1. The taro, in Japan, has many synonyms, as the cultivating history is old. So we tried, since 1945, to survey the 205 varieties of the taros collected all over this country, China, Formosa and southern islands, and to classify them in order to clear up oecological significance according to shape of leaves and tubers, chromosome numbers, and bud mutation.
2. Green colour degree, distribution of anthocy-an, crookedness of neck part of leaf-stalk, wilting of leaves and maturity are clearly different among varieties.
3. Shape of the corm has three types: round, cylindrical, and multiple-headed. Lateral tuber varieties(the corm is coarse and does not develope so large) and Akame in corm-tuber varieties (both corm and tubers are able to eat) are round, corm varieties (growth of lateral tubers is poor and few) are cylindrical, and only Yatsugashira and Shogaimo are multiple-headed. Lateral-tubers of Hasubaimo, Kurojiku and those of round shape line of Ishikawa-wase and Dotare are, too, in good shape.
4. Outbreak of lateral tubers is, naturally, different by maturity speed but only Hasubaimo is late in spite of its early maturity. Of second lateral tubers from first tubers, Kurojiku group has the most numbers.
5. From shape of flower organs, we think Eguimo and Okinawa-aoguki are Colocasia antiquorum Schott. var. trpica of Engler classification, the other lateral tuber varieties are var. globulifera, and corm-tuber varieties are var. esculenta.
6. Bud mutation is lately found; mutation of
anthocyan ditstribution on petiole at Yamagatataimo, Tonoimo and Akame, and mutation of out-break of lateral tubers at Yatsugashira. But we have not found the mutation changed to a remote variety group.
7. Chromosome numbers of these varieties studied by the Kyushu university have their important significance in this classification.
8. The varieties, in Japan, are classified from these above standpoints as follows.