162 
PACIFIC SCIENCE, Vol. XI, April, 1957 
One of the most comprehensive reclassifica- 
tions of Eugenia was proposed by Niedenzu 
(1893). He established several segregate gen- 
era which were based mainly on the char- 
acters of the flower. His system was adopted 
by many workers, but, because of the lack of 
distinct generic limits, many other botanists 
continued to consider Eugenia in a broader 
sense. Later studies of the members of the 
group led more and more botanists to reject 
Niedenzu’s classification and to return the 
segregated genera to Eugenia. 
Because of the great number of species 
described in Eugenia the group has become 
rather unwieldy, and Merrill and Perry (1938*2, 
1938^, 1939) proposed a new systematic 
treatment of the group. They redefined some 
of the earlier proposed segregates of the 
genus. The new systematic treatment which 
they proposed is based on the structure of the 
seed. According to them, species of Eugenia 
are diagnosed as having a pericarp which is 
easily crushed and "the seed is free, the testa 
is smooth, chartaceous to cartilaginous and 
mostly lustrous, and the cotyledons are me- 
chanically inseparable, i.e., they have grown 
together in such a way that often the line of 
their opposing faces is scarcely distinguish- 
able." Syzygium, one of the segregate genera, 
is described as having "fruits that when dried 
are not too easily broken, and, when opened, 
the embryo (not the entire seed) falls out 
leaving the roughish seed coat more or less 
loosely adhering to the pericarp; the embryo 
has two distinct cotyledons usually attached 
near the middle of the opposing faces which 
conceal the hypocotyl within." 
In a monograph of the species of Eugenia 
in Malaya, R. M. Henderson (1949) critically 
analyzed Merrill and Perry’s classification. 
Henderson, in his study, decided that neither 
the degree of adherence of the seed coat to 
the cotyledon, nor the nature of the cotyle- 
dons themselves is consistent enough to be 
regarded as a good generic character. In fact, 
he gives detailed descriptions of the seeds of 
several different species which illustrate an 
intergrading series from those with two dis- 
tinct and separate cotyledons to those with 
the cotyledons completely fused. On the basis 
of these observations he rejects the transfer 
of species from Eugenia to Syzygium. 
In 1953 Ingle and Dadswell published the 
results of their studies of the wood anatomy 
of several Pacific Myrtaceae. 3 They concluded 
that the anatomy of the wood of Eugenia 
provides ample characters for splitting the 
genus. They added that the "suggested split 
on anatomical grounds involved only the two 
groups Eugenia A [corresponding to Eugenia 
sensu strictu] and Eugenia B [including the 
genera Acmena , Cleistocalyx and Syzygium ] . ” 4 
The most recent treatments of the genus 
have been inconsistent. By far the majority of 
workers (Airy Shaw, 1949; Backer, 1945; 
Degener and Ludwig, 1952) have accepted 
Merrill and Perry’s classification without any 
apparent critical study. A few botanists 
(White, 1945; Amshoff, 1942), seemingly in 
doubt as to the validity of this system, have 
hesitated to adopt it. 
EVALUATION OF THE RECENT 
GENERIC TREATMENTS 
The Hawaiian flora, although predomin- 
antly Asiatic in its affinities, includes repre- 
sentatives of the tropical regions of both the 
Old World and the New World. Because of 
the uniqueness of the Island flora in this 
respect, the Hawaiian Islands may be con- 
sidered an ideal region for evaluating the re- 
cent treatment of Eugenia. 
The two most important features used by 
Merrill and Perry in segregating the genera 
are: (1) the nature of the embryo, and (2) the 
degree of adherence of the testa to the coty- 
ledons. On the basis of these characters 
3 See also Dadswell and Ingle, 1947. 
4 Since this paper went to press Kathleen M. Pike 
published the results of her studies in the pollen 
morphology of the Myrtaceae. {Austral. Jour. Bot. 
4(1): 13-53, 1 ph, 1956.) She found that the pollen 
grains of "Eugenia A” are distinct from those of 
"Eugenia B” and thereby adds support to the group- 
ings of Ingle and Dadswell based on the wood anatomy. 
