A new phenomenon of distribution of foreign elements in clam shells (maximum quantity at surface and along edges of the shell) is reported and interpreted. The new finding makes it possible to reconstruct the history of contamination in the waters surrounding the organism concerned. It further suggests inadequacy of previous techniques in using mollusk shells for measuring the degree of water pollution. Bivalve shells have been shown to be useful in monitoring heavy-metal pollution in the marine environment; shell composition changes as a result of changing water chemistry (Bertine and Goldberg, 1972; Goldberg et al., 1978; Miramand et al., 1980). Apparently, shells provide more sensitive and more irreversible clues for monitoring changes in ambient heavy metals than the soft tissues (Miramand et al., 1980; Koide et al., 1982). We report here on the composition of metals and sulfur in the shell, and illustrate the usefulness of shell chemistry as a tool for deterrnining the history of the chemical environment of the organism concerned. Individuals of the clams Meretrix lusoria (Roding) were collected from a shallow-water culture farm near an estuary in southern Taiwan (120°E, 23N). This species inhabits sandy beaches exposed to air for a few hours during low tide. These clams are found in an anoxic environment with gray-to-black organic matter. Shell samples were first separated from living soft tissues, washed with distilled water, exposed to ultrasonic radiation for 2 h to remove possible foreign materials, and then air-dried prior to scanning electron microscopy (SEM). Shells of 3 to 4 cm in width which represent 8 mo to 1 yr of growth were used. Point-count analyses by SEM (JSM 35-CF), operating at 15 to 25 KV, coupled with energy dispersive analysis of X-ray (EDAX) technique, were made on the Q Inter-Research/Printed in F. R. Germany outer surface, as well as on intermediate and inner layers exposed in the fractured clam shell. Sample size was estimated to be several microns. Occasionally, an electron beam was scanned on a specific area (e.g. 3 X 3 pm) in order to control the measurements. The distance from hinge to the measuring spot was estimated on the magnified electron image of the shell. In the inner and intermediate layers only Ca was detected, X-ray diffraction (CuK,) analysis indicated aragonite to be the major substance in the intermediate layer. Significant amounts of Ca, S and Fe, with minute amounts of Cu, Zn, C1 were detected in the outer layer (periostracum). The significant amounts of Fe and S on the shell are expected since the habitat of the clam, similar to other subtidal anoxic marine sediments, contains abundant amounts of soluble sulfide (H,S and HS-) and acid-labile sulfides (FeS) (Howarth and Giblin, 1983). Ca. S, Fe, Cu and Zn were analyzed next from the umbo to the posterior edge of the shell (following the growth direction), and the relative amounts of S, Fe, Cu, and Zn over Ca were estimated from the EDAX peak height (i.e. counts). Ca was used as reference because it is the major constitutional element of the shell. Fig. 1 and 2 show a general trend of increasing S, Fe, Cu and Zn towards the outer edge (younger portion) of the shell. This is in contrast to the previous concept, i.e. that pollutants accumulate with increasing age, which implies that the older portion should be richer in metals than the younger portion of the shell. Surface sorption processes are assumed to be responsible for the accumulation of metal pollutants in the shell (Miramand et al., 1980). Such assumption cannot be used to explain our data either. If surface sorption processes were the main operating mechanisms, we should expect a uniform distribution of for188 Mar. Ecol. Prog. Ser 18: 187-189, 1984