Since GARNER and ALLARD in 1920 found the three categories of plants, i, e. short-day plants, long-day plants and ever-flowering plants according to the relation between the flowering and the day-length, many plants have been studied to be classified among these three types. It may, however, be considered that there are two different stages in the course of flower formation, namely the differentiation of flower bud and the subsequent development of differentiated bud to flowering, and that these stages would receive the effects of day-length independently. To make clear this relationship some experiments have been carried out by the author during these six years from 1931 to 1936, using sixteen kinds of plants.
1. Cultivated pears in China belong to three main Pyrus spp., and the following names: (1) Tsiu tzu Li Pyrus ussuriensis, MAXIMOWICZ. (2) Pei Li Pyrus Bretschneideri, REHDER. (3) Sha Li Pyrus serotina, REHDER.
With respect to the root shape of the radish, MALINOWSKI (1916) previously indicated that the long is due to several partially dominant cumulative factors, which are absent from the recessive round shape. UPHOF (1924), however, found a unifactorial difference in all crosses where different root shapes were concerned. The F1 was intermediate in shape, and an approximate 1:2:1 ratio was always obtained in the F2. In order to make an accurate determination of this fact the present writer has attempted a similar experiment to those of the previous investigators with long and round varieties in the Japanese and Chinese radish. Perhaps the most direct, accurate, and concise description of shape in radish roots is given by root index (Root index=length/width). In the investigation the root shape was determined by this index, and the, results obtained were examined by statistical methods. The F1 was distinctly more vigorous than either parental form, and the root shape was approximately intermediate, but nearer the round parent. In the F2 a trimodal curve appeared. Accordingly, as UPHOF'S paper showed, it is fair to believe that the 1:2:1 ratio of round, intermediate, and long is obtained in the F2. While in the F2 the extracted long was more round than the long parent. It seems probable that in addition to the single major factor which causes the difference between round and long shape, there is another independent factor which affects root shape in the same general way but to a slighter degree. Consequently, it is concluded that the round shape of root is incomplete dominant over the long, the main difference being due to a single factor, a lthough another factor may produce minor effects.
1. In the results of this study the carbohydrate-nitrogen ratios have not proved of special significance in flower bud formation. 2. In the comparison of chemical composition between early summer headed and late summer headed shoots and also between shoots of long growth and those of medium growth of the Chojuro pear trees, the high content of soluble nitrogen was associated with flower bud formation. 3. The results of the study with short-day plants such as cosmos and chrysanthemum show that the content of carbohydrates was decreased by short days and flower bud formation in these plants was, therefore, associated with decrease in carbohydrates on the contrary to the general belief that the accumulation of carbohydrates favors flower bud production. 4. Within the limits of the data obtained in this study, high content of nitrogen, though not in every case, but on the whole, seems favorable for flower bud formation.