1. The growing direction of the left valve of the oyster changes upward at the projected or hollowed portion of substratum. This phenomenon is remarkable in the smaller species and the southern population of certain species and also in the species found on the upper portion of the intertidal zone. 2. Generally if the left valve bends upward, the right valve is flat. But, if the left valve is flat, the right valve becomes deep. The right valve of larviporous species is also of deep form when the left valve bends upward. 3. The morphological variation of the oyster shells parallels the variation in weight of the shell valves. Namely in the case of the upwardly bended left valve, the left valve is heavier than the right valve, and if the left valve is flat, the right valve is the heavier. In the species which have symmetrical valves as larviporous species, both the left and right valves are almost equal in weight. 4. The relation of the form and weight between the left and right valve is characteristic in each species. Namely the degree of the said variation became smaller in the order of Crossostrea echinata, C. gigas, C. nippona, C. rivularis, O. circumpicta and O. denselamellosa. 5. The shell variation of C. nippona is rather similar to that of C. gigas.
The observations on the water-striders were made on the water surface of a small river in which Aquarius elongatus (A. e.) and Aquarius paludum (A. p.), were found. It seems that the population of each species shows a tendency to occupy its domain on the water surface in which the water-striders are skating, and the domain of one species overlaps that of another. The domain of each population is transformed by the individual numbers of each population. When the individual numbers of A. e. population are fewer than those of A. p. population, the individuals of A. elongatus are around the domain formed by A. p. population. When the individual numbers of A. p. population are fewer than those of A. e. population, the individuals of A. paludum occupy a small area of the corner of the domain formed by A. e. population. The individual of A. elongatus is more powerful than that of A. paludum, but the domain of each water-strider population is determined by the population densities of each species. Therefore, the domains formed by the populations of two species on the surface of the water seem to be determined by territorial behaviour of them.
The author previously mentioned that there might be a causal relation between the quantities of cultivation of the host plants and the tendencies of food preference of the common cabbage butterfly larvae. To ascertain this relation, the present experiment was planned. The common cabbage butterflies were reared successively for ten generations with cabbage for food. Tendencies of food preference of the fifth instar larvae (LC) between the cabbage and radish leaves were examined at the second, fifth and ninth generations. Also the larvae which were reared by radish (LR) for only one generation following the culture of one or five generations on cabbage, were examined similarly. The results are : 1. The larvae (LC) at the second or fifth generation cultured on cabbage showed no clear preference between the cabbage and radish leaves. This behavior bore some resemblances to that shown by the larvae collected directly from field at Sasayama. The larvae at the ninth generation preferred decidedly the cabbage leaves, which seemed to indicate that the change in the tendency of food preference had been established (Tables 1,2,3). 2. When the larvae, whose parents had been reared by cabbage, were cultured by radish, they preferred the radish leaves. Whereas when the larvae, whose ancestors had been reared by cabbage for five generations, were cultured by radish, they showed no clear preference between cabbage and radish leaves (Tables 4,5). 3. Therefore, a conclusion may be suggested that the larvae reared by a specific host plant successively for a certain generation will prefer that host plant.
To understand the exact relation between bees and flowers, further biological studies have been made in ten observation areas described in the previous paper. In this paper the following phenomena are taken into consideration. 1) From the studies on the correlation between bees and flowers in each area, flowering plants are distinguished into two types, those visited by a large number of bee species and those by a small number. From April to June, the flowering plants visited by the largest number of bee species are as follows : Brassica campestris subsp.>Astragalus sinicus> Ranunculus acris. But some differences are recognized in the personnel or species-construction of visitors to each flower species. On the other hand, the species which visits the greatest number of flowering plant species during April to June is Halictus scitulus, followed by Anthophora acervolum villosula>Bombus diversus>Bombus ardens. In other words, these bees are the most dominant pollinators in this district. But there are some differences between two kinds of species-construction of flowering plants visited by the bees belonging to Halictidae, and by the bees belonging to Bombidae or Anthophoridae (Table 1). 2) The seasonal transition of the species of visitors to each species of flowering plants were studied from the standpoint of both quantity and quality (Tables 2-7). For each species of flowering plants, is shown the period of time when the respective species is visited by the largest number of bee species (Table 2). From the total records on the respective species of bee which visits successively to a definite species of flower (Table 3), it is supposed that there is a close specific correlation between certain bees and flowers for a long time. Moreover the details are recorded on seasonal transition of flowers visited by bees with fairly long active life and that of bees with fairly long blooming period (Tables 6 and 7).