1. The ecological survey of a breeding bird community was carried out in the broad-leaved deciduous forest of Chichibu-Tama National Park in central Japan, and the food consumption of each species population was estimated by a procedure based on the feeding experiments of captive birds. 2. The feeding experiments were carried out on Uroloncha domestica, Cyanopica cyanus japonica and Corvus corone orientalis. The food intake was 2.7 and 2.4gm dry weight per bird per day or 11.1 and 9.8 Kcal per bird per day in female and male Uroloncha, respectively, fed on great millet. It was 11.8 and 15.2gm dry weight per bird per day or 57.2 and 72.8 Kcal per bird per day in Cyanopica fed on a paste food and sikworm, respectively. Corvus fed on horse-mackerel consumed the food averaging 38.2gm dry weight per bird per day or 221.4 Kcal per bird per day. 3. The faeces output was 0.34 and 0.31gm dry weight per bird per day or 1.18 and 1.08 Kcal per bird per day in female and male Uroloncha, 7.18 and 4.25gm dry weight per bird per day or 27.02 and 14.51 Kcal per bird per day in Cyanopica fed on a paste food and sikworms, and 15.3gm dry weight per bird per day or 41.4 Kcal per bird per day in Corvus. 4. The interspecific comparison of digestible energy was made on available data, and the relation between digestible energy and body weight was discussed. It was suggested that the logarithm of digestible energy was roughly proportional to the logarithm of body weight according to the equation Y=2.03W0.706. 5. The digestive coefficient of various food organisms was also compared and some discusssions were made. 6. The available data on the food consumption of the wild birds were examined with respect to that of captive birds and the relation between food intake and body weight. 7. The breeding bird census was carried out by the method of territory mapping on a belt-shaped plot, which was 2750m long, in 1957 and 1958. 8. Number of species was 28 in 1957 and 29 in 1958. The total density was 505+ pairs per square km in 1957 and 636+ pairs per square km in 1958. The most abundant species was Erithacus cyane, followed by Parus major, Phylloscopus occipitalis and Parus ater (only in 1958). 9. The total biomass was about 45kg per square km in both years, much higher than that of similar forest of eastern North America. 10. The pattern of distribution corresponding with the pattern of physiographic and/or vegetational feature of the habitat in the census area was seen in some species. 11. Accoriding to the feeding niches observed, 26 species were divided into 7 or 8 feeding groups, and the feeding niches of some species were described in some details. 12. The food consumption of every species population in the forest was calculated on the several assumptions. The group which consumed the food organisms most abundantly was "Ground" group and those species that consumed the food much more than others were Phasianus soemmerringii, Erithacus cyane and Garrulus glandarius. 13. Some discussions based on these calculated values were made.
Japan lies apart from the continent and many migration routes to and vir this archipelago are innevitably sea-crossing, for example, Japan Sea and China Sea crossings, 425-480 miles. Many stray autumn migrants of land birds reach the Bonin and Volcano Is., possibly aided by oceanic wind drift after crossing 500-600 miles and more may perish on the ocean. They seem to survive the winter on the islands but the summer records are extremely few and spring migration from the south is not evident. This is supported by the fact that actually no Passerines reach Marianas, N. Micronesia, though some palaearctic ducks further reach W. Carolines and even Marshall Is. There are, on the other hand, some northern migrant species which reach Palau Is., SW. Micronesia, from the Philippines crossing 580 miles of the sea, perhaps also aided by winds. Except the waders, only a few American ducks could reach the Marshall Is., E. Micronesia, after flying vast Pacific Ocean.
The Steganopodes are heterogenous Order comprising such distinctive adaptive groups as the Cormorant, the underwater swimmer, the Gannet, the deep diver from air, and the Frigate Bird, the extreme aerial flier. I could examine the samples of types which were brought back from the Galapagos Is. by Mr. Koichi Sekiguchi. They are Nannopterum harrisi (a well known flightless cormorant), Sula nebouxii and Fregata magnificens. Various parts of their skinned body were measured and analysed by taking proportions, based on proposed 'Alphabetical measurements of bird body (carcass)' (See Table 1). Nannopterum, with degenerated wings, is characteristically long-bodied, with short, somewhat broad but shallow chest, extremely long belly, long and most massively muscled pelvic and tibial parts. The neck muscles, particularly basal part of m. collis anticus, are strongly developed for underwater use. Sula has the body of log-type, with long pectoral part, (although it was rather short and broad in an example of S. sula (juv.) examined), the pelvic-tibial part with medium amount of muscles, though much less than in Nannopterum, and the fairly thick and long neck. Fregata's body is of an unique short aerial type, with no sign of aquatic adaptation. The chest is extremely deep, broad and short, the pelvic-tibial part being poorly developed. The neck is relatively (for thoracic length) short and weakly muscled. But, in spite of these adaptive peculiarities of each three type, the indexes in % of the 'Cervical length (ai)', the 'Dorso-pelvic length (ci)', the 'Thoracic length (fi)', the 'Pelvic length (qi)', the 'Pectoral length (ji)' and the 'Belly length (ei)' for the 'Basic total length (a+c)' are very constant among them, showing an Order-specific trend of the bodily proportions, except the deviations shown by Nannopterum in Pectoral and Belly length indexes owing to its special adaptation. Among other birds, a 'Standard-type trend of proportions' was noticed among taxonomically different Orders, while some others, such as the heron and the penguin, revealed very specific or aberrant proportional trends, as in the case of Steganopodes. Therefore, this method of analysis of morphological adaptive radiation way be of significance. The peculiarities in the pectoral muscles in the three Steganopodes are also described which supplement the author's previous paper (Auk, 78 (2)). The possession by flightless Nannopterum of m. pect. major. profundus (though only 1mm. thick as are other parts) which is characteristic in large soaring birds, suggests the cormorant's soaring origin. The highly soaring habit of Anhinga related to cormorants is in support of this. Only in Fregata, the medius layer of the pect. major is developed, but the very massive proximal part of the pect. major lateralis in Sula, is well suggestive of the history of development of the medius layer of Fregata.