216 
PACIFIC SCIENCE, Vol. VI, July, 1952 
result of this overgrazing. This can be demon- 
strated statistically. The only collectors to 
find more than one species were Pickering and 
Brackenridge (collecting in 1840), Remy (in 
1851-1855), and Hillebrand (1850-1871). Of 
the 14 species, only 3 have been collected 
more than once (not counting the revisit by 
the same collector, Hosaka, to the known 
type locality of /. Hosakae). Since 1872 there 
have been only five collections of Isodendrion, 
by five different collectors, representing four 
species. Other botanists who have collected 
widely in the islands, as Rock, Skottsberg, 
Fosberg, Mitchell, Cowan, Webster, Wilbur, 
et al., have failed to find the genus. In a sense 
this revision of Isodendrion has almost been a 
study in paleontology, as the investigation 
has been almost wholly of old collections in 
herbaria of species now extinct. It would have 
been gratifying to have many and abundant 
collections for each species. Such has not been 
the case. Though Hillebrand and others have 
grouped under I. pyrifolium any collection 
with more or less ovate leaves, it has been 
learned from study of the flowers, fruits, and 
stipules that there has been much speciation 
resulting in narrow endemism. No species is 
known to occur on more than one island, and, 
furthermore, there is local endemism to a 
particular mountain range or even to a few 
adjacent mountain valleys, as is so character- 
istic among the older Hawaiian plant genera. 
Since the best characters are in the flowers and 
stipules, it is very helpful to find that the 
stipules are long persistent and that the 
perianth is marcescent and persistent even on 
mature fruits. These characteristics have made 
it much easier to study and understand the 
species of Isodendrion. 
Extinction of the Species 
One cannot always prove that a given 
species of plant is extinct in its natural habi- 
tat, but the data presented in the discussion 
of distribution show that there is a great 
probability that the majority of the known 
species of Isodendrion are extinct, as well as 
the one or two probably lost species from 
Maui, once recorded by Hillebrand. Of the 
14 known species, nearly two thirds have not 
been collected since 1871 and are assumed to 
be extinct. The remaining species have been 
found and collected between 1895 and 1950, 
and most of these are probably extant, two of 
them having been found two and three times. 
These living species are: I. Forbesii, I. macu- 
latum, and /. suhsessilifolium of Kauai; 1, 
waianaeense of Oahu; and /. Hosakae of 
Hawaii. 
Phytogeny 
As is indicated by the arrangement in the 
key, the species fall into two groups. The 
first one has glabrous sepals, petals, and 
stipules. It appears that this is the group of 
more primitive species. The fully glabrous 
ones are only three, 1. laurifolium and /. longi- 
folium of Oahu and 1. maculatum of Kauai. 
These all have elongate, more or less ob- 
lanceolate blades. There is little basis for 
choice between these, but it is suggested that 
/. laurifolium represents the most primitive 
species. 
The group of really hairy species, which is 
considered more advanced, contains three 
species on Oahu, two on Hawaii, and one 
each on Niihau, Molokai, and Lanai. 
Acknowledgments 
Although parts of this study were done 
while visiting other herbaria, the bulk of it 
has been accomplished in Honolulu through 
the generous loan of herbarium specimens. 
With cordial thanks, acknowledgment is 
made of loans from the United States Nation- 
al Herbarium, Washington; the Gray Herbar- 
ium, Harvard University, Cambridge; the 
Royal Botanic Gardens, Kew; the Museum 
National d’Histoire Naturelle, Paris; and the 
Herbier Barbey-Boissier, Geneva. 
Explanations 
The location of the specimens studied and 
cited is indicated by the standard abbrevia- 
tions (Lanjouw, 1939). The great majority of 
