290 
PACIFIC SCIENCE, Vol. XXII, July 1968 
temperate flora with a less changing landscape. 
The rainfall of southern California is such that 
landslides though they occur are not the rule. 
This is quite the opposite of the situation gen- 
erally found in the tropics. 
The foregoing interpretation has been based 
on the assumption that the flora of Hawaii has 
developed largely from species which arrived 
through long distance dispersal. Campbell 
(1919, 1933) felt that the Hawaiian Archi- 
pelago is a remnant of a much larger area which 
was once in more or less direct connection with 
the South Pacific. He maintained that the isola- 
tion occurred in early Tertiary and that the 
older genera are derived from the South Pacific 
before isolation, and that the younger genera 
such as those belonging to the Compositae 
arrived after isolation by long distance dispersal 
and largely from the east. 
Skottsberg (1925) proposed that the flora of 
Hawaii represents the derivatives of an ancient 
antarctic continental flora. He speaks of a 
"Jurassic Hawaii” and apparently felt that such 
did once exist. The hypothesized sub-antarctic 
continent of Skottsberg and pre-Tertiary central 
Pacific land mass of Campbell are interesting 
conjectures. However, even if one assumes that 
they did occur, the Hawaiian Archipelago has 
been isolated for at least 5,000,000 years. 
Fosberg (1948) indicated that 5-10 million 
years would be enough time to account for the 
present flora, assuming one arrival by long 
distance dispersal on an average of once every 
20,000 years. 
If there is a high proportion of outcrossing 
in the species of the Hawaiian flora one would 
expect to find many polymorphic species. As 
Baker (1953) has pointed out, in outbreeding 
populations there is an almost continuous in- 
tergradation between different populations while 
in inbreeding populations there are much greater 
discontinuities between populations in both 
morphological and physiological characters. 
Nearly continuous intergradation between popu- 
lations is frequently found in the Hawaiian flora. 
Hillebrand (1888:xxv) states, "A comparison 
of the Hawaiian flora with that of any other 
country brings out at once a striking difference 
in the great number of varieties in all the species 
of leading genera.” And Fosberg (1948:107) 
commented, "The reputation for polymorphism 
enjoyed by Hawaiian plants has led taxonomists 
to avoid undertaking major problems on them.” 
Polymorphism, inherent in outcrossing breeding 
systems, is further evidence for a truly high 
proportion of outcrossing. The outcrossing has 
been effected at least partially through dioecism 
which has developed in at least 14 different 
families in less than approximately 5,000,000 
years. 
SUMMARY 
It is proposed that Baker’s Law has been 
effective during the development of the Hawai- 
ian indigenous flora. There is no reason to 
assume differently and there is good support 
for this view in the present structure of the 
Hawaiian flora. At the same time it can be 
demonstrated that there is a significantly higher 
percentage of dioecism in Hawaii than is found 
in continental floras. Dioecism is a very effec- 
tive mechanism for outcrossing and in species 
in which it occurs there is assurance that an 
open recombination type is present. While self- 
incompatibility of hermaphroditic species occurs, 
the same plant family usually does not show 
self-incompatibility and dioecism. It is true that 
dioecism alone is not a measure of outcrossing 
in a flora since other outcrossing mechanisms 
occur. However, by estimating dioecism, one 
knows that outcrossing occurs at least as much 
as would be indicated by the proportion of 
dioecism. Thus, there is no reason to deny the 
applicability to the Hawaiian indigenous flora 
of Baker’s Law of self-compatibility and long 
distance dispersal; yet there is an active out- 
crossing breeding system at work in Hawaii, 
where one would expect the contrary. The 
dilemma can be resolved if we recognize that the 
dioecious habit can develop relatively quickly 
and may then be perpetuated if it has greater 
survival value than hermaphrodism. Several 
workers have experimentally produced dioecious 
species or races. 
ACKNOWLEDGMENTS 
The author gratefully acknowledges the sug- 
gestions and criticism of H. G. Baker, University 
of California at Berkeley and Robert W. Long, 
University of Hawaii, both of whom kindly 
