抄録
We may consider two cases in correlations of rice characters in local phase, the first according to the results here obtained, was the local correlations between the mean rice characters and environmental conditions, and the second was the local correlations vetween the characters of a rice variety and environmental characters. We refer here only to the first case. 2. The characters which showed distinct positive correlation with mean heading date of rice are as follows; accumulated temperature for the period from seeding to heading. Mean temperature of the season of cultivation, hours of sunshine of the season of cultivation, harvesting date, seeding date, trsnsplanting date, period of maturity, panicle length and weight of thousand grains. The characters which showed no correlation with mean heading date of rice were as follows; culm length, number of panaicles, yield per tan, the period from seeding to heading and the period from planting to heading, and the character which showed a negative distinct correlation was latitudial index. 3. The character which showed positive correlation with mesn culm length of rice was panicle length, and that showed negative correlation was number of panicles per tsubo, and the other characters were independent. 4. The characters that showed positive correlations with mean numbers of tillers per plant were as follows; number of panicles per tsubo, plant growth index 1(culm length x no. of panicles per tsubo), and plant growth index II (culm length x no. of panicles per plant), hours of sunshine of the cultivation period and yield per tan. 5. Taking in account the correlation coefficient of rice characters in local phase, the influence of number of panicles per plant was found greater than that of culm length on the mean yields per tan of the locality. 6. From the conception of soil fertility the author separated the conception of the crop yielding efficiency. And he thought that the mean yield per tan of the prectural experiment station should indicate the crop yielding efficiency of the locality, and the characters which showed distinct positive correlations with the yield per tan crop yielding effiency were as follows; culm length, number of panicles per plant, growth index 1, plant growth index II, difference of daily temperature, markings of soil profile, markings of surfacial soil and fertility expressed by Fukuoka method. And the characters which showed a little negative correlation with yield per tan were as follows; the periods from seeking to heading, the periods from planting to heading, quantity of nitrogen in the standard fertilizer of the station and percentage of total nitrogen in the soil.
