4. Scatter diagram showing variation pattern 
in hybrid populations at Nanakuli, Oahu. Staminal 
index plotted on horizontal scale, petal length on ver- 
tical scale. Other characters indicated according to key. 
Further explanation in text. 
relatively few observations which could be made 
on tomentosum provided no direct evidence that 
honeybees, or any other pollen-gathering insects, 
visit flowers of this species. 
LOCAL NAMES AND USE 
There are two Hawaiian names for Gossypium 
tomentosum: ma ( o ("yellow-green”) and hulu 
hulu ("hairy hairy”). Two independent sources 
suggest that the plant was used by the ancient 
Hawaiians as a source of green dye for their 
kapa cloth. Malo (1903:43, 74), writing about 
1836, stated definitely that ma'o was Gossypium 
tomentosum, and that its flower was used as a 
dye to color "tapa and the loin cloths of the 
women etc.” Bennett (1840:217), who was a 
surgeon on a whaling ship, stated that "a delicate 
green-yellow dye” obtained "from an infusion of 
the flowers of the cotton plant” was used for 
dyeing kapa in Oahu. Later in the same narrative 
(1840:255) he said: "Several exotic kinds of 
the cotton-shrub ( pulu or soft of the natives) 
grow wild in the plains, and include that rare 
species, the yellow or nankeen cotton ( Gossyp- 
ium religiosum ) called by the natives marou.’ T 
Brigham (1911:50, 144) referred to Bennett’s 
statement, and assumed that the latter had con- 
