The rice stem maggot, Chlorops oryzae MATSUMURA, has two generations a year in northern Japan and has three in the south. There seems to be local variation in the period of fly emergence both in the two and in the three generation districts. The authors conducted some experiments on growth of larvae and pupal duration, and were able to attain some contributions to the geographical variations in the occurrence of this insect.<br>The winter hosts of the rice stem maggot with parasitic larvae were sent to Takada from various places. These places are Omagari belonging to the two generation districts, Murakami and Kanose located in border zone of the two and the three generation districts, Matsudai located in mountainuos area of the southern parts of Niigata Prefecture, and Hatano belonging to the three generation districts. The period of the first fly emergence, the duration of larva and the duration of pupa of these materials were compared to those of the Takada's population.<br>The results obtained are as follows:<br>1. The larvae of Takada grew most speedy in spring, those of Hatano were secondary speedy, and the growth of larvae of Omagari, Murakami and Matudai was greatly delayed.<br>2. The period of emergence of the first generation fly of Takada was earliest, and that of Hatano was secondary though delayed about 10 days from the former. In other places that period was more delayed, and in the Matsudai, Murakami and Kanose populations the period from the first to last emergence of the fly was long.<br>3. The wintering larvae from some places were inoculated into young rice plants and reared in a greenhouse. It was found that the emergence of the fly from Takada was earliest, that from Hatano was secondary, and that from Murakami and Omagari was most delayed.<br>4. The pupal duration of winter generation of Takada was about 10 days, those of Hatano and Omagari were respectively about 11 and 12 days, and those of Murakami and Kanose were both about 12.5 days.<br>5. The larval duration of the first brood of Takada was less than thirty days, being shortest in all places examined. It was secondary in Hatano, and longest in Omagari where all the individuals needed more than fifty days to pupate.<br>The larval durations of Matsudai and Kanose had great variances; some larvae needed only less than twenty days to pupate, but there were the larvae which needed more than fifty days.<br>6. The larvae of winter generation of Omagari were brought to Takada in autumn of 1956 and reared there throughout 1957. The period of fly emergence of the first brood in 1958 of that population was similar to that of the population brought to Takada in autumn of 1957.<br>7. It can be stated from these results that there are remarkable geographical variations in growth of the rice stem maggot. The difference between the population of Omagari and those of Takada and Hatano is very significant. It may be concluded that there are two races in the rice stem maggot, the one has three generations in a year and distributes in southern parts of Japan, the other has only two and distributes in the north. But there also seems to be considerable local variations in some provinces of three generation districts.<br>8. Female flies of the first generation of Takada and male flies of the same generation from Omagari were crossed. The larval durations of the next generation were greatly variable and their variability was similar to that of the Kanose and Matsudai populations.<br>It is considered from this fact that in the border zone of the two and three generation districts, there are populations which derive from the natural hybridization between the two races.