384 The Philippine Journal of Science 1915 
when the vast majority of apparently distinct varieties are still 
botanically unknown and when the named varieties, and even 
species, are so incompletely known and described that they are 
often totally unrecognizable among their very numerous unknown 
congeners. Undoubtedly some of the Philippine varieties will 
be found to be identical with Indo-Chinese, Amboinese, and other 
Malayan forms. But such reference made on the strength of 
any existing literature would be merely adding to the confusion 
already existing. For instance, there are all shades of red ba- 
nanas occurring throughout the tropics, representing a whole 
series of well-marked varieties, apparently quite distinct in the 
structure of their flowers. As a basis for classifying these, 
Schumann’s monograph furnishes us the following: 
Var. rubra (Firminger). Bacca. 15 cm. longa obscure rubra, matura 
flavido-rubra ; truncus, petiolus, medianus et flores obscure rubri, Ram- 
Kela Indorum. 
Var. violacea Bak. Truncus et saepe folia subtus plus minusve violacea. 
Var. saguinea Welw. ex Bak. Folia et bacca intense sanguinea. 
These descriptions are so general as to be of no value for the 
critical distinction of varieties. There is nothing in literature 
that would make it possible to know if the variety violacea of 
Baker, 1893, is the same as the variety violacea of Blanco. The 
probability is that they are not at all the same. None of Blanco’s 
numerous varieties are mentioned by K. Schumann. 
DESCRIPTION OF SPECIES AND VARIETIES OF BANANAS 
Since the time of Linnaeus, who described M. sapientum, the 
banana, and M. paradisiaca, the plaintain, many species of ba- 
nanas have been described by botanists. Due to lack of access 
to living plants or lack of good herbarium material these descrip- 
tions have very rarely been complete or clear, and this has 
rendered the group an extremely difficult one for study. Many 
of the descriptions are based on fruits alone, nothing being said 
of flower structure, which furnishes many characters of first 
importance. In connection with this work I have found that it 
is possible to make very good herbarium specimens which will 
represent leaves, bracts, and flowers. When these are accom- 
panied by photographs and drawings, as should always be the 
case, then the study of varieties may proceed, and gradually a 
foundation be laid for the grouping of races into varieties and 
varieties into subspecies and species. Hundreds of varieties of 
bananas and plantains are now in cultivation throughout the 
tropics and wild forms are numerous. The specific origin of the 
cultivated varieties is unknown, and can only be guessed at after 
