Notes on Banana Varieties in Hawaii 
N. W, SlMMONDS^ 
Knowledge of the botany of the banana 
has been advanced considerably in recent years 
by the taxonomic researches of Cheesman 
(1947 et seq.). It is now clear that the edible 
bananas have originated from three wild spe- 
cies of which one, Musa acuminata Colla, is 
by far the most important. This species be- 
longs to the section Eumusa of the genus 
Musa and is a variable (but constantly diploid) 
species with 2n = 22 chromosomes. Many edi- 
ble varieties have derived from it, some being 
diploid, some triploid (2n = 3x = 33). Musa 
halbisiana Colla is another diploid Eumusa, 
and it has contributed to the origin of edible 
bananas by hybridity with M. acuminata. 
Upon triploid cultivars of such hybrid origin 
were founded the two Linnaean species Musa 
sapientum and M. paradisiaca. (This statement 
includes a modification of Cheesman’s views 
on the subject, but the point is immaterial for 
the present purpose.) The third species (fol- 
lowing MacDaniels, 1947, rather than Chees- 
man, 1949 ) is M. troglodytarum L., the fe'i 
banana of the Pacific, perhaps more com- 
monly referred to as M. fehi Bert, ex Vieillard. 
It is a member of the section Australimusa and 
is diploid with 2n = 2x = 20 chromosomes. 
The wild progenitor of this complex cultigen 
has not yet been identified, and, indeed, it 
may well turn out that more than one wild 
form was involved in its origin. The section 
Australimusa is still very poorly known taxo- 
nomically, and, until this deficiency is rem- 
edied, we are not likely to understand the vari- 
ability or origins of the fed banana, as it may 
conveniently and noncommittally be called. 
^ Banana Research Scheme, Imperial College of 
Tropical Agriculture, Trinidad, British West Indies. 
(Abbreviated in the text as I.C.T.A.) Manuscript re- 
ceived August 11 , 1953. 
Edibility, therefore, has had independent 
origins in different sections of the genus 
(Cheesman, 1947, 1948; Dodds, 1946). 
The wild bananas are all native to tropical 
Asia and Australasia, and the edible deriva- 
tives must have had their origins somewhere 
in that vast area. Malaya was almost certainly 
the home of some edible Eumusas; the fe'i 
banana presumably came from an area much 
farther south and east, in the islands where 
the wild Australimusas grow. Spread in cul- 
tivation must have been entirely by suckers 
after the process of evolution of partheno- 
carpy had gone far enough to impose seed 
sterility (Dodds and Simmonds, T948), and 
it must have been by this means that Hawaii 
received her cultivated bananas. Once seed 
sterility has been established in a crop plant 
(or rigorous vegetative propagation is prac- 
tised), somatic mutation is the only source 
of genotypic variability. Such mutation, or 
sporting, is well known to occur in bananas 
(e.g., see Baker and Simmonds, 1951; Sim- 
monds, 1953 ) and has been a significant 
source of variability among the old Hawaiian 
varieties as Pope (1926), by classifying the 
Hawaiian varieties in groups, implicitly rec- 
ognised. Evidently the Hawaiian cherished 
the mutants as he cherished mutant forms of 
taro and cane. 
In Hawaii, as elsewhere in the Old World 
tropics, two banana cultures are superposed 
one on the other: an indigenous pre-European 
culture of bananas introduced, used, and 
named by the Hawaiian people; and a later 
culture of varieties introduced in recent times 
by Europeans. The two groups may conve- 
niently be called "Hawaiian” and "alien,” 
respectively, and with the foregoing remarks 
in mind we may proceed to an annotated list 
226 
