In the present paper the authors study especially the following two problems on the method of science education. 1) We aim to survey the actual states of the scholary attainments of science among the school children in some communities and to analyse them in quality and quantity, excuting the achievement test confining itself to the side of the knowledge of science. 2) It is essential to analyse the factors which are supposed to prescribe the differences in the attainments among schools, such as the domestic environments the arrangements for experiments etc. Here, however, we take up and clarify chiefly the views of science in those communities and afterwards examine their regitional differences, carrying out two kinds of questionarie: one for the children and the other for their parents. The results are as follows: 1) The result of the standerdized achievement tests in science which were carried out in four different communities, shows that the lower decline of scientific knowledge was markedly seen in rural communities. Though discriminative power on test items and the forms of providing problems were investigated by the application of G. coefficient method, no definite tendency was seen. Therefore, the problems were divided into two groups, one consisting of difficult ones, the other of easy ones, by judging from the degree how far the rate of correct answers was got out of the average, and the marked difference between two groups was studied. By these studies, it is concluded that unless scientific phenomenon is realy experienced through both observation and experiment by the pupils, they can hardly get the true comprehension and the scientific reasoning which lies in the background of the phenomenon. Furthermore, when this standerdized achievement test was tested in the interview situation, the increasing of scores was remarkably seen in rural communitied. Through this fact, it is denoted that there lie many problems which need much consideration, where scientific knowledge is to be tested with the from of paper test. 2) The views of science among these children and their parents were surveyed in our primary schools from the method of rank order. As for the order of importance of science towards all subjects, showing no difference among schools, ranking in the second order with sociology following after Japanese and Arithmetics. As for the order of choice of liking and disliking, it was found that boys liked it while girls rather disliked it. The survey about the attitude towards science showed some differences among communities, but these differences were not so significant that we could not conclude that there are fundamental differences between the rural and urban districts in views of science.