tached and probably of foreign origin. The nodules did not appear to be part of the coral but were so well attached that when one of them- was removed for counting it could not be arated from the coral without being broken. low sphere weighing 1 mg yielded 100,000 d/m. sep- This unashed hol- It is possible that these bodies were cystg produced by the coral itself, either. for the purpose of walling off irritating, highly radioactive particles, or that they were rapidly-growing neoplastic growths which had concentrated a great amount of radioactivity since the time of the blast. See Figure 22c. Photographs and autorediographs of Heliopora and of the above samples of Acropora collected at Bogallua are shown in Figure 22a and b. The specific activity of the Acropora was 7,000,000 d/m/g and of the nodule 100,000,000 d/m/g, i.e., 100,000 d/m/mg. After the autoradiographs were made another piece from the same sample of Heliopora was used in an attempt to complement the results with quantitative data. outer layer, about one millimeter thick, The thin the dense median por- tion corresponding to the least dense portion in the autoradio- greph, and the relatively porous central portion were separated from one another and ashed for counting. The resulting spec- ific activities were 3,400,000 d/m/g, 160,000 d/m/g and 1,000,000 a/n/zg respectively. It seems likely that the radioactivity found in the median portion lined small cavities which are present in the skeleton rather than actually being incorporated in the coralline material. aut - 94 - & wo pivbe