The fossiliferous, marine Upper Cretaceous sequence of fine sandy siltstone or silty sandstone lithology (Upper Yezo Group) is well exposed in the lower stream of the Pombets, central Hokkaido (Fig. 1).

The age of the uppermost part of the sequence is determined as the Coniacian, because of the occurrence of Inoceramus uwajimensis and Baculites cf. yokoyamai. The main part of the Upper Yezo Group in this area is Upper Turonian, as has been hitherto determined (Fig. 3).

Throughout the sequence the Upper Turonian comprises many well-preserved ammonites which are studied paleoecologically in this paper. The Upper Turonian ammonite assemblages are roughly classified into the following three in accordance with the differences in faunal composition and diversity, i.e. the assemblages of the lower, middle and upper parts (Fig. 4). The assemblage of the lower part is characterized by the common occurrence of a baculitid Sciponoceras intermedium with associations of collignoniceratids, scaphitids, tetragonitids and others. The assemblage of the middle part is similar to that of the lower part in the major species diversity, but differs in the absence of collignoniceratids. The upper Upper Turonian assemblage is rich in collignoniceratids (Lymaniceras planulatum and species of Prionocyclus) with rare associations of heteromorph and smooth or weakly ornate morphotypes. The comparison with previous works on the Turonian ammonite assemblages in other areas of Hokkaido (Matsumoto, 1965; Tanabe et al., 1978; Tanabe, 1979) suggests the ascending biofacies change from a shallow sea fades to a less shallow and more offshore "transitional facies," and again to shallow sea fades in the Upper Turonian of the Pombets area. The upper Upper Turonian ammonite assemblage characterized by L. planulatum, Prionocyclus aberrans etc. is also distributed in the Bibai (north of Pombets) and Sentarozawa (west of Pombets) areas, but has not yet been found from the Manji-Hatonosu areas (south of Pombets) (Matsumoto & Harada, 1964; Obata & Futakami, 1975, 1977). In contrast, the upper Upper Turonian assemblage in the Manji-Hatonosu areas and also in the Katsurazawa (eastern Ikushumbets) area is represented by the abundant occurrence of Reesidites minimus, which is very rare in the Pombets area.

These lines of evidence strongly suggest a segregation in habitats between R. minimus and L. planulatum in Hokkaido in the late Turonian time.

An ammonite from Madagascar has been recently donated to the Museum by Mr. Kazuo Kawakami and provided to me for study. It is called Torquatisphinctes sp. aff. T. primus Spath and described in Japanese, with illustration (Pl. 1, Fig. 1).

Remarks are given as follows: (1) It is necessary to examine the variations of numerous "species" from Madagascar and Cutch to give a final conclusion of classification. (2) Dimorphism is probably present in Torquatisphinctes and this specimen may represent a macroconch. (3) The suggested age from this specimen is the Kimmeridgian (Upper Jurassic). (4) The species can support the intimate palaeogeographical relation between Madagascar and the west coast of India in the Kimmeridgian time.

A huge specimen of Inoceramus from the Turonian of the Obira area, Hokkaido, donated to the Museum by Mr. Hiroharu Kokubu and another from the Turonian of the Hobetsu area, Hokkaido, by myself are described and illustrated under I. (I.) aff. iburiensis and I. (I.) iburiensis respectively.

On this occasion the specimen illustrated by Nagao & Matsumoto, 1939, pl. 31, fig. 1 is desgnated as the lectotype of I. (I.) iburiensis Nagao et Matsumoto and a specific diagnosis is given on the basis of this and other specimens. This species is somewhat similar to but can be distinguished from the nearly contemporary I. (I.) lamarcki Parkinson and the Coniacian species I. (I.) percostatus Müller. It is suggested that the group of I. (I.) lamarcki is represented in the Middle Turonian by similar but different subgroups of species between separate provinces of the North Temperate Realm and that they may have differentiated directly or indirectly from some unknown common ancestor(s) of the Early Turonian.

The male and female genitalia of Endoclita excrescens (Butler) and Hepialus fusconebulosa (De Geer) are described and figured. The musculature of male genitalia of Endoclita excrescens (Butler) is described and figured. The mode of firm connection between male and female genitalia, which has not been recorded in Lepidoptera, is effected in the following sequences: (1) the mesosome was inserted into the dorsal groove of central process (E. excrescens) or grappled the dorsal sclerotized portion of central process (H. fusconebulosa), (2) the valvellae were applied to the lateral walls of seminal gutter (E. excrescens) or the depression of subanal plate (H. fusconebulosa), which was opened by the inflation oftegumen, (3) the valvae clasped the weakly sclerotized portions of female diaphragma (E. excrescens) or lateral portions of female ninth tergum and sternum (H. fusconebulosa). The characteristics of these copulation mechanism in Hepialidae and their phylogenetic significance are shortly discussed.