In the spring and summer, up to 20,000 pairs of White Ibises (Eudocimus albus) congregate to breed on Pumpkinseed Island, Georgetown County, South Carolina (Christy et al. 1981, Bildstein 1987, Frederick 1987). Early in the breeding season, the adult White Ibises feed primarily on fiddler crabs (Uca spp.), which they obtain on salt marshes within 8 km of the colony. Later, when raising nestlings, they feed almost exclusively on crayfish (Cambarinae), which they catch in freshwater areas up to 50 km away from the breeding colony. As part of an ecological study of White Ibises, we required an estimate of the time and energy expended by adults that fly back and forth between the nesting island and the foraging areas. We observed flight speeds directly, and calculated the energy expended in short foraging flights. The methods can easily be adapted for use on other species. Flight speeds were measured by ornithodolite, an optical device, whose construction was described by Pennycuick (1982a). An ornithodolite run consisted of two or more timed, three-dimensional determinations of the bird's position in space. Groundspeed were then obtained by measuring the distance and time between successive positions. To convert groundspeed into airspeed, we read wind speed and direction from an anemometer immediately after each run. All of these data were recorded in digital form for later analysis. The details of operation, and analysis of the data, were essentially as described by Pennycuick (1982b), with the variations described by Pennycuick (1987a). Observations were made from an 18.5-m-high steel tower located at the edge of the North Inlet Marsh, Hobcaw Barony, 5 km north of Pumpkinseed Island. For details of the area, see Christy et al. (1981). The top of the tower supported a platform with railings and a small hut. The ornithodolite was tripod-mounted on the platform, with its eye level 16.4 m above the level of the marsh, and about 10 m above the tops of a clump of trees that surround the tower. The anemometer mast was fixed to the railings on the windward side of the platform, with the sensor head clear of the top of the hut. The observation site was within the brackish water foraging area used by the White Ibises; and observations were made on birds flying past the tower to or from the nesting area. Observations were made on 20-22 April 1987, during the initial phase of the breeding season, which is a period of nest construction, egg laying, and early incubation. Body measurements.-The measurements needed for flight calculations (mass, wing span, and body frontal area) were made on captive White Ibises taken as nestlings from Pumpkinseed Island, and maintained at the Savannah River Ecology Laboratory, Aiken, South Carolina. There was marked sexual dimorphism (Table 1; see also Kushlan 1977 and Bildstein 1987). The body frontal areas were only about 65% of the values predicted by a regression equation for hawks and ducks (Pennycuick et al. 1988). This difference is somewhat larger than in other species measured to date, and may reflect the more elongated and slimmer ibis body. Observed and calculated flight speeds.-We made 82 observations of airspeeds in steady flight (Fig. 1). Observations in which the bird was about to land were excluded. The mean air speed was 13.1 m s-1, with a standard deviation of 1.8 m s-l. We were not able to discriminate between the sexes. We estimated the minimum power speed (V,,,) and maximum range speed (Vm,) (Fig. 1). The estimates were different for males and females because of the dimorphism (Table 1). Estimated values in Table 1 were from the method presented by Pennycuick (1975). Parasite power was calculated from measured body frontal areas (Table 1), combined with drag coefficients calculated by the method of Pennycuick et al. (1988). This makes only a small difference to the calculated speeds, in comparison to the simplified method for calculating equivalent flat-plate area (Pennycuick 1975).